Advanced Computational and Communication Paradigms

Elephant detection system is a subject of interest these days. Expert systems may be designed to enhance the efficiency of these systems. Artificial Neural Networks (ANNs) may be implemented in order to accomplish the task and research may be carried out in this field. In this paper, a method is proposed which detects the presence of elephants near the railway tracks and instantly activates the simulator to drive away elephants from the railway tracks. The simulators may be virtual fire or cracker sound. Elephants are scared of bee sound. So, bee sound may also be used as simulators. The proposed ANN used here is an unsupervised type of ANN, where a weight detection algorithm has been designed to get rid of the ambiguity.

In this work, an attempt is made to segment endocardium and epicardium of left ventricle in normal and myocardial ischemic cardiac magnetic resonance (CMR) images using local region with optimized boundary driven level set. Myocardial ischemia (MI) is a cardiac disorder that results in deprivation of oxygen supply to myocardium and can be analyzed by study of abnormal anatomical changes in CMR. This study is carried out on short-axis view CMR images from Medical Image Computing and Computer-Assisted Intervention (MICCAI) database. The edges are computed by simple Laplacian and Laplacian of Gaussian (LOG) operator. LOG is optimized to obtain enhanced edges of endocardium and epicardium. The quality of edge is validated with edge preservation index (EPI) and gradient magnitude similarity deviation (GMSD) measure. Local region with optimized boundary (LROB) driven level set is utilized for simultaneous segmentation of endocardium and epicardium of left ventricle in CMR images. The results are compared with local region (LR) driven and LR with LOG-driven level set. Further, the efficacy of the segmentation is validated with different similarity measures. The optimized LOG image visually shows better endocardium and epicardium contours. Optimized LOG with a higher EPI and lower GMSD provides better enhanced edges compared to Laplacian and LOG functions. The computed similarity measures for LR with LOG-driven level set are significantly higher compared to LR-based level set for segmentation of endocardium and epicardium. Further, LROB-driven level set shows higher similarity measures than LR with LOG-driven level set. Thus, LROB-driven level set provides better segmentation accuracy for epicardium and endocardium of left ventricle than LR-based level set and LR with LOG-driven level set. The efficiently segmented endocardium and epicardium could aid the diagnosis of myocardial ischemia with their ability to quantify anatomical changes in LV.

Recent advances in digital image processing, computer vision, etc., have led to many approaches to the classification of fine art painting. The focus was made to develop an automatic painting classification system based on their genre. The expanding database of digital painting images makes it imperative to develop an automated method to annotate paintings with metadata as painter, genre, painting tool used, style, etc., so that problems like image retrieval, searching, organizing, and artistic recommendations become convenient and efficient. The aim was to classify painting database into five genres. The database consisted of 1229 digital painting images. The method adopted for the task was feature extraction from the images, using each feature individually for classification and then combining the features based on weighted majority voting to form an ensemble classifier. It was observed that Local Binary Pattern (LBP) was the best performing feature. The ensembled model achieved an accuracy of 80.41%. We have included analysis of our work and have discussed the performance of the various features deployed for the painting classification task.

An improved image encryption scheme that uses affine transform and asymmetric keys in Fourier transform domain has been proposed. The scheme is validated for grayscale images through numerical simulation using MATLAB 7.14. In the proposed scheme, the decryption keys are different from the encryption keys and are obtained by phase truncation Fourier transform method. The performance of the scheme is evaluated in terms of the metrics such as correlation coefficient, mean-squared error, and peak signal-to-noise ratio. We have carried out the sensitivity analysis relative to the affine transform parameter, which serves as an additional security feature. The robustness of the scheme is demonstrated by showing its resistance against noise and occlusion attacks. Since this is the first study that uses affine transform in the phase-truncated Fourier transform based asymmetric cryptosystem, it provides a new scheme for image encryption with enhanced security.

This chapter is focused on the implementation aspects of adaptive feedback canceller algorithms and their computational complexity when reducing misalignment and convergence rates. When an adaptive algorithm filter was used for modeling the acoustic feedback, there was wide misalignment due to a fixed step size. Through the use of the prediction–error method (PEM), the bias in the algorithm for an adaptive filter was reduced. The PEM used a variable step size and a full range of adaptive filters were used as a trade-off between the misalignment and the convergence speed. Various performance measures were considered in order to study these algorithms: misalignment, maximum stable gain, added stable gain, and the algorithm execution time. The disadvantage of misalignment and convergence when changing step size has been addressed using a new algorithm that has automatic step size adjustment. This new algorithm demonstrated effectiveness in controlling misalignment. The findings reveal that the misalignment, maximum added gain, and added stable gain improved with the use of the new adaptive filter algorithm. Despite this, the PEM did not satisfy user requirements and so a new system named AFC-PEM MPVSS is proposed. Furthermore, work has been done to measure the quality of signal.

Skin cancer is one of the atrocious diseases observed in the western part of the world due to exposure to the ultraviolet (UV) rays approaching from the sun- and human-made tanning beds. The survival rate of skin cancer is very high if it is detected at an early phase and treated surgically. To detect it, preprocessing of affected skin lesion images is essential. Here, we are representing a technique for preprocessing of skin lesion via contrast enhancement followed by anisotropic diffusion and sigmoid function. In this method, we critically normalized the skin lesion images followed by removing Gaussian noise and preserving some feature by anisotropic diffusion. For more improvement of it, we applied sigmoid function in the spatial domain of the skin lesion image. Here, we critically consider different parameters of anisotropic diffusion and sigmoid function. This innovative method has been successfully used in various low contrast affected skin lesion images. All most in all the cases, it gives the satisfactory results in terms of MSE PSNR, and SSIM values. This proposed method can be used to improve the quality of low contrast images in medical science, satellite imaging, and different industries. The said technique can be applied successfully in various applications.

Interval data are found in many real-life situations involving attributes like distance, time, etc. Mining closed frequent intervals from such data may provide useful information. Previous methods for finding closed frequent intervals assume that the data is static. In practice, the data in a dynamic database changes over time, with intervals being added and deleted continuously. In this paper, we propose an incremental method to mine frequent intervals from an interval database with n records, where each record represents one interval. This method assumes that intervals are added one by one into the database and each time an interval is added to the database, our proposed method will mine all the newly generated closed frequent intervals in O(n) time.

The present work proposes an unsupervised method for identifying named entities from a corpus of crime reports containing information on crime against women in Indian states and union territories and subsequently discovers substantial relations among the identified named entities. For discovering the relations, different types of entity pairs have been chosen and similarity among them has been measured based on the intermediate context words. Depending on the similarity score, clustering technique has been applied that forms several clusters of named entity pairs. Each cluster consists of a representative entity pair and relation of that representative pair corresponds to the relation of the whole cluster formed, leading to the relational labelling of the clusters. This method does not desire any time consuming richly annotated corpora and the result with high F-measure values depicts the effectiveness of this method.

