Volumetric - Prostor's model. Space models of space Space and dynamic models

Іsnuє class zavdan, in some particular importance nabuvaє іnformatsija strukturі і rusі ob'єktіv foldії ї sceni (videoposterezhennya v zakrytih primіshchennyah, іѕtsyah great skuchennya people, rukhom robotechnіchnіchіchіnіkіv, zоrоzhennі ruhіvі transport t.d.d.). Consequences of images are a collapsible information resource, structured in space in hours and shared information in visually rich signals, the form of representation in a computer and physical models of dynamic objects, phenomena, processes. The new technical possibilities of digital image processing allow to frequently correct such specificity of images, victoriously reaching the cognitive theory of human perception of natural images.

The analysis of the space-time obsyagu of these allows showing both static and dynamic signs of objects of caution. And here the task of recognition can be found, as a classification of the totality of the stations, as a classification of trajectories, the solution of which can be found by classical methods of recognition, because temporal transitions can give rise to a transformation of the image, which is not described by the usual analytical deposits; Also, the order of the recognition of dynamic objects is blamed for the recognition of active events and subdivisions, for example, for the detection of unauthorized events in places where people are crowded, or the scenes assigned to the genre for indexing in multimedia data bases. To look at the task of recognizing objects and subdivisions behind image sequences as a single process, then the most important hierarchical process with elements of parallel processing of the skin layer.

Improving the technical means of collecting and processing information from visually static images (photos) and video sequences will require further development of methods and algorithms for their processing, analysis of situations and recognition of images of objects. Pochatkov, the theoretical formulation of the problem of recognition of the image is carried out until 1960-1970. and endorsed by a number of robotic authors. Setting the task of recognizing images can be changed depending on the task of recognizing objects, tasks for analyzing scenes until the task of recognizing the problems of the machine dawn. In this case, the system accepts intellectual solutions, which are based on the methods of recognizing the images of that image, and the input information of a complex type. Before it, one can see it as an image taken in a wide range of the electromagnetic spectrum (ultraviolet, apparently, infra-red and in), so is the information in looking at sound images and location data. Regardless of the physical nature, such information can be given from looking at real images of objects and specific images. Radiometric data - the whole flat image of the scene, presented in a perspective view of an orthogonal projection. The stench is shaped by a path of vimiryuvannya іntensivnostі elektromagnіtnіh hvil pevnogo spectral range, scho vіdobrazhayutsya аbo vіpromіnuyuyutsya by the objects of the scene. Sound vicorist photometric data, vymiryanі in the visible spectral range - monochromatic (yaskra) * or color images: As well as coordinates for all points of the scene, such an array of location data can be called an image of the depth of the scene. Use simplified image models (for example, models of affinity projections, represented by weakly promising, para-perspective, orthogonal and parallel projections), in which the depth of the scene is respected by a constant value, and the location image of the scene does not carry core information. Sound information may have a different additional sub-character.

Most of the photometric data are quickly collected. Location information, as a rule, is calculated for data, which is taken from special attachments (for example, a laser ranger, a radar) or from a stereoscopic method for analyzing images of yaskravikh. Due to the difficulties in the operational processing of location data (especially for scenes with the shape of visual objects, which are easily changed), the task of describing the scene is overestimated by one visual image, that is. The task of the monocular zor sprinyatt scene. It is impossible to deduce the geometry of a scene from a single image in a single image. Just for the sake of singing the borders to complete simple model scenes and the obviousness of a priori information about the expanse of the expansion of objects, we try to induce a new three-dimensional description of one image. One of the ways to get out of this situation is the processing and analysis of video sequences, taking into account one or several video cameras, installed unruly or moving around in space.

Thus, the image is the main form of presentation of information about the real world, and further development of methods for transforming the semantic analysis of both images and sequences is needed. One of the most important directions in the development of such intellectual systems is the automation of the choice of methods for describing the transformation of an image with the improvement of its informational nature and the purpose of recognition at the early stages of image processing.

The first research papers from the USA (Louisiana State University, Carnegie Mellon University, Pittsburgh), Sweden ("Computational Vision and Active Perception Laboratory (CVAP), Department of Numerical Analysis and Computer Science), France (INRIA), Great Britain (University of Leeds) , FRN (University of Karlsruhe), Austria (University of Queensland), Japan, China (School of Computer Science, Fudan University) s image sequences and recognition of dynamic objects were published in the 1980s. appear in Russia: in Moscow (MDU, MAI (DTU), MFTI, DerzhNDI AS), St. Petersburg (SPbDU, GUAP, FSUE GOI, LOMO), Ryazan (RGRTU), Samara (SDAU), Voronezh (VDU ), Yaroslavl (YarSU), Kirov (VDU), Taganrozi (TTI SFU), Novosibirsk (NSU), Tomsk (TDPU), Irkutsk (IRDU), Ulan-Ude (VSTU) and in. at tsіy galuzі, as an academician of the Russian Academy of Sciences, Doctor of Technical Sciences Yu. M. Mestetskiy , d.t.s. B. A. Alpