Skip to main content
Disciplines
Automotive Business IT + Informatics Construction + Real Estate Electrical Engineering + Electronics Energy + Sustainability Insurance + Risk Finance + Banking Management + Leadership Marketing + Sales Mechanical Engineering + Materials
Events
DE
EN
Books
Journals
Topic Page
Marketing
Start single access now
Access for companies
EXTENDED SEARCH
Log in
Top

2018 | Book

Theoretical and Practical Solutions in Remote Sensing Read chapter Read first chapter

Computer Vision in Control Systems-3

Aerial and Satellite Image Processing

Editors: Prof. Margarita N. Favorskaya, Prof. Lakhmi C. Jain

Publisher: Springer International Publishing

Book Series : Intelligent Systems Reference Library

Part of: Springer Professional "Wirtschaft+Technik" , Springer Professional "Technik" , Springer Professional "Wirtschaft"

Login to get access
insite
SEARCH

About this book

The research book is a continuation of the authors’ previous works, which are focused on recent advances in computer vision methodologies and technical solutions using conventional and intelligent paradigms.
The book gathers selected contributions addressing aerial and satellite image processing and related fields. Topics covered include novel tensor and wave models, a new comparative morphology scheme, warping compensation in video stabilization, image deblurring based on physical processes of blur impacts, and a rapid and robust core structural verification algorithm for feature extraction in images and videos, among others. All chapters focus on practical implementations. Given the tremendous interest among researchers in the development and applications of computer vision paradigms in the field of business, engineering, medicine, security and aviation, this book offers a timely guide.