Achievement of high-capacity data hiding using a digital media is an important research issue in the field of steganography. In this paper, we have introduced a novel scheme of data hiding directly within the video stream using bit plane slicing through (7, 4) Hamming code with the help of shared secret key. In the proposed scheme, a secret logo image is embedded within the cover video stream for authentication and ownership identification through Hamming code based video steganography. Each frame of secret video has been separated into individual three basic color blocks (R, G and B) and then partitioned into (3 × 3) pixel blocks. After that, each color block is sliced up into 4 bit planes starting from LSB plane. The pixels’ positions of cover images are randomly selected by Pseudorandom Number Generator (PRNG) using a shared secret seed value and data embedding performed using (7, 4) Hamming code. As a result, 36 bits secret data can be embedded within a (3 × 3) pixel block which is almost eight times greater than Ramadhan and Khaled’s scheme (Systems, applications and technology conference (LISAT), 2014 IEEE Long Island, 2014) [1]. Here, we achieve a high payload with good visual quality stego video. Furthermore, the video compression is lossless so the video file size is strictly preserved for post-data embedding.

Research on biophotons imaging techniques has now advanced to a level where it is possible to capture biophotons in true color two-dimensional image form. On the other hand, biophysicists have proved that thought process has definite impact on biophotons emission from human biofield. These two developments are enough to predict the prospects of development of a next generation security system to protect the vital sensitive installations and precious human lives without the need of the present day frisking and scanner aided security system. This paper is devoted to describe a conceptual design for such a security system. To demonstrate its functioning, true color images of thought forms from Theosophical Society literature are used. Most of the images were used in experiments for recognition of their patterns in our three recent papers where it was shown that thought forms had patterns of human behavior which matched with the comments attributed to the images in theosophical texts. These images were captured by theosophists using their power of clairvoyance, and published nearly 100 years ago. The three factors, viz., (i) patterns recognition in our papers, (ii) advances made in biophotons imaging techniques, and (iii) changes in thought process resulting in changes in human biofield, are the essential components on which a conceptual design of the next generation security system is proposed.

In this paper, we have proposed a new data hiding method for grayscale image based on the block pixel intensity range. This work is inspired by the Varasaki’s data hiding method. Here, the cover image is symmetrically divided into blocks of size $$3 \times 3$$3×3. The intensity range of each block is computed and this intensity range is divided into four zones. For each block, the neighboring pixels are modified to place into a zone according to the message bit so that the message can be reconstructed losslessly. Our proposed method gives better quality stego image than LBP-based data hiding.

This paper presents the development of a fatigue detection system that would be capable of detecting an individual’s level of alertness through live video acquisition. The approach is to build a nonintrusive system that uses computer vision methods to localize face, eyes, and iris positions to measure level of eye closure within an image, which, in turn, can be used to identify visible eye signs associated with fatigue leading to a sleepy state. The aim here is to detect this state early enough and issue a warning or alert in the form of an alarm.

A bit level block cipher based symmetric key cryptographic technique is proposed here. Entire plaintext file reads two characters at a time and according to the binary representation of ASCII value of these characters, 16-bit blocks are created and they are represented in a 4 × 4 spiral matrix. Then using random shuffling of spiral matrix and multiplicative inverse, 8-bit ciphertext blocks are generated for each 16-bit plaintext block. So here 50% compression is achieved in terms of size of ciphertext file. Apart from compression, RSSMMI technique has several advantages like formation of symmetric key dynamically and randomly, security, and equal applicability of this technique for large number of files of almost any type.

A concatenated approach which utilizes the strength of Particle Swarm Optimization (PSO) and Grey Wolf Optimizer (GWO) is proposed for odor source localization by a team of mobile robots. Odor plume is modeled by using the Gaussian distribution. Robots continue random search within the workspace to locate the plume. When one of the robot enters in the vicinity of plume, robot’s new positions are calculated by applying concatenation of PSO first then Grey Wolf Optimizer second and vice versa. In order to prevent getting stuck at local minima, concept of search counter is used. Proposed approach is compared with Refined Hybrid PSO and the simulation result shows the validity of the proposed approach over the other.

Distance feature has great significance in recognizing facial expressions. Identifying accurate landmarks is a vital as well as challenging issue in the field of affective computing. Appearance model is used to detect the salient landmarks on human faces. These salient landmarks form a grid on the human face. Distances are determined from the one landmark point to another landmark point in grid and normalized. A novel concept of corresponding stability index is introduced which eventually is found to play important role to recognize the facial expressions. Statistical analysis such as range, moment, skewness, kurtosis, and entropy are calculated from normalized distance signature to supplement the feature set. This enhanced feature set is supplied into a Multilayer Perceptron (MLP) to arrive at different expression categories encompassing anger, sadness, fear, disgust, surprise, and happiness. We experimented our proposed system on Cohn-Kanade (CK+), JAFFE, MMI, and MUG databases to training and testing our experiment and establish its superiority performance over the other existing competitors.

This paper presents the comparison between performance of an Interval Type-2 Subsethood Neural Fuzzy Inference System (IT2SuNFIS) (Sumati et al. Expert Syst. Appl. (Elsevier) 60:156–168, 2016 [27]) using two different similarity measures, implemented on a parallel platform. The inputs to both the system are fuzzified using interval type-2 fuzzy sets (IT2 FS) with Gaussian primary membership function (GPMF) having identical mean but different variance. The signal aggregation of type-2 based activation is performed using product operator. A comparison between subsethood and mutual subsethood has been carried out experimentally, showing better contrast handling capacity of later over former. Consequently, the performance of both the models is tested and compared on a benchmark application of Mackey-Glass time series prediction. It is observed that the performance of mutual subsethood based IT2SuNFIS is better than subsethood-based IT2SuNFIS in terms of result accuracy.

In this paper, we present an efficient method of classification of sentiment in social media texts, each consisting of single or multiple sentence(s) that most of the time includes pop culture texts. In our experiment, we present an architecture that derives vector representations (i.e., word2vec) of the phrase level sentences. We use some combination of quantitative and qualitative methods for training a recurrent neural network with empirically cross-validating gold-standard array of lexical features, which are precisely synced with sentiment in microblog-like pieces. We leverage a new technique that expands upon previous works on sentence-level lexical sentiment classification, using recurrent fuzzy neural network and use it jointly with a Recursive Neural Network to further improve the classification. We have tested our algorithm against the other state-of-the-art methods on various platforms for better demonstration of our experiment with satisfactory and competitive results.

Internet of Things (IoT) hegemonies all other technological development the world has seen so far. 6LoWPAN is emerging as the next generation protocol of IoT. 6LoWPAN protocol enables the resource constrained embedded device to connect to the Internet through IPv6. Routing Protocol for Low Power and Lossy Networks (RPL) is specifically designed as a routing protocol for resource constrained device and it is adaptable with the 6LoWPAN. There are many attacks which make RPL insignificant to achieve its desired objectives. Neighbor attack is momentous and is capable of disrupting possible routing path. In this paper, we proposed intrusion detection system which can detect neighbor attack in RPL protocol and a secure root process to prevent the effect of attack on this protocol. The IDS is incorporated by considering location information and received signal strength to identify malicious node. We have also incorporated secure root process which can rectify disruption in routing path after detecting attacks. This method can act as an optimum method for resources constrained environment.

The identification of features responsible for smell of a molecule has been a long-standing challenge. We use cheminformatics and opinion dynamics based optimization algorithm to identify feature subsets of a molecule, which can predict how pleasant a molecule will smell. We have also compared it to standard feature selection techniques. The features identified reveal that three classes of features are primarily responsible for pleasantness. The work may open up some innovative inroads into feature identification and their physical understanding into the olfactory stimulus-percept problem.

