Multimedia Technology and Enhanced Learning

In order to solve this problem, a heuristic network similarity measurement model based on cloud computing is proposed. First, the heuristic network data is collected, and then the spherical harmonic function method is used to match the network data similarity measurement. After the above work, the heuristic network similarity measurement model is built according to the structure balance theory. Thus, a heuristic network similarity measurement model based on cloud computing is constructed. In the experiment, the quality of service node genes obtained by the two models was tested. The experimental results show that the service node genes obtained by the model are better and meet the design requirements.

In order to improve the preparation ability of lithium ion screen precursor, an anti-interference high definition lithium ion screen precursor preparation data dynamic migration and information enhancement method based on dynamic migration equilibrium modulation is proposed. The dynamic migration transmission channel model of HD lithium ion screen precursor preparation is constructed, and the anti-interference design of HD lithium ion screen precursor preparation is carried out by using interference filtering algorithm. The dynamic migration state characteristic of the preparation data of HD lithium ion screen precursor is extracted, and the Porter interval equilibrium method is used to control the dynamic migration of HD lithium ion screen precursor preparation data. The dynamic migration optimization of preparation data of anti-interference HD lithium ion screen precursor is realized, and the anti-interference and safety of lithium ion screen precursor preparation are improved. The simulation results show that the preparation of HD lithium ion screen precursor by this method is safe and has strong anti-interference ability.

The characteristic distribution performance of big data, the exchange characteristic of lithium ion screen in cloud computing environment, quantitatively reflects the running state of lithium ion screen exchanger, in order to realize the effective monitoring of lithium ion screen exchange process. A fast extraction algorithm of Li-ion screen exchange feature big data based on big data is proposed. Big data acquisition of lithium ion screen exchange characteristics is realized in lithium ion screen exchange array, and the statistical analysis model of big data mining is constructed. In big data distribution subspace, the spectral feature extraction method is used to extract the spectral stripe feature of Li-ion screen exchange feature big data, and the extracted spectral stripe feature is fuzzy clustering and mining by adaptive neural network learning algorithm. Big data rapid extraction of exchange characteristics of lithium ion screen was realized. The simulation results show that the method has high accuracy in fast extraction of exchange features of lithium ion screen, strong resolution of exchange characteristics of lithium ion screen, and has good application value in high precision measurement of exchange characteristics of lithium ion screen.

In view of the low recall of the traditional network text data classification algorithm, an artificial intelligence based network text data classification algorithm is designed. Before feature extraction, text information is preprocessed first, and word stem is extracted from English. Because there is no inherent space between Chinese words, word segmentation is carried out to complete the preprocessing of network text data. On this basis, an evaluation function is constructed to evaluate each feature item in the input space independently, and to reduce the dimension of the features of the network text data. Finally, the artificial intelligence method is used to classify the network text data, and the most similar training text is found through similarity measurement in the network text data training set. The experimental results show that the designed algorithm based on artificial intelligence has higher recall than the traditional algorithm, and can meet the needs of network text data classification.

In order to improve the security of online social network security and privacy data and shorten the delay of privacy data protection, this paper proposes a method of online social network security and privacy data protection in the era of big data. In order to enhance the security of online social network security and privacy data, a network security and privacy data architecture is constructed. In view of the risk of online social network security and privacy data loss, this paper proposes a set of security and privacy data backup processing scheme. Complete the formulation of online social network security and privacy data protection scheme; Combined with the current attack methods commonly used by attackers, the shortcomings of the traditional privacy protection algorithm in protecting the security of online social network privacy data security are concluded. Design online social network security privacy data privacy protection algorithm; Finally, complete homomorphic encryption of online social network security and privacy data is adopted to realize online social network security and privacy data protection in the era of big data. Experimental results show that the proposed privacy data protection method has a shorter delay than the traditional one.

Aiming at the problem that the transmission error of the traditional algorithm is too large, a dynamic data migration security enhancement algorithm for optical fiber network communication is proposed. Firstly, the migration data in the optical fiber network is collected. According to the data transmission process, the electric domain compensation, optical domain compensation and optical domain electric domain hybrid compensation in the PMD compensation unit design scheme are used to calculate the path of data safe migration. The dynamic migration data is collected by the variable electric delay algorithm, and then the amplitude of migration data is preprocessed according to the transmission characteristics of optical pulse in the optical fiber, to the preprocessing formula, use the PSO search mechanism to calculate the migration security feature data, and calculate the dynamic migration amount of n input layers, and finally get the dynamic migration security enhancement algorithm. The experimental results show that compared with the traditional methods, the error of dynamic data migration security enhancement algorithm is smaller when transmitting enhanced data.