Table of Contents

Frontmatter
Chapter 1. Theoretical and Practical Solutions in Remote Sensing
Abstract
The chapter presents a brief description of chapters that contribute to theoretical and practical solutions for aerial and satellite images processing and the fields close to this scope. One can find the original investigations in the novel tensor and wave models, new scheme of comparative morphology, warping compensation in video stabilization task, image deblurring based on physical processes of blur impacts, fast and robust core structural verification algorithm for feature extraction in images and videos, among others. Each chapter involves practical implementations and explanations.
Margarita N. Favorskaya, Lakhmi C. Jain
Chapter 2. Multidimensional Image Models and Processing
Abstract
The problems of developing mathematical models and statistical algorithms for processing of multidimensional images and their sequences are presented in this chapter. Different types of random fields are taken for the basic mathematical image model. This implies two main problems associated with image modeling, namely, model analysis and synthesis. The main attention is paid to the correlation aspect, i.e. evaluation of the correlation function of a random field generated by a given model and, vice versa, development of a model generating a random field with a predetermined correlation function. For this purpose, new models (tensor and wave) and new versions of autoregressive models (with multiple roots) are suggested. The problems of image simulation on the curved surfaces are considered. The suggested models are used to synthesize the algorithms of multidimensional image processing and their sequences. The tensor filtration of imaging sequences and recursive filtration of multidimensional images, as well as the asymptotic characteristics of efficiency of random field filtration on grids of arbitrary dimension are suggested. The problem of object and anomaly detection on the background of interfering images is considered for the images of any dimension, e.g. for multi-zone data. It is shown that four equivalent forms of the optimal decision rule, which reflect various aspects of detection procedure, exist. Potential efficiency of anomaly detection is analyzed. The problems of alignment and estimation of parameters for interframe geometric image transformations are considered for multidimensional image sequences. A tensor procedure of simultaneous filtration of multidimensional image sequence and their interframe displacements are suggested. A method based on a fixed point of a complex geometric image transformation was investigated in order to evaluate large interframe displacements. Options for adaptive image processing algorithms are also discussed in this chapter. In this context, pseudo-gradient procedures are taken as a basis, as they do not require preliminary evaluation of any characteristics of the processed data. This allows to develop the high-performance algorithms that can be implemented in real-time systems.
Victor Krasheninnikov, Konstantin Vasil’ev
Chapter 3. Vision-Based Change Detection Using Comparative Morphology
Abstract
The chapter addresses the theoretical and practical aspects of the scene change detection problem with the use of computer vision techniques. It means detecting new or disappeared objects in images registered at different moments of time and possibly in various lighting, weather, and season conditions. In this chapter, we propose the new scheme of Comparative Morphology (CM) as a generalization of the Morphological Image Analysis (MIA) scheme originally proposed by Pyt’ev. The CMs are the mathematical shape theories, which solve the tasks of the image similarity estimation, image matching, and change detection by means of some special morphological models and tools. The original morphological change detection approach is based on the analysis of difference between the test image and its projection to the shape of reference image. In our generalized approach, the morphological filter-projector is substituted by the comparative morphological filter with weaker properties, which transforms the test image guided by the local shape of reference image. Following theoretical aspects are addressed in this chapter: the comparative morphology, change detection scheme based on morphological comparative filtering, diffusion morphology, and morphological filters based on guided contrasting. Following practical aspects are addressed: the pipeline for change detection in remote sensing data based on comparative morphology and implementation of change detection scheme based on both guided contrasting and diffusion morphology. The chapter also contains the results of qualitative and quantitative experiments on a wide set of real images including the public benchmark.
Yu. Vizilter, A. Rubis, O. Vygolov, S. Zheltov
Chapter 4. Methods of Filtering and Texture Segmentation of Multicomponent Images
Abstract
Some modern video systems, for example, remote sensing systems analyze the multicomponent images. Limitations of on-board technical and energy resources and video data transmission by low power and over long distance lead to strong image distortions. The filtering is used to recover the distorted by noise images for subsequent tasks of image processing, such as detection of texture regions and objects of interest, estimations of their parameters, classification, and recognition. Multicomponent images can be represented as the multidimensional signals and have significantly greater statistical redundancy than one-component images. This redundancy would be appropriate to improve a quality of image restoration. Special cases of multicomponent images are color RGB images, each color component of which is a g-bit digital halftone image. The nature of the statistical relationship between elements within the digital halftone image and among the elements of color components allows to suggest an approximation for 3D color images using a Markov chain with several states and for bit binary image applying a 3D Markov chain with two states. The proposed filtering method is based on an approximation the multicomponent images using a 3D Markov chain and on an efficient use of statistical redundancy of