Technology has evolved a lot in the last decade, and various devices have come up for assisting us in our day-to-day life. There has always been a need for simplifying the User Interfaces (UI) of such devices so that they can be easily interacted with, and a speech based UI can be a potential solution. Speech recognition is the task of identification of words from voice signals. Every language consists of a set of atomic sounds called Phonemes which builds up the entire vocabulary of that language. Speech recognition in Bangla is rather a complicated task due to the complex nature of the language like the presence of compound characters. In this paper, a Bangla Phoneme recognition system is proposed towards the development of a Bangla Speech recognition system based on Line Spectral Frequency-Grade (LSF-G) features derived from standard line spectral frequency values. The system has been tested on a Bangla Swarabarna Phoneme dataset of 3290 clips and an accuracy of 94.01% has been obtained with an Ensemble learning based approach.

In MANET, wormhole link creates an illusion in such a way that two remote regions are directly connected through nodes which seems to be neighbors; however, these are actually distant from one another. The attackers using wormhole can easily manipulate the routing priority in AODV to perform eavesdropping, packet modification, or packet drop. Hence, a two-phase wormhole link detection procedure in AODV routing protocol is proposed to identify the malicious link for avoiding such erroneous transmission. Initially, the round trip time (RTT) and corresponding round trip bit transfer (RTBT) of each link are determined. Here, if RTBT of any link is greater than a dynamic threshold value (RTBT_TH), then such victim link is marked as a suspicious link. Next, the amount of power required to send a packet of certain size by each node is obtained to verify whether the transmission power of suspected links is reasonably high compared to other links. Various experimental results are carried out to validate the proposed work as well as to show an improvement obtained by the proposed approach in terms of several performance metrics.

In this paper, image contrast enhancement is achieved by combining together the local–global image statistics and multiscale morphological filtering (MMF). The proposed method has been executed on two different sets of images, and the result has been compared with that of some existing standard methods, namely histogram equalization (HE), contrast limited adaptive histogram equalization (CLAHE), and multiscale morphology in order to have an outlook on the relative performances. The experimental results manifest that the proposed method produced results superior to the methods compared.

Cloud computing offers to its users, in theory, infinite computing through utility computing. Scheduling tasks in heterogeneous resources poses a formidable challenge albeit an increase in available computing capacity. A class of tasks termed as Bag-of-Tasks (BoT) is a predominant workload in any large-scale distributed system. We propose an agent-based approach, BISA (BoT Intelligent Scheduling agent), which chooses an appropriate scheduling heuristic and in time learns which policy will provide the most optimal schedule. The agent-based framework is presented and the cloud environment is simulated in CloudSim using threads. The working of the BIS agent is presented, and results from the training phase are divulged and discussed. The results obtained show the framework presented could provide near-optimal solution in minimizing the makespan of a BoT using the most appropriate scheduling heuristic in a given scenario.

The Hodgkin–Huxley (H-H) model of axonal membrane is one of the most inspiring and popular neuron models. Many variants of this model are present in current literature. However, this model does not encompass the post-synaptic membrane response to synaptic functions that are responsible for the dynamic membrane behavior. The paper thus presents an extended version of the H-H model that incorporates a “conjoint” synaptic model representing the complex synaptic activities and describes how the post-synaptic membrane behaves in presence of a synaptic entity as an input. The simulation is performed in MATLAB environment and the results are presented in the form of graphs.

Due to the unstructured nature of review text, it is very hard to develop an automated opinion mining application to compare various product models based on their various aspects to make a purchase decision. Over the year, various data mining techniques have been proposed to extract aspects of the products. In this paper, we have proposed a technique based on the nonnegative matrix factorization to extract aspects of a product category. Performance of our proposed method has been compared with a very popular aspect extraction technique based on probabilistic latent semantic analysis. We have also given a comparison between common aspects of a particular model under a specific product category from various manufacturers. These comparisons are based on the sentiments expressed by the users on these aspects. These sentiments expressed in various aspects have been extracted using an unsupervised technique.

This part of the research paper emphasizes on the detection of behavioural pattern from the extracted prosodic and geometrical features using human-gait-speech data. The clusters of these above-extracted features are mapped for the detection of behavioural pattern using soft-computing technique and c-means clustering method. Here, only four features of human-gait and four features of human-speech pattern are used for the formation of clusters. These clusters are mapped to close vicinity with minimum distance measurement, which in return is helpful for the proper classification and decision process, with a positive outcome for the detection of behavioural pattern. The mapping has been done with proper mathematical analysis over each feature of human-gait-speech pattern. The four prosodic features (extracted from human-speech pattern) are speech duration, speech rate, pitch and speech momentum, whereas the four geometrical features (extracted from human-gait pattern) are step length, walking speed, energy or effort and gait momentum, which are clustered. Here, five different natural languages (Hindi, Bengali, Oriya, Chhattisgarhi and English) have been employed for the completion of this part of research, when the subject is talking while walking. The classification process is being carried out with the help of a human-gait-speech-model (HGSM), using Baye’s theorem and support vector machine of artificial neural network. The mapping process has been carried out using adaptive-unidirectional-associative-memory (AUTAM) technique with an acceptable limit. The decision process for the detection of behavioural pattern has been done using revolutionary algorithm called genetic algorithm. Three behavioural patterns have been detected with three class-based moments: happy moments, normal moments and sad moments. An algorithm, called behavioural pattern detection algorithm using human-gait-speech pattern (BPDAHGSP), has been proposed. The complexity measures have been done, and the performance of the overall authentication system has been found very helpful for promoting global biometrical security system using nominal number of features.

This paper formulates a genetic algorithm to judge the mutation of path for packets in the network. The predefine routes for network lines to transfer packets, transportation links in liquids and bearer links are static phenomenon. The algorithm discovers mutation in prescribed path due to some breakage in the communication link. The breakage occurs due to natural calamity or for any technical issues. At this juncture in the mid-way dumping or returning, the packet/package is not wise. Here, a new path has to be redefined from the current node and the packet has to be delivered to the destination. It is good to have delay in delivery instead of been dumping. Because resending the packet is far more time consuming, this makes the phenomenon more dynamic and less time-consuming. The concept is also useful in robotics for path selection. Simulation for path is performed using MATLAB. Results are similar to Dijkstra’s algorithm. Dijkstra’s algorithm is a static algorithm, whereas the proposed algorithm is dynamic one.

In this article, a facial expression recognition technique has been proposed. In this method, initially, a face image is segmented into different sub-images like left eye, right eye, mouth, nose, etc. Then, two-dimensional principal component analysis (2DPCA)-based transformation is performed on each sub-image separately to extract the features. Lastly, classification operation is made to categorize the expression of faces. To demonstrate the effectiveness of the proposed method, results are compared with other existing PCA- and 2DPCA-based methods in terms of overall classification accuracy. Results for the proposed method are found to be encouraging.

Active vision plays an important role for many applications using better understanding of our surrounding environments. Using active vision, it is possible to gather new knowledge or changes in the environment by autonomously moving of eyes, and nearly all animals have this routine biological capability including human. In active vision, it is needed to repeatedly move the gaze from one fixation point to another fixation point to obtain high-resolution image and track the object robustly. The main reason to keep gaze still during fixations is the need to avoid the blur that results from the long response time of the photoreceptors. In this paper, we propose an active vision system which is constantly tracking an object by moving its gaze. In our proposed approach, human–eye movement technique, i.e., saccade and smooth pursuit mechanism, has been used to obtain high-resolution image on the sensor and track that particular object in real time.