In order to solve the problems of the traditional anti-jamming algorithm of broadband wireless communication, such as poor anti-jamming performance and high bit error rate, an anti-jamming algorithm of broadband wireless communication based on Embedded MCU is proposed. In the broadband wireless communication based on embedded single-chip microcomputer, the m-sequence of communication signal data is constructed, encoded and decoded. Finally, CRC redundancy test and error correction are carried out for the decoded communication data, so far the design of broadband wireless communication anti-interference algorithm based on embedded single-chip microcomputer is completed. Through the contrast experiment, compared with the traditional anti-jamming algorithm of broadband wireless communication, the experimental results show that compared with the traditional anti-jamming algorithm of broadband wireless communication, the proposed anti-jamming algorithm of broadband wireless communication based on Embedded MCU has lower bit error rate, which shows that it has better anti-jamming ability.

In the traditional basketball teaching mode, teachers could’t solve the problems in the training of students in time, which affected the training effect of students. Therefore, the basketball teaching mode based on ITbegin cloud platform is designed. According to the difficulty of basketball teaching and training content, it is divided into basic content and advanced content. On this basis, made basketball teaching and training ITbegin course. Published it in the ITbegin cloud platform through the Internet. Students watched the course by downloading or online, complete the training and timely feedback the training progress and problems encountered in the training, adjusted the teaching mode in time, ensured training quality. The test results showed that: In the case of similar physical quality of the test personnel, the overall level of the basketball teaching mode based on the ITbegin cloud platform was higher than that of the traditional teaching mode. It is proved that the design of basketball teaching mode based on ITbegin cloud platform was more suitable for practical training projects.

In view of the poor teaching effect of the traditional sports teaching method, this paper designs the sports auxiliary teaching method based on the multimedia network. By introducing extended knowledge, using visual method and language method, improving teaching means, making multimedia courseware with flash, combining the teaching content and teaching needs of sports, changing teaching mode, applying advanced teaching facilities, and completing the design of auxiliary teaching method of sports based on multimedia network. The results show that, compared with the traditional teaching methods, the application of multimedia network in physical education teaching, its teaching effect has been significantly improved.

In this paper, a tightly-coupled light detection and ranging (LiDAR)/dead reckoning (DR) navigation system with uncertain sampling time is designed for mobile robot localization. The Kalman filter (KF) is used as the main data fusion filter, where the state vector is composed of the position error, velocity error, yaw, and sampling time. The observation is provided of the difference between the LiDAR-derived and DR-derived distances measured from the corner feature points (CFPs) to the mobile robot. A real test experiment has been conducted to verify a good performance of the proposed method and show that it allows for a higher accuracy compared to the traditional LiDAR/DR integration.

In order to improve the ability of human-computer interactive fire extinguishing training, a visual simulation model of man-machine interactive fire extinguishing training based on virtual reality technology is proposed. Taking the large-scale emergency scene fire drill as the research object, the virtual scene image reconstruction model of human-computer interaction fire fighting training is constructed, the boundary feature detection and particle tracking filter processing are carried out for the scene image of human-computer interaction fire fighting training, the dynamic structure of the scene image of human-computer interaction fire fighting training is reorganized with fuzzy edge feature extraction method, and the virtual reality simulation model of human-computer interactive fire fighting training scene image is established. The multi-person cooperative control method is used to simulate the virtual reality in the process of human-computer interactive fire fighting training, and the 3D simulation image of virtual scene fire drill in large-scale emergency scene is followed and rendered, and the optimization design of virtual reality VR simulation model of human-computer interactive fire fighting training is realized. The test results show that the virtual scene simulation of human-computer interaction fire fighting training using this method has good cooperation, high image fusion performance and strong reconstruction ability of fire fighting training scene simulation.