multicomponent images. This method requires small computational resources and is effective with signal-to-noise ratio at the input of receiver up to –9 dB. Real images have areas with varying degrees of detail and different statistical characteristics. The authors propose to improve the accuracy of the statistical characteristics of each local region within an image and between the color components to improve a quality of the reconstructed image. A sliding window is used to estimate the local statistical characteristics of an image. The proposed method allows to detect the small objects and contours of objects more accurately in image distorted by white Gaussian noise. A method of texture regions’ detection on the reconstructed images based on Markov random fields is proposed. An estimate of the probability of a transition between image elements is used as the texture feature. The method efficiently detects the texture regions with different statistical characteristics and makes it possible to reduce the computational costs.
E. Medvedeva, I. Trubin, E. Kurbatova
Chapter 5. Extraction and Selection of Objects in Digital Images by the Use of Straight Edges Segments
Abstract
New method for finding geometric structures in digital gray-level images is proposed. The method is based on grouping straight line segments, which correspond to the edges of the object. It includes extraction of straight line segments by oriented filtering of gradient image and gives the ordered list of segments with the endpoints’ coordinates for each segment. Adaptive algorithm for straight edge segments extraction is developed that uses angle adjustment of oriented filter in order to extract the line corresponding to the real edges accurately. This algorithm permits the extraction and localization of artificial objects with the rectangular or polygonal shape in digital images. Perceptual grouping approach is applied to extracted segments in order to obtain the simple and complex structures of lines using their crossings. Proposed approach uses the points of intersection of ordered segments as the main property of object structure and also takes into account some specific properties of grouped lines, such as the anti-parallelism, proximity, and adjacency. At the first step, the simple structures are obtained by lines grouping taking into consideration all crossing lines or only part of them. At the second step, these simple structures are joined allowing for restrictions. Initial image is transformed to a collection of closed rectangular or polygonal structures with their locations and orientations. Structures obtained by this method represent an intermediate-level description of interesting objects, which have polygonal view (buildings, parts of roads, bridges, and some natural places of landscape). Application with real aerial and satellite images shows a good ability to separate and extract the specific objects like buildings and other line-segment-rich structures.
V. Yu. Volkov
Chapter 6. Automated Decision Making in Road Traffic Monitoring by On-board Unmanned Aerial Vehicle System
Abstract
The study is dedicated to solving the target issues of the ground traffic monitoring aided by the Unmanned Aerial Vehicles (UAV) based on applying the on-board computer vision systems. The classification of the road situations using images obtained after Traffic Accident (TA) is based on the feature set, facts, and attributes specified directly and/or indirectly on a possible situation class. The hierarchical structure of description of a road situation observable after the TA event is developed. For decision making, the production model of knowledge representation and corresponding Knowledge Base (KB) is offered to use. The issues related to decision making for recognition of the occurring traffic situations have been considered. The analysis of the strategies have been carried out based on the principles of minimizing the overall losses, limiting the admissible UAV flight altitude, and ensuring the required class recognition reliability. The models describing the functional criteria of the losses, flight safety of the UAV, and reliability of class recognition have been proposed. It has been shown that applying the minimum loss criterion ensures considerable savings of resources under different ratio of the loss quotients. The example for classification of a road incident using the real images is given.
Nikolay Kim, Nikolay Bodunkov
Chapter 7. Warping Techniques in Video Stabilization
Abstract
Digital image and video stabilization are crucial issues in many surveillance systems. Good stabilization of the raw data provides a successful processing of visual materials. At present, the main approach directs on the search of the trade-offs between 2D and 3D stabilization methods in order to derive the benefits of both techniques. Our contribution is twofold. First, the multi-layered motion fields are applied in the warping during stabilization. For this purpose, the term “Structure-From-Layered-Motion” was introduced. Second, the warping and inpainting of the frame boundaries are executed using a pseudo-panoramic key frame and the multi-layered motion fields. Such inpainting permits to restore fast the cropped stabilized frames up to the sizes of the original non-stabilized frames. The dataset Sports Videos in the Wild, as well as the additional non-stationary video sequences, were used in experiments, which demonstrated good visibility results with a preserving of the frame sizes.
Margarita N. Favorskaya, Vladimir V. Buryachenko
Chapter 8. Image Deblurring Based on Physical Processes of Blur Impacts
Abstract
Main methods of image deblurring, as well as their advantages and disadvantages, are considered in this chapter. It is revealed that these methods do not take into account the physical processes that occur during a blur impact. It is shown that instead of the continuous function of illuminating intensity it is convenient to consider its discrete-analog modification, in which the size of a discrete will be equal to the size of a photosensitive element. In the case of the stationary images, such replacement will not impact on the formed video signal in any way. On the contrary, when objects move relative to the fixed sensor, the processes of blur may appear. Due to these reasons, the models of various blur types, such as the linear, non-linear, and vibrational models, were built. Note that these models are subdivided on the models with the small and large blur according to the ratio of movement displacement of an object and its length. The constructed models are the systems of the simple non-uniform algebraic equations with non-singular matrixes that allows to build the deblurring algorithms. The proposed algorithms were tested using natural images obtained from the imaging device layout.
Andrei Bogoslovsky, Irina Zhigulina, Eugene Bogoslovsky, Vitaly Vasilyev
Chapter 9. Core Algorithm for Structural Verification of Keypoint Matches
Abstract
Outlier elimination is a crucial stage in keypoints-based methods, especially in extreme conditions. In this chapter, a fast and robust “Core” Structural Verification Algorithm (CSVA) for a variety of applications and feature extraction methods is developed. The proposed algorithm pipeline involves many-to-one matches’ exclusion, the improved Hough clustering of keypoint matches, and cluster verification procedure. The Hough clustering is improved through an accurate incorporation of translation parameters of similarity transform and “partially ignoring” the boundary impact using two displaced accumulators. The cluster verification procedure involves the use of modified RANSAC. It is also shown that the use of the nearest neighbour ratio may eliminate too many inliers, when two images are matched (especially in extreme conditions), and the preferable method is a simple many-to-one matches exclusion. The theory and experiment prove the propriety of the suggested parameters, algorithms, and modifications. The developed cluster analysis algorithms are robust and computationally efficient at the same time. These algorithms use some specific information (rigidity of objects in a scene), consume low volume memory and only 3 ms in average on a standard Intel i7 processor for verification of 1,000 matches (i.e. magnitudes less than the time needed to generate those matches). The CSVA has been successfully applied to practical tasks with minor adaptation, such as the matching of 3D indoor scenes, retrieval of images of 3D scenes based on the concept of Bag of Words (BoWs), and matching of aerial and cosmic photographs with strong appearance changes caused by season, day-time, and viewpoint variation. Eliminating a huge number of outliers using geometrical constraints allowed to reach the reliability and accuracy in all solutions.
Roman O. Malashin
Chapter 10. Strip-Invariants of Double-Sided Matrix Transformation of Images
Abstract
In this chapter, a strip-method suitable for reducing pulse interference in communication channels, cryptography, steganography, and other applications is considered. The invariants to fragmentation and double-sided matrix transformation of images provide the noise immunity and transmission security. The chapter contains new definitions of invariants, as well as invariant images of the first and second types. Moreover, tasks of analyzing and synthesizing both invariants and corresponding transformation matrices are set forth too. The criteria of their existences are derived and methods for creation of invariant images using eigenvectors of transforming matrices are proposed. Some cases of complex and multiple eigenvalues of a direct transformation matrix are considered. It was proposed to solve the problem of finding the matrices of direct and inverse transformations by means of a given set of invariant images. The solution of the task of arraying the matrix of double-sided transformation according to a given set of invariant images is suggested.
Leonid Mironovsky, Valery Slaev
Chapter 11. The Fibonacci Numeral System for Computer Vision
Abstract
One of the most important challenges when creating efficient systems of technical vision is the development of efficient methods for enhancing the speed and noise-resistance properties of the digital devices involved in the system. The devices composed of counters and decoders occupy a special niche among the system’s digital tools. One of the most common ways of creating noise-proof devices is providing special coding tricks dealing with their informational redundancy. Various frameworks make that possible, but nowadays, an acute interest is attracted to noise-proof numeral systems, among which the Fibonacci system is the most famous. The latter helps generate the so-called Fibonacci codes, which can be effectively applied to the computer vision systems; in particular when developing counting devices based on Fibonacci counters, as well as the corresponding decoders. However, the Fibonacci counters usually pass from the minimal form of representation of Fibonacci numbers to their maximal form by recurring to the special operations catamorphisms and anamorphisms (or “folds” and “unfolds”). The latter makes the counters quite complicated and time-consuming. In this chapter, we propose a new version of the Fibonacci counter that relies only on the minimal representation form of Fibonacci numerals and thus leads to the counter’s faster calculation speed and a higher level of the noise-resistance. Based on the above-mentioned features, we also present the appropriate fast algorithm implementing the noise-proof computation and the corresponding fractal decoder. The first part of the chapter provides the estimates of the new method’s noise-immunity, as well as that of its components. The second problem studied in the chapter concerns the efficiency of the existing algorithm of Fibonacci representation in the minimal form. Based on this examination, we propose a modernization of the existing algorithm aiming at increasing its calculation speed. The third object of the chapter is the comparative analysis of the Fibonacci decoders and the development of the fractal decoder of the latter.
Oleksiy A. Borysenko, Vyacheslav V. Kalashnikov, Nataliya I. Kalashnykova, Svetlana M. Matsenko
Metadata
Title
Computer Vision in Control Systems-3
Editors
Prof. Margarita N. Favorskaya
Prof. Lakhmi C. Jain
Copyright Year
2018
Electronic ISBN
978-3-319-67516-9
Print ISBN
978-3-319-67515-2
DOI
https://doi.org/10.1007/978-3-319-67516-9

Premium Partner

    Image Credits