While robotic manipulators are becoming a common sight in today’s industries and fast paced production lines, it is becoming difficult to develop foolproof methods for automation of these manipulators, owing to their geometric and structural variety. Creating a common algorithm for these manipulators would help in setting a base standard for their automation. Trio Motion coordinators are most widely used for robotic manipulators in recent times. The objective of this paper is to create a simple interface based on Visual Basic programming language to coordinate directly with the robot’s motion coordinator by bypassing all other programming methods which are otherwise needed for sending commands to the robot. This interface can be easily adapted for further tuning methods and also for more or lesser degrees-of-freedom robotic manipulators. MATLAB has been used for detecting the weld line in the image using image processing techniques. A suitable artificial neural network has been used to give forward kinematic solutions with image coordinates as the input.

The imbalanced load between clusters is a key issue in distributed computing environment. All existing dynamic load balancing algorithms are post-active, as balancing activities start after system turn into imbalanced state. The better approach is to design pro-active load balancing algorithm which starts working with scheduling algorithms. It helps scheduling algorithms to schedule incoming jobs in such a way that system remains in balanced state. The pluggable to scheduler dynamic load balancing algorithm (P2S_DLB) is designed and evaluated over priority scheduling algorithm in our previous research work. The P2S_DLB is pro-active dynamic load balancing algorithm. In this paper, we have measured and evaluated the performance of P2S_DLB over First Come First Serve (FCFS), Shortest Job First (SJF), and Earliest Deadline First (EDF) scheduling algorithms. The experimental result shows that algorithm has improved the cluster utilization and decreased the imbalance level of distributed computing environment in case of all the three scheduling algorithms.

The recent advancements in Hadoop MapReduce scheduling techniques have demonstrated significant outcomes. The continuous tradeoff between the data-job locality and synchronization results in the higher efficiency for the framework. Thus, large number of scientific and enterprise applications have adopted the parallel and synchronized mechanism through Hadoop framework. However, with this adaptation, a large number of datacenter-based nodes are been deployed, significantly causing the increase of energy consumptions. Henceforth, the demand of the recent research is to enhance the overall efficiency of Hadoop jobs and to decrease the energy consumption without degrading the performance. The recent advancements have demonstrated by many strategies by improving the Map and Reduce job allocation techniques; conversely, the same improvement can also be achieved through multilevel queues. Hence, this work constitutes the multilevel queue with custom load balancing to demonstrate the improvement in overall performance of Hadoop job scheduling. The work results in a significant improvement of Hadoop jobs in terms of execution times and energy consumption.

This paper proposes a semi-automated tool to build ontology from text. The tool consists of an analyzer to parse the given text and a mapper that maps NLP triple to RDF triple under user supervision. The resulted RDF triple is then validated through “triple validator” for its existence in the ontology. The triple is augmented to the ontology if it does not exist. System learns during this process and provides better mapping suggestions with time, making ontology building faster.

Medical images are poorly illuminated and often suffer from low contrast so it needs enhancement before further processing. In this paper, we have introduced an ideal approach for medical color image enhancement which is based on type-2 fuzzy set with unsharp masking based post-processing. The proposed technique has been tested in various poor contrast medical color images and found the results superior to the other traditional state-of-the-art algorithm.

In today’s entertainment industry which is becoming increasingly competitive, music producers, record labels are striving hard to give the next big hit song and capture the viable music market. We propose to formulate factors and dependency variables which would form the basis of hit prediction in big data environment. The audio features such as pitch and tempo are analyzed in tandem with statistical parameters such as root mean square energy, slope, period frequency, and musical topographies like acousticness, loudness, and instrumentalness. This is a preliminary experiment where the simulated ratings are paralleled with ground truth obtained from Billboard, Spotify, and Radio Mirchi rankings over a period of 5–10 weeks. The paper covers a wide area of tracks from USA, UK, Australia, and India, and proposes to arrive at a consensus to the factors contributing to the success of the track according to their topography. While acousticness plays a vital role in US and India countdowns, British are highly influenced by the danceability and the energy components of the track. The paper provides a cushion for hit prediction classification of musical tracks in big data applications.

Dynamics of the dust-ion-acoustic anti-kink waves (DIAAKWs) in an unmagnetized multicomponent electron–positron–ion (e-p-i) dusty plasma consisting of negatively charged static dust, positively charged positrons following Maxwellian distribution, inertial ions, and q-nonextensive distributed electrons of two distinct temperatures are studied. The ion kinematic viscosity is intended and the reductive perturbation technique (RPT) is employed to obtain the Burgers equation. Using Galilean transformation, the Burgers equation is reduced to a system. Analyzing vector fields and corresponding potential energy of the dynamical system, the stability and instability of the equilibrium points are discussed. The physical parameters $$q_1, q_2, \mu _1, \mu _2,\sigma _1,\sigma _2, \eta $$q1,q2,μ1,μ2,σ1,σ2,η, and U affect significantly on the characteristics of the DIAAKWs.

With the advent of the widespread use of Internet for communication, there is an increase in exchange of personal secured information via the Internet wherein a large variety of sensitive information is exchanged between the end users which affect the privacy and safety of the data during storage and communication. Visual cryptography, an example of a secret sharing scheme encrypts the secret digital information into a number of shares wherein the decryption is performed by overlaying the shares generated and by utilizing the HVS (Human Visual System). General access structures efficiently hides the data by defining an access structure of qualified sets which only can produce the hidden information. Since no transmission is noise free, quantum neural network is used to extract the original information even when the information is not clearly visible to the human eye. The paper proposes a methodology to conceal multiple secret in a pair of shares and the Q’tron network to improve the security of information systems.

Distributed approach in a network is a prime attribute to achieve quality throughput. Many real-life infrastructures share such distributed network structures. Recently, researchers are focusing on different prime attributes of these distributed networks and meaningfully analyzing them to retrieve essential information toward throughput enhancement. These structures exhibit different constructs. Some of those are static and some are dynamic. They also contain strategic groups within it. Appropriately, identifying these groups is the key essence of community detection. Present work applies a novel mechanism for graphical analysis of network structures to detect overlapping communities. Experimental findings and comparative analysis with existing methods show efficacy of the present algorithm.

The existence of chaotic and periodic structures of dust-ion-acoustic nonlinear waves (DIAWs) is reported through dynamical system approach in a dusty plasma with dust-ion collision effect. Employing the reduction perturbation technique (RPT), the damped modified Korteweg-de Vries (DmKdV) equations are derived. The nonlinear wave phenomena of perturbed DmKdv equation is studied in presence of an applied external force. Chaotic behavior is found in some critical composition for the said perturbed system. It is seen that dust-ion collisional frequency acts as a controlling parameter to chaos and changes the structure from chaotic motion to periodic motion.