In order to realize the visual effect optimization design of landscape virtual interactive planning in the resettlement area, the optimal design method of landscape virtual interactive planning in the resettlement area based on situational awareness is proposed. The feature sampling model of landscape virtual interactive planning optimization is established, the virtual reality simulation in landscape virtual interactive planning design is carried out by MPI visual simulation tool, the virtual interactive planning feature construction of landscape virtual interactive planning is carried out in Vega Prime software, and the virtual interactive planning information sampling model and block information fusion model of landscape virtual interactive planning are established. Create, edit and run virtual interactive planning optimization program of landscape in resettlement area, combine with cross-compiling method to simulate virtual interactive planning information of landscape in resettlement area, create 3D visual environment of virtual interactive planning of landscape in resettlement area in real-time interaction, and realize virtual interactive planning optimization design of landscape in resettlement area in virtual reality simulation environment. The simulation results show that this method can effectively realize the visual optimization design of virtual interactive planning of landscape in resettlement area, improve the visual feature expression effect of virtual interactive planning of landscape in resettlement area, and has good application value in virtual interactive planning design of landscape in resettlement area.

In order to improve the reconstruction ability of time series data under probability statistical model, a time series data reconstruction method based on machine learning is proposed. The time series data distribution structure model under probability statistical model is constructed. The spatial multi-sensor information sampling method is used to sample the time series data information flow under the probability statistical model, and the phase space reconstruction method is combined to reconstruct the time series data information structure under the probability statistical model. The probability statistical model is established to decompose the time series data, and the distributed grid computing method is used to extract the big data association features of the time series data under the probability statistical model. Combined with the adaptive weight learning method, the optimal control of the scheduling is carried out. The big data cross-domain scheduling of the time series data under the probabilistic statistical model is realized under the support vector machine learning mode. The simulation results show that the method has good adaptability to time series data cross-domain scheduling under the probability and statistics model, and the load balance of data output is strong.

The traditional logistics vehicle scheduling method only estimates the total scheduling of batch vehicles, without considering the capacity limit of single logistics vehicle. It causes the problem of waste in vehicle transportation. Therefore, a vehicle scheduling method based on monitoring image under time constraints is proposed, using the time displayed in the monitoring image to constrain, the dynamic scheduling model is established by setting up the time window scheduling model to schedule the vehicle tasks within the time window conditions. According to the images obtained from the monitoring, combined with the need to divide several stages under the time constraints, the vehicles to ensure the logistics transportation can be scheduled according to the actual situation, make a highly optimal decision, achieve the maximum vehicle load rate, and ensure the smooth implementation of the dynamic strategy of the non full load logistics vehicle scheduling under the time constraints. Finally, the simulation test results show that the proposed method can improve the efficiency and rationality of logistics vehicle scheduling, the algorithm is stable and reliable, and has strong practicability.

In this paper, a short-term load forecasting model based on PCA dimensionality reduction technology and grey theory is proposed. After the correlation analysis between meteorological factors and load indicators, the data is carried out by combining PCA dimensionality reduction technology and grey theoretical load forecasting model. In this paper, the validity of the load data verification model in a western region is selected. The analysis of the example shows that compared with the general gray prediction model GM (1, 1), the accuracy of the model prediction result is much higher, which proves the model. Effectiveness and practicality.

The traditional double-camera face recognition attendance image preprocessing algorithm can not distinguish the target from the complex background. In order to solve this problem, an artificial intelligence based double-camera face recognition attendance image preprocessing algorithm is proposed. First, artificial intelligence technology is used to extract the features of face recognition attendance image, and then spatial denoising algorithm is used to remove the noise of face recognition attendance image. On this basis, multi-channel texture weighting algorithm is used to realize the double-camera face recognition attendance image preprocessing. Therefore, a double-camera face recognition image preprocessing algorithm based on artificial intelligence is completed. In the experiment, the infrared image of the face is tested to see whether the evaluation factors obtained by the two algorithms can distinguish the target from the complex background. Experimental results show that the algorithm has a short computing time and can distinguish targets in complex background in a short time.

The conventional methods of military intelligence assessment could not comprehensively analyze the value of military intelligence. To this end, a military intelligence value assessment method based on big data analysis was proposed. Big data analysis technology was introduced to determine data value density k; overall system architecture was established; operation mode was optimized, relevant analysis technology was formulated; military intelligence value evaluation was achieved. Experimental data showed that the application of big data analysis technology could comprehensively analyze the value of military intelligence.

Aimed at the problem of volatile energy consumption of the traditional grid energy-saving control model, design a less volatile energy consumption of the power grid energy-saving control model, by building a basic calculation model, to meet energy-saving targets and related constraints, using the priorities and time calendar solving basic measurement model, in order to realize the energy-saving power generation dispatching; according to the results of energy-saving power generation scheduling, a power load stratification probability prediction method based on empirical mode decomposition and sparse Bayesian learning is used to establish a load forecasting model for load sampling. Based on the sampling results, the emission control cost and the reserve capacity cost, the nominal purchase cost of the non-renewable energy unit and the nominal purchase cost of the renewable energy unit are respectively constructed, and then the energy-saving control model of the power grid is constructed. In order to prove that the energy consumption fluctuation model of the power-saving control model is small, the model is Compared with the traditional grid energy-saving control model, the experimental results show that the energy consumption volatility of the model is less than that of the traditional power grid energy-saving control model, which reduces the nominal power purchase cost of renewable energy units and is more suitable for power grid energy-saving control.

Sensor network coverage is one of the basic problems in the Internet of Things. Coverage is one of the important indicators to measure the performance of sensor network nodes. Through the research on coverage problems, we can seek ways to improve the quality of sensor network services. A wireless sensor network node control effect is proposed. A distributed algorithm based on probability model is first constructed to optimize the probability perception algorithm. The above-mentioned two-dimensional space algorithm is extended to three-dimensional space, and the greedy heuristic algorithm is used to obtain the control solution to achieve the optimal control of the current wireless sensor network. In addition, the matlab simulation program is written, and the algorithm is compared with the simulation results of the average algorithm and the random algorithm. The simulation results of the proposed algorithm have significant advantages.

Substation equipment status maintenance background. Solve the problem of high cost in equipment maintenance. A decision model for substation equipment maintenance based on association set decomposition is proposed. Under the premise that the state of the device state is known. Analysis of the basic structure of substation equipment. And predict the operating status of the device. Starting from the functional association between devices, the association set is the basic unit. Realize the time-varying maintenance decision of the equipment. The conclusion is obtained by the model verification experiment: Compared with traditional equipment overhaul models. The method of substation equipment maintenance decision model based on association set decomposition can save 16.5% maintenance cost.

Due to the frequent occurrence of network security incidents, causing unnecessary losses to people, frequent network security incidents are worrying. For the problems of Internet application network information security, attackers use attacks to continuously threaten them. This paper studies the method of information security monitoring and early warning mechanism for Internet application network based on particle swarm optimization. Based on the support vector regression machine, a network security prediction model with multi-group chaotic particle optimization is established. The prediction results are obtained through the network information security monitoring and early warning mechanism, and the prediction results are analyzed and summarized. The results show that the Internet application network information security prediction model based on particle swarm optimization algorithm can provide guidance for the development of network security solutions and strategies, enhance the initiative of network security defense, reduce the losses caused by network attacks, and have better practicality Sex.

Multiple sclerosis can damage the central nervous system, and current drugs are difficult to completely cure symptoms. The aim of this paper was to use deep learning methods to increase the detection rate of multiple sclerosis, thereby increasing the patient’s chance of treatment. We presented a new method based on hu moment invariant and artificial neural network trained by particle swarm optimization. Our method was carried out over ten runs of ten-fold cross validation. The experimental results show that the optimization ability of particle swarm optimization algorithm is superior to the genetic algorithm, simulated annealing algorithm and immune genetic algorithm. At the same time, compared with the HWT+PCA+LR method and the WE-FNN-AGA method, our method performs better in the performance of the detection.

It is significant to classify teeth categories in dental treatment. A novel teeth classification method was proposed in this paper, which combined fractional Fourier entropy and feedforward neural network. Firstly, fractional Fourier transform was performed on the teeth CT images and the obtained spectrums were used to extract entropies as the features. Then, a feedforward neural network was employed for automatic classification. To train the parameters in the network, improved hybrid genetic algorithm was leveraged. Experiment results suggested that our method achieved state-of-the-art performance.

With the risk of hearing loss being higher than before since the digital device is more popular, it becomes more urgent to identify the sensorineural hearing loss from the view of changes in internal brain structure. Based on 180 brain MRI of three categories of hearing loss balanced dataset, one schema with fractional Fourier transform entropy and direct acyclic graph support vector machine is proposed and applied to identify the features and predict the categories of hearing loss. The experiments prove this schema rather promising when the dataset is not large since the overall accuracy is up to 94.06 ± 1.08% which is higher than those of some previous methods in scope of traditional machine learning.

Gingivitis is usually detected by a series of oral examinations. In this process, the dental record plays a very important role. However, it often takes a lot of physical and mental effort to accurately detect gingivitis in a large number of dental records. Therefore, it is of great significance to study the classification technology of gingivitis. In this study, a new gingivitis classification method based on wavelet entropy and support vector machine is proposed to help diagnose gingivitis. The feature of the image is extracted by wavelet entropy, and then the image is classified by support vector machine. The experimental results show that the average sensitivity, specificity, precision and accuracy of this method are 75.17%, 75.29%, 75.35% and 75.24% respectively, which are superior to the other three methods This method is proved to be effective in the classification of gingivitis.

In order to solve the problems that indoor mobile robots have parking during the traveling process and the Extended Kalman filter (EKF) receives too much influence on parameter selection, this paper proposes an Interacting Multiple Model (IMM)-EKF/Particle Filtering (PF) adaptive algorithm for the tightly inertial navigation system (INS)/Light Detection And Ranging (LiDAR) integrated navigation. The EKF and PF calculate the position of the robot respectively, then the smaller Mahalanobis distance-based filter’s output is selected as the initial value of the next iteration, which improves the accuracy of the positioning for the robot. Based on that, the two motion equations of the static and normal motion models are dsigned at the same time. A Markov chain for converting the two state of the model, and the weighting filtering result of the filtered is used to provide distance estimates. The real experimental results show that the IMM-EKF/PF adaptive algorithm improves the positioning accuracy of mobile robots in the presence of parking.

Nowadays, significant changes have taken place in the field of information technology and industry and robot research is also deepening. The realization of multi-robot flocking control problem has far-reaching significance. This paper mainly introduces the development status of flocking control at home and abroad and summarizes several commonly used distributed flocking control strategies. In this paper, on the basis of summarizing the development of flocking research at home and abroad, forecasts its development prospect in the field of aviation and so on.

With the development of the national economy, the demand for electricity in various industries is expanding. It is necessary to ensure the safe operation of the high voltage transmission line. How to prevent and detect natural disasters and accidents that endanger transmission lines in a timely manner has become an important basic work to ensure power supply. Identifying high-voltage transmission lines first requires mathematical modeling of high-voltage transmission lines. Based on the mathematical model constructed, the image processing method is used to remove the blurred images in the images and restore the true background of the images. The establishment of mathematical models for high-voltage transmission lines has been relatively complete. This paper focuses on the analysis of existing methods for automatic identification and localization of foreign bodies on transmission lines and the existing research on deblurring, de-fogging, image denoising, image enhancement, etc. method. With the rapid development of deep learning, there are more and more methods for identifying high-voltage transmission lines and image restoration. More people will be engaged in this research in the future.

With the development of science and technology, using electronic fence to replace traditional physical fence becomes a trend. Research in the image processing-based electronic fence is gaining more popularity due to its low cost, low power consumption, and intelligence. In this paper, we are interested in the study of three types of scenarios with an image processing-based electronic fence. The scenarios contains national border protection, safety management in manufacturing field and vehicles safety plans. The state of the art frameworks of the three scenarios are reviewed, and the trend of these fields is also discussed.

In view of the defects and shortcomings of traditional Automated Guided Vehicle (AGV) robots in the localization mode and working scene, this paper studies the tightly-coupled integrated localization strategy based on inertial navigation system (INS) with ultra wide band (UWB). This paper presents an interactive multi-model (IMM) to solve the influence of non-line-of-sight (NLOS) on positioning accuracy. In IMM framework, two parallel Kalman filter (KF) models are used to filter the measured distance simultaneously, and then IMM distance is obtained by weighted fusion of two KF filtering results. This paper adopts the tightly-coupled combined method, and performs indoor positioning by extending Kalman filter (EKF). Experiments show that the method can effectively suppress the influence of NLOS error and improve the localization accuracy.