Cattle crossbreeding improves the breed quality and milk yield of the progenies and hence improves the nutrition security and livelihood of the farmers. It is a complex data-intensive process and requires the animal breeder to maintain 7–10 forms having 230 data fields. Each data field may have 5–100 options to be filled in, which is a challenging task to be done on a small screen mobile phone by low-literate breeder. We followed an iterative design process to improve the overall user experience. We interacted with the end users, understood their environment, and analyzed the existing methods. We analyzed all the forms and data recorded by them in last 3 years. We analyzed major parameters like cattle species, pregnancy stages, cattle locations, etc., and based on the unique combination of these values, we defined various “contexts”. For every context and data field, we identified “context-sensitive” vocabulary (word library) extracted from the available records. So, for the selected data field, we recommended the values based on the context, from the vocabulary. During further iterations, the user choices were also learnt, and values substituted by the users, if any, against the recommendation, were added to the vocabulary. We added autocomplete suggestion feature to make it more user-friendly. We used the word usage count to rank the suggested words with the words used maximum ordered first in the recommendation. This design approach minimizes the user data entry efforts, improves the speed, and reduces error, especially by the low-literate animal breeders who are not quite comfortable with text typing on smaller screens.

The problems related to different sorting methods over a signed permutation is initiated and raised by comparative genomics. Especially, translocation in comparative genomics is often deal with large-scale mutation between species. Related to that Hannenhalli has proposed the first polynomial time algorithm in 1996 for calculating translocation distance between genome. Here our algorithm improves algorithm given by Bergeron, Stoye and Mixtacki for sorting by translocation problem. Not only that, our algorithm is a betterment of the algorithm given by Tannier et al.

Vance (Am. Nat. 112:797–813, 1978) modeled a food web consisting of two prey and one predator with competing prey species. Gilpin (Am. Nat. 113:306–308, 1979) explored that for biological feasible parameter values, Vance’s model can produce chaotic dynamics. In the present paper, we consider a modification of Vance’s model by incorporating predator switching into the model for the same set of parameter values considered by Gilpin. We observe that if we increase switching intensity above a threshold value, then the system becomes stable from chaotic oscillations and enhances the persistence of the system.

Security in the recent technology world is an important issue to be taken care of and to be encountered with preventive measure on various aspects. Secret sharing is a technique used in visual cryptography (VC) which divides the image into many shares and by overlapping those shares original image is revealed, but it creates a threat when an intruder get shares with which the image is going to be decrypted easily. However, in the present project work, an extremely useful bitwise operation is performed on every pixel with the help of key. The key is provided by new concept of sterilization algorithm. Initially, Red, Green, and Blue channels get separated from image and are going to be encrypted on multiple levels using multiple shares; it converts an image into unreadable format and by combining all the shares in proper sequence the original secret image is revealed.

This paper presents a method based on mathematical morphology for enhancing the contrast of a color image using total ordering with Hilbert space filling curves. The method defines a total ordering of three-dimensional (3D) space (RGB space) using Hilbert 3D space filling curve and then applies morphological operators on the color image to obtain the contrast enhanced image. The output obtained through the above method has been compared with the outputs obtained through marginal morphology and vector morphology based on a distance measure. Experimental results show the efficiency of the proposed method in terms of enhanced contrast and better time complexity.

Management of religious crowd gatherings is a complex and essential task. Lack of planning and management has resulted in many unfortunate incidents and deaths in the past. Though local authorities deploy crowd management personnel to avert and control this crisis, some unforeseen events still lead to crowd panic resulting in a sudden rush of individuals. In this paper, we have utilized Agent-Based Modelling technique to model a near to real scenario of crowd gatherings at Alopi Devi temple of Allahabad, India. We have attempted to incorporate certain heuristics such as closing and opening of exit and entry gates which can be beneficial in controlling crowd-related disaster. These heuristics are then simulated over panic situations to analyse their effects in terms of numbers of victims. We have used NetLogo, an Agent-Based Modelling tool, to design and simulate our model. The simulation results establish the applicability of our methodology.

Recommender systems attempt to predict the future behavior of a particular individual based on her past preferences. Today any individual may have more than one profile that he/she maintains on various websites, and leveraging all this data on the preference of an individual from various domains (cross-domain) can help us in making better user models that can be used to make better and improved recommendation. A cross-domain recommender system thus aims to improve the recommendation of a target domain extracting and using the metadata from many source domains. Building scalable recommender systems is always a challenge in the presence of Big Data, and this is compounded for cross-domain recommenders. In this paper, we aim to tackle the problem of scalability in cross-domain recommendations. We have performed various experiments to divide the datasets into smaller clusters and then running a recommendation algorithm using the attributes in the dataset to return the best recommendations.

This paper proposes a document classification model using feature learning (Coates, Demystifying unsupervised feature learning, 2012) [5] approach based on semantics of the documents. In the learning phase, basic vocabulary (BV) for each document class consisting of nouns has been created by proposing a novel approach. The classification phase searches unique words in the BVs and if found, the corresponding sentence becomes a basic sentence (BS). A tree with unique words of the BS is inserted in the respective forest. Associated words of the children are used to continue the tree formation process until no new node is generated in the tree. Finally, we assign the test document to a class which has a clearly dominant percentage of sentences in the respective forest. The proposed algorithm is compared with various feature-based classification models and satisfactory performance has been observed.

Although various quantifiers of software component reusability have been proposed, these metrics have seldom been utilized in existing literature to analyze source code file reusability transpiring within a single product family. Such metrics can be effortlessly employed to develop reuse prediction models which can support the software practitioners in obtaining information regarding the total cost involved in developing a novel version of a prevailing software or upgrading an existing software version by estimating the total reusable code files in advance without being compelled to scrutinize the complete codebase. In view of this, this research work aims to examine the efficacy of seven meta-classification techniques in the development of such reuse prediction models on four software datasets constructed from four successively released versions of software using appropriate reuse metrics. We also evaluate the predictive performance of these meta-classifiers against the statistical technique of logistic regression and rank these techniques using the Friedman statistical test.

The finding in this chapter is that two oblique solitary waves collide in different angles. The KdV equations are derived, and the collisions figures are drawn with different angles. The multicomponent plasma here is constituted by +ve and −ve charged dust and hydrogen ions, and remaining components are hot and cold electrons. The solutions of the KdV equations are calculated by using Hirota’s method. The phase shifts are derived. The effects of density ratio and temperature ratio have investigated on the phase shifts. There are many areas to use these results like study on mesosphere, modern research in plasmas, etc.

Nowadays, signed product cordial labeling and total signed product cordial labeling play an important role in graph labeling. For the graph $$G=(V,E)$$G=(V,E), all the vertices are label by 1 or $$-1$$-1 with the restriction that the difference between total number of vertices labeled by 1 and $$-1$$-1 is less than equal to 1, where edge of the graph $$G=(V,E)$$G=(V,E) labeled by the multiplication of the label of the end vertices also with the restriction that the difference between total number of edges labeled by 1 and $$-1$$-1 is less than equal to 1. We investigate some interesting result under the above labeling scheme on some new type of complex graph. In this paper, we apply signed product cordial labeling and total signed product cordial labeling on the graph obtained by Cartesian product between balanced bipartite graph $$K_{n,n}$$Kn,n and path $$P_r$$Pr. We have shown that the time complexity of the algorithms is superlinear.

Named Data Networking (NDN) is an instantiation of Content-Centric Networking (CCN) that focuses on the limitation of the current working IP-based Internet architecture. Like any other networks, NDN suffers from many threats that include denial-of-service attack (DoS) or distributed Dos (DDoS). To exhaust the NDN router resources or content provider, DDoS attack can be triggered. This survey paper addresses the interest flooding attack that is one of the DDoS attack types in NDN which tries to overflow the Pending Interest Table of the NDN router and exponentially decreases the NDN router performance.

In this paper, the three-parameter characterization of intuitionistic fuzzy sets and normalized hamming distance are employed to develop mathematical programming-based TOPSIS techniques in interval-valued intuitionistic fuzzy settings. A pair of linear fractional programming models are generated which are simplified for producing intervals to measure relative closeness coefficients of alternatives. Possibility degree matrix is obtained by pairwise comparisons of closeness coefficients and optimal degrees are estimated for final ranking of alternatives. The proposed approach is illustrated through a numerical example.

In Formal Concept Analysis, stability is an important utility measure to rank concepts. However, computation of stability is considered to be a hard problem. Efficient algorithms having good bounds to estimate stability holds promise. In this paper, an effective graph-based technique is proposed to estimate stability. Our estimation algorithm has a polynomial time complexity of $$O({|A|^2})$$O(|A|2) where |A| is the number of vertices.

This paper proposes a feature extraction technique called Gaussian probabilistic fuzzy generalized two-dimensional Fisher’s inverse linear discriminant analysis (GPFG-2DILDA) method based on fuzzy set theory, Gaussian probabilistic distribution information and inverse LDA. Like the FG-2DLDA, the proposed GPFG-2DILDA method also maximizes class separability along x- and y-axes directions simultaneously. The proposed method first calculates fuzzy membership matrix by fuzzy k-nearest neighbour (Fk-NN) algorithm. These values are combined with the training samples to obtain the global mean and class-wise mean training images. Thereafter, fuzzy membership values are integrated into intra-class and inter-class scatter matrices along both (x- and y-) directions. Similarly, Gaussian probabilistic distribution information is incorporated into the intra-class scatter matrices. Finally, by solving the eigenvalue problems of these scatter matrices, we find the optimal Gaussian-fuzzy inverse projection vectors, which actually used to generate more discriminant features and to solve the binary classification problem. The GPFG-2DILDA method has been evaluated on the AT&T face database to demonstrate the efficacy of the proposed method over some state-of-the-art face recognition methods.

In the recent trends, secure and efficient plan is required to get the access control of shared data in distributed systems. Different ABE schemes are provided for secure data sharing but, in CP-ABE scheme, a set of attributes are associated with each user the data on to user was encrypted based on access structure formed from the set of attributes. Any user who satisfies the access structure of the encrypted text they can only the decrypt the encrypted text. In data outsourcing environment, if you use this attribute-based encryption (ABE) results some challenges in case of attribute and user revocation. In smart grid provides sensitive data sharing policies and schemes because it has to deal with sensitive information and maintains flexibility for giving information about the policies used for the data protection or authentication details of data owner and receiver. A movement toward secure and efficient data sharing with fine grain access control. We propose a scheme that enables the “secure data sharing and key distribution” with communication channels, and the users can obtain their secret keys from the “key generation center (KGC)” and our scheme can accomplish “fine grain access control” and the user in the users’ list can use the data available in the cloud and restricted access is only allowed for the revoked users. Our proposed scheme enables the secure user revocation, key distribution, data confidentiality, and fine grain access control in cloud.

In quantitative biology, discovering a class when presented with a large bimolecular dataset poses a big problem. However, ensemble learning approach has been helpful in various complex areas of decision-making. So, in this paper, we propose a cluster-based ensemble learner called adaptive cluster-based ensemble learner (ACEL) which incorporates the prior knowledge of the datasets into the cluster ensemble framework. ACEL computes the cluster boundaries using three diverse clustering algorithms to obtain clusters for classification decision. ACEL learns by transforming the obtained clusters into rules and performing adaptive rule tuning to optimize the classification decision. The cluster-based classification results are then processed using majority voting algorithm. The proposed approach is compared with other supervised benchmark algorithms using seven problems from the field of biology. The experiments performed on benchmark datasets show that ACEL works effectively in classifying datasets.

Due to the advance Internet and increasing globalization, the electronics forms of information grow in a rapid manner. Extracting the useful hidden information from those multiple documents is a recent challenge. Hence, efficient and automated clustering algorithm which is effective in identifying topics plays the main role in information retrieval. In this paper, the analysis regarding the large unstructured text document corpus using our proposed map-reduce algorithm has been performed, and the results show the advantage of the proposed method by detecting clusters of document features within less computation time and provides premier solution for increasing the precision rate of retrieval in information extraction.

Today social networking has emerged as a prominent area of networking. It provides a platform where people can exchange information, sentiments, and expressions. Taking this as a core, the current research work was initiated wherein WhatsApp chats of different individuals were taken and effort was made to identify social inclination/pattern between them. The newly proposed WAPiS Algorithm: WhatsApp Pattern Identification Algorithm indicating Social Connection has been formulated to achieve this. The algorithm has been designed and developed using C#–MongoDB combination and the results indicate the existence or identification of the social contact between the WhatsApp text of the sampled individuals.

Today, with the increase in technology and computational power everything is getting digitalized. More and more private data is being loaded into our computers. These data are sent to various websites for suitable applications. Now, the problem arises while transferring these data. Sending these data through insecure channels like the Internet without using proper encryption methods can lead to data loss or data hacking. In order to protect this data, various encryption methods are being implemented. One such encryption methodology used in this arena is DNA cryptography. This paper gives a brief overview of DNA cryptology along with a new algorithm based on the fusion of symmetric-key cryptography, DNA nucleotides, and XOR operation is proposed. This algorithm is very much efficient and one main striking feature of the algorithm is the security which can be set as per sender requirements.

The programmed way of establishing and validating the existing person upon their corporal and observable characters are termed as biometric technology. Due to the accuracy of the iris recognition, it becomes more dominant in the available biometric techniques. The current study aims at a new technology for iris recognition which helps us to identify a human by the iris from various places. This method is more valid and protected when compared with the other biometric technologies. In this biometric, the human characters are used which will not change during the lifetime of that particular individual. The time taken for identification of individual human is very less. Iris recognition uses the uniqueness of the eye and the information is stored in the iris database. The movement of the human individual is recognized by the gait analysis and iris recognition is more liable in the existing biometric systems. The study mainly focuses on the iris preprocessing, edge detection, and feature extraction, finally gait and iris fusion classification techniques in the research area. Without the help of a particular individual, we can use the gait analysis for iris recognition. The security places such as banks, places which are used in elections, military installations and even airport, where more restriction to provide the details of the human use this biometric technology.

The swimming efficiency of a microswimmer composed of a rigid spherical head fixed to an elastic filament with tapered cross section is analyzed in Stokes flow regime. The small-amplitude harmonic base-angle actuation and linear taperings are assumed. The formulation of elastohydrodynamic equations and solution procedure are adapted from the theory of swimming developed by Lauga. The numerical results on the propulsion speed and the energy efficiency of the swimmer are presented as a function of the filament length and taper ratio. The results demonstrate that by introducing a linear filament tapering both the propulsion speed and the energy efficiency can be enhanced with an appropriate trade-off with the filament length.