Considering Fuji apple, the relationship between the near infrared spectrum and the soluble solids content (SSC), which is one of the important indexes to measure the internal quality of apple, is studied in this paper. In order to reduce the computational complexity and to improve the accuracy of modeling, this paper adopts the wavelet packet threshold denoising method for spectral spectrum processing, and uses the method of wavelet packet analysis (WPA) to filter the characteristic wavelength of the spectrum. Moreover, a prediction model of SSC is proposed based on BP neural network due to its characteristics of anti-noise, anti-interference, strong nonlinear conversion ability and the good capacity in handling nonlinear measured data with uncertain causality. Finally, the simulation results show that wavelet packet analysis can not only reduce the calculation of modeling variables, but also Improve modeling accuracy of the BP neural network model. The proposed method can make a better prediction of the SSC of apple.

The accuracy of pedestrian positioning is helpful to ensure the pedestrian safety in both indoor and outdoor environments. Improve the accuracy of pedestrian positioning is a key research issue. In order to solve the problem that the indoor and outdoor pedestrian navigation is not continuous and the accuracy is low, a pedestrian seamless navigation and positioning method based on BDS/GPS/IMU is proposed. In outdoor environment, in order to improve the availability of dynamic positioning when the single-system observation geometry is not ideal, the key techniques such as the differential coordinates and time benchmark in BDS/GPS positioning are studied, and a method of eliminating time difference by the independent combination difference in the system is proposed. This method simplifies the operation steps and overcomes the current compatible positioning difficulties without the time difference between BDS and GPS. It can be seen that the more the number of visible satellites, the better the space geometric distribution. In the combined positioning experiment results of BDS and GPS, the number of visual satellites are about 6–8 when GPS is used alone. When using the combined system, the number of visible satellites increased to about 16. The increase in the number of visible satellites greatly improves the observation geometry. For the Position Dilution of Precision (PDOP), the maximum PDOP of the dual system is 2.7, which is significantly lower than that of the single system, and the observation geometry performance is greatly improved. In the effective positioning time, for BDS/GPS, the positioning accuracy of elevation (U) direction is better than 4 cm, and the positioning accuracy of North (N) and East (E) direction is better than 2 cm. Based on the analysis of pedestrian gait characteristics, a multi-condition constrained zero-velocity detection algorithm is proposed. For the error of the inertial sensor error is accumulated over the time, the zero velocity update (ZUPT) algorithm is implemented to correct the cumulative errors by using to the designed extended Kalman filter (EKF) with the velocity and angular velocity information as the measurements. The results show that the accuracy of dual-mode positioning system of BDS compatible with GPS is better than the single-mode GPS positioning, the outdoor position accuracy can reach centimeter level, and under ZUPT compensation the indoor error ratio is 1%, which can achieve more accurate pedestrian seamless navigation.

This paper is based on the Chinese sign language video library, and discusses the algorithm design of video classification based on handshape recognition of key frames in video. Video classification in sign language video library is an important part of sign language arrangement and is also the premise of video feature retrieval. At present, sign language video’s handshape classification work is done manually. The accuracy and correctness of the results are quite erroneous and erroneous. In this paper, from the angle of computer image analysis, the definition and extraction of key frames are carried out, and then the region of interest is identified. Finally, an improved SURF algorithm is used to match the area of interest and the existing hand image, and the classification of the video is completed. The entire process is based on the actual development environment, and it can be used for reference based on the classification of video image features.

Sign language plays a significant role in smooth communication between the hearing-impaired and the healthy. Chinese fingerspelling is an important composition of Chinese sign language, which is suitable for denoting terminology and using as basis of gesture sign language learning. We proposed a Chinese fingerspelling recognition approach via Hu moment invariant and RBF support vector machine. Hu moment invariant was employed to extract image feature and RBF-SVM was employed to classify. Meanwhile, 10-fold across validation was introduced to avoid overfitting. Our method HMI-RBF-SVM achieved overall accuracy of 86.47 ± 1.15% and was superior to three state-of-the-art approaches.

One challenging task for the blind is to input Braille while by no way could they sense the location information on touch screens. The existing Braille input methods are suffering from problems including inaccurate positioning and lack of interactive prompts. In this paper, touch gestures are recognized by trained RBF network while combined gestures are modelled. By doing so, the Braille input concerning multi-touch gesture recognition is then implemented. The experimental results show that the method is effective and blind people can friendly input Braille with almost real-time interaction.

Patients with severe mental disorders are sudden and aggressive, and the means may be more cruel. The data shows that the number of serious mental disorders is increasing. In order to prevent the occurrence of accidents and disasters in patients with mental illness, active intervention should be carried out to design an early warning and monitoring system for serious mental disorders. By collecting administrative departments of health and family planning at all levels, it is necessary to cooperate with the political and legal, public security, civil affairs, human resources, social security, and the Disabled Persons’ Federation. The established platform information is used to summarize multi-source heterogeneous data. Establish an early warning monitoring model, classify 10 risk factors from four levels, establish a risk factor assessment model, and set up different levels of treatment intervention programs. It is described from the perspectives of design ideas, design principles, and architecture design. The construction of an early warning and monitoring mechanism for serious mental disorders can effectively integrate the high-quality resources of mental health institutions at all levels, guide the rational allocation of resources, improve the management of serious mental disorders, detect the morbidity of patients with mental disorders early, and promptly intervene to reduce the risk of accidents.

We proposed a new method for generating synthetic CT on abdomen from modified Dixon (mDixon) MR data of abdomens to address the challenges of PET/MR attenuation correction (AC). AC is necessary in process of PET/MR but MR data lack photon attenuation, thus multiple methods are proposed to generate synthetic CT. However, these existing methods requires advantaged MR sequences which needs fine acquisition and huge cost consumption. To address this problem, we proposed a new method for generating synthetic CT using Patch Learning (SCG-PL). Global model of SCG-PL is transfer learning and patch model is semi-supervised classification. The advantages of our method can be summarized into two points. (1) Patch learning is a gradual learning process with gradually updating global model on remodeling patch model, so our SCG-PL method is gradually capable of generating synthetic CT. (2) Semi-supervised classification adopted in the process of patch learning, only small amount of labeled data is needed in SCG-PL, which greatly reduced the workload of radiologists. The experimental results indicate that proposed SCG-PL method can effectively generate synthetic CT image from challenging abdomen images using mDixon MR sequence data only.

In view of the high degree of personalization of embedded 3D printing products, traditional 3D printing is not applicable. This paper presents an embedded three-dimensional printing technology based on high elastic strain wireless sensor. The whole method framework includes mechanical system, control module and visual module. Firstly, three non-collinear points on the high elastic strain wireless sensor are used to align the guide plate and the model. Then, according to the position and direction of the guide hole on the high elastic strain wireless sensor, the mechanical system is controlled to guide the model guide hole to move to the center of the visual module. The characteristic parameters such as roundness, length-width ratio, diameter and center distance of the guide hole are analyzed to determine whether the guide hole is qualified. The experimental results show that compared with the traditional three-dimensional printer, the three-dimensional printer designed in this paper shortens the production cycle and improves the print resolution.

Fruit image recognition plays an important role in the fields of smart agriculture and digital medical treatment. In order to overcome the disadvantage of the deep belief networks (DBN) that ignores the local structure of the image and is difficult to learn the local features of the image, and considering that the fruit image is affected by the change of illumination, we propose a new fruit image recognition algorithm based on Census transform and DBN. Firstly, the texture features of fruit images are extracted by Census transform. Secondly, DBN is trained by Census features of fruit images. Finally, DBN is used for fruit image recognition. The experimental results show that the proposed algorithm has a strong feature learning ability, and the recognition performance is better than the traditional recognition algorithm.

The problem of clutter suppression is gaining importance because of many disadvantages. However, conventional clutter suppression methods cannot eliminate the great disturbances to radar system caused by wind turbines. An improved ridge regression algorithm is investigated to accurately estimate the spectral moment of the weather signal contaminated by wind turbine clutter (WTC) in this paper. Firstly, a weighted regression model is introduced to solve the problem that the strong collinearity of the data in the regression model leads to unstable parameter estimation. Then the optimal regression parameter in the model is obtained by generalized cross validation (GCV) to improve the estimation accuracy of weather signal. Theoretical analysis and simulation results show that the spectral moment recovered by the proposed algorithm has better accuracy and stability in lower SNR.

Because of its overall performance, the Extreme Learning Machine (ELM) has been very concerned. This paper introduces the ELM algorithm into the clutter mitigation for weather radar, and proposes a wind turbine clutter mitigation method. Firstly, building training samples. Secondly, the model parameters for ELM are examined and optimized aim to improve its overall performance. Finally, the optimized ELM algorithm is used to recover the weather signal of the contaminated range bin. Simulation results show that the proposed algorithm can realize the precise recovery of the weather signal.