A healthy mind requires healthy body and healthy body needs healthy food. Therefore, to provide healthy food to the rapidly growing population of a country is a challenge within the limited fertile land. To fulfill the healthy food demand requires the high production of yield. To achieve high production, it is essential for the farmers to monitor the field from time to time for a good yield, as a little miss can cause disaster and the entire efforts could go waste. The manual monitoring of the field is quite tough and expensive. Therefore, in this paper Internet of Things (IoT) applications are addressed to monitor the field. An Arduino microcontroller board with the soil moisture, temperature, and humidity sensors is used to collect the data from the remote field on the fly. The data once received is analyzed by applying cascade forward and function fitting neural network. Further, the data is tested against an already trained dataset of the field in normal conditions collected over a period of 1 year. The test data, when applied to the trained data, provides a dataset which is used for analysis of the ideal condition of the field. In case of alarming changes in the field properties, a signal is generated and the farmer is informed to take necessary action. A preventive action can be taken to save the crop and also maintain the productivity of the field.

In recent years, Vehicular Ad hoc Networks (VANETs) and Flying Ad hoc Networks (FANETs) are evolving rapidly. VANETs and FANETs are special types of Mobile Ad hoc Networks (MANETs). VANET uses vehicles as mobile nodes for the communication. VANET provide communication during emergency situations like accidents to avoid its possibility by sending alert messages to the drivers. FANET is a collection of unmanned aerial vehicles that communicate without any predefined infrastructure. FANET being the most searched and researched topic nowadays is finding its scope in flying objects like drones used for military applications such as border surveillance and for civil applications such as disaster management, traffic monitoring. In VANETs and FANETs, routing is challenging when Quality of Service (QoS) parameters needs to be satisfied. In this paper, VANETs and FANETs’ routing protocols implementing optimization techniques (like Ant colony optimization, bee colony optimization, and particle swarm optimization) are surveyed. The differences between VANETs and FANETs are clarified first and then routing protocols for VANETs and FANETs are discussed.

This work presents the cloud-assisted secure WBAN for healthcare application. There are various security issues associated with WBAN, which need to be solved to provide a secure real-time health monitoring system. Through this implementation, the patient’s vital signals can be accessed in a secure manner in real time remotely by sensors and networks without visiting doctor’s clinic or hospital. Here, we provide the cost-effective solution for the transmission of the patient’s health data to doctor with proper confidentiality, authenticity, freshness, and security using cloud computing. In this work, the biosignals of patients and doctors are used to provide authenticity and vital signals are encrypted by using Advanced Encryption Standard (AES) for the secure m-health application. We have experimentally analyzed the average end-to-end delay for secure healthcare application is 14.59 and 19.31 ms in off-peak hours and peak hours respectively. This delay is only 5.84% in off-peak hours and 7.72% in peak hours of permissible delay of 250 ms for medical application.

Android is used as an operating system by a vast range of devices such as mobile phones, tablets, TV sets, watches, etc., providing users with a collection of android applications to, carry through, their daily needs. Android security has been built upon a “Permission-based mechanism”, which gives access to the critical information, of an Android device to an application, and often does not allow the user to understand the privacy implications of those accepted permissions. User’s overriding desire to complete the primary task to install an Android application reduces their attention to the installation-time permission requests. This study on Android applications will help the users to gain better understanding of the required and used permissions in Android environment, giving them a preferable user control to effectively maintain confidentiality, integrity, availability, security, and privacy of user’s data.

A Maglev car travels by floating on a magnetic track with the attractive and repulsive force of magnets to create both lift and propulsion. The paper presents outlines of an application on magnetic levitation transportation technology to design a new car and study of it utilizing computer vision. This vehicle is specially designed to give high comfort to the physically challenged people and to move in a profoundly secure condition. The vehicle is planned with a full resting beds compartment and can move at a very high speed. Inductrack a frictionless, attractive levitation framework is proposed to be used here for super speeds. It offers a smooth contactless trip to the goal. Helper components are just required when the vehicle needs to back off and needs to stop at a station. The financially savvy Halbach arrays of magnets are utilized to produce high magnetic fields to suspend and move the vehicle. Computer animation of the whole circumstance shows up and observed it to be a good vehicle with high speed. Computer vision system study is utilized to detect cars, monitoring vehicle, analysis of car video information for a safe journey, and to avoid accidents. Tracking a car in a vulnerable environment is difficult due to noise generation. To overcome this problem, a method is proposed for noise removal from the car videos for tracking it in a vulnerable environment. It is found that the proposed methods give better result compared to the existing one. Thus, it shows that computer vision can greatly help the Maglev vehicle to provide more safety, high alerts, and high comfort to the passengers.

Lung cancer is a malignant tumour having uncontrolled lung cell growth. Papanicolaou (Pap)-stained cell cytology from Fine Needle Aspiration Cytology (FNAC) is the most followed approach for lung cancer diagnosis. However, the manual assessment of cytopathology slides under light microscopy is time consuming and suffers from feature ambiguities including inter-observer variability. Here, an automated computer vision approach is presented for identifying and classifying cancerous cell nuclei from pap-stained microscopic image of lung FNAC sample. The proposed methodology adopted Gaussian mixture model-based hidden Markov random field technique to segment cell nucleus. Later, bag-of-visual words model was used for nucleus classification, where scale-invariant feature transform feature were extracted from segmented nucleus for training a random forest classifier model. The adopted nucleus segmentation-cum-classification model was able to precisely segment the nucleus and classify them in two class, viz. Small Cell Lung Cancer (SCLC) and Non-small Cell Lung Cancer (NSCLC). The segmentation process achieves a sensitivity of 98.88% and specificity of 97.93%. And also, the nucleus classification model was able to perform with a sensitivity of 97.31%, specificity of 99.54%, and accuracy of 98.78%.

A skin ulcer is a clinical pathology of localized damage to skin and tissue instigated by venous insufficiency. Precise identification of wound surface area is one of the challenging tasks in the dermatological evaluation. The assessment is carried out by clinicians using traditional approach of scales or metrics through visual inspection. The manual assessment leads to intra-observer variability, subjective error and time complexity. This paper evaluates the performances of supervised and unsupervised segmentation techniques used for wound area detection. The unsupervised methods used for evaluation were namely K-means, Fuzzy C-means and Gaussian mixture model. On the other part, random forest was implemented for supervised classification. Several filtering methods were used to generate image feature set from wound images to train random forest. The Gaussian mixture model with classification expectation–maximization clustering method achieved the highest weighted sensitivity of 95.91% and weighted specificity of 96.7%. The comparative study shows the superiority of proposed method and its suitability in wound segmentation from normal skin.

Nowadays, intelligent systems are quite common to assist the driver in its navigation chores. But, if accurate information is not provided to the driver on time then it is of no use and the delay can even lead to disastrous consequences. To improve behavioral realism in real life navigation scenarios we need intelligent and speedy decisions. Such decisions rely on the information being extracted from real-time images. We need to select the best available features/information that can be used as neural network inputs and ultimately predict the next move of the vehicle. Little work has been done on selection of best features of navigation images and research needs to be done in this gray area as its results have direct impact on the classification accuracy and generalized performance of automotive navigation planning. This article proposes a novel approach to find the best possible set of features from real-time navigation images by using Boruta Algorithm and Earth Algorithm so as to improve the prediction power of vehicle to either move or stop in the next course of action. The results obtained were cross-validated by using three classifiers: Logistic Regression, Support Vector Machines (SVM), and K-Nearest Neighbor on the basis of parameters: Classification accuracy and prediction performance. It was identified that the proposed model maintained the classification accuracy and performed with more superiority by getting rid of the irrelevant features of images and thereby reducing the training time as compared to computations done on the basis of original feature set. The prediction speed of the proposed model was found to be much better than the model without feature selection. The accuracy of this novel approach was also found to be improved by few ensembles using Generalized Linear Model (GLM) wrapper.

With the extensive use of machine vision methodologies, computer-assisted disease diagnosis has become a popular practice for the medical professionals. Detailed analysis of wound bed area and precise identification of the wound tissue regions are the most desirable aspects of an automated wound assessment applications. This study proposes a supervised wound tissue classification method, where a deep neural network classifier model is trained by the colour, texture and statistical features which are extracted from different tissue regions. The proposed classification process considers three types of tissue, viz. granulation (red), necrotic (black) and slough (yellow) and a total of 105 features are used for the classification. A pixel-based feature extraction approach is implemented to extract features from the tissue region, where a mask window of size $$9\,\times \,9$$9×9 runs over each pixel of the tissue regions for feature extraction. The proposed deep neural network achieves accuracy 99.997215%, sensitivity 99.998006%, specificity 99.996625% and F-Measure 99.997316%.

In CBIR techniques, image retrieval based on object-based features are more precise to retrieve appropriate relevant images. So, in this paper, a CBIR technique is proposed using extracted combined shape and color features from image object region. In this particular work, some significant statistical parameters are calculated from image object or shape region by gray-level co-occurrence matrix and simultaneously, color features are extracted from the color object using color autocorrelogram. Initially, RGB color images are transformed into YCbCr color space, and subsequently, the active contour is employed on Y-component to obtain the foreground and the background regions. Shape or object feature is located in the foreground region of Y-component and gray-level co-occurrence matrix provides some statistical parameters. We have also computed some statistical parameters from the background region to improve the image retrieval performance. Afterward, an intermediate color object image is reconstructed by combining foreground image region along with chrominance components for deriving the prominent color information. We have employed color autocorrelogram over this newly constructed intermediate image. Finally, all the computed features are combined together to form the ultimate feature vector. The proposed technique is tested over two benchmark databases, i.e., Corel-1K and GHIM-10K and we have achieved satisfactory results in object-based images.

Recent growth in depth camera technology has significantly enhanced human motion tracking. Human future behaviour and intention forecasting become a challenging task due to high-dimensional interactions with the physical world. Prognostic methods that estimate ambiguity are, therefore, critical for supporting appropriate robotic responses to the numerous ambiguities posed within the human–robot interaction environment. Beyond autonomous agents, we will also see our surroundings—buildings, cities—becoming equipped with ambient intelligence which can sense and respond to human behaviour. In this paper, we present different deep learning models that can forecast the navigational behaviour of multiple classes (i.e. pedestrian, car, cycle) by considering influencing factors such as the neighbouring dynamic subjects and social behaviour of the classes under investigation. The results show that our approaches outperform the existing state-of-the-art forecasting models.

Charles Darwin once wrote: ‘It is certainly not true that there are in the mind of man any universal standards of beauty with respect to the human body’. The relation between facial beauty and the golden ratio is a known fact. In this paper, we have tried to establish the relation between face beauty and the golden ratio. Finally, we try to improve facial beauty using the golden ratio-based geometric transformation and some filtering operations. The work is divided into two parts: 1. verification of the relation between face beauty and golden ratio, 2. application of golden ratio for face beautification. The first part of the paper is based on the verification of a neoclassical theorem of beauty and the golden ratio based on the symmetry of the face, using various machine learning tools. Verification of the ratings is done using SCUT-FBP dataset. We used 450 images for the training purpose out of the 500 images and the rest 50 images are used for testing the data. The second part of the work is to beautify a face based on mathematical calculations and improve the skin texture, removes blemishes, and change the facial features according to the golden ratio. Test results show the significant improvement in facial beauty due to the application of the golden ratio.

MANETs (Mobile Ad hoc Networks) are wireless networks that are deployed for a particular purpose or short-term use. Because of the lack of central coordination, MANETs share an inherent trust relationship among the nodes forming the network. Each node implicitly trusts its neighbour to forward packets in the network till the packets reach their destination. Further, each node in a MANET can monitor its neighbours by keeping track of the packets passing through the neighbours. This characteristic ability of a node in MANET makes it possible to develop a trust model that can correlate with the innate trust shared among the nodes. This paper looks at developing such a trust model which is applied to all the nodes in the network. The trust model works like an Intrusion Detection System (IDS), which seeks to detect blackhole attacks in the system, and then identify and mitigate the malicious attacker.

The paper presents a metric to automatically compute a score for machine-generated questions and transforms the questions which are having a lower score value, unacceptable, and ungrammatical into a human appealing form. Questions are unacceptable due to the flaws like incorrect grammar, selection of wrong wh-phrase, partial selection of answer phrase, negation, etc. Identifying such infirmities in the question is a challenge. Here, our attempt is to automatically detect and correct the flaws present in the question. We named this system as the Automatic Question Quality Enhancer (AQQE). By employing a multiple linear regression model, AQQE first computes the score (in range of 1–rejected to 5–accepted) for 174 questions. Higher score value tells the acceptance and lower score value shows the rejection of the question. AQQE’s challenge is to enhance the quality of questions having a lower score. Out of 174, human evaluator had identified 84 questions as acceptable and 90 (51.72%) as an unacceptable. Performance of AQQE is judged with precision and recall and it is found well acceptable. AQQE enhanced 79(87.77%) questions are accepted by the human evaluator and 11 (6%) questions can be accepted with further modifications.

This paper suggests an innovative technique for quality-type linear barcodes from the face image. This procedure calculates the distinction in gradients of image shine and then it requires finding the average of the gradients into a finite number of intervals using normalization. After this, the result of quantization is converted into the limits of decimal digits from 0 to 9 and table is translated into an ultimate linear barcode. A theoretical analysis shows that the upper part of the physiognomy is not affected by a remark (like a change in physiognomy utterance, change of face range size, change in eye range (gaping and occlude eyes, after mirror rotation of input image, changes with human age)) ensuring the stability of representing its features. So, this method generates the standard-type linear EAN-8 barcode of the top part of face image (upper 70% and 75% of face image). However, in this work, a technique which is suggested is established based on the idea that the distances between attributes of a human face by comparing the feature edges of human faces that have been done by determining its image gradient using window technique. The determined gradient data is further reduced into a smaller data set for the representation of a histogram of each face. From this histogram, the barcode is generated for each face image. This method is tested by using “Face94”, “YaleB Face Database”, “Face database FERET”, FG-NET dataset. A generated barcode holds knowledge about human’s face and can be used for recording, finding, recognition, and search for peoples.

Privacy preservation of cloud data is gaining importance since majority of digital records are transferred to cloud-based storage. Familiarization of cloud storage is rapidly growing and people are choosing cloud storage services as suitable alternatives for personal setup. Potential increase in the cloud usage creates a concern for data security. Cloud records demand robust encryption standards. Protection is needed both from external hackers and internal intruders. This article reports a bit-reversal encryption mechanism for privacy preservation of digital image data deployed on cloud. This encryption mechanism deals with different types of digital images toward secure storage in cloud infrastructure. Experimental results show that the method is quite simple and easy to implement to preserve the privacy and security of user data either off premises or on premises cloud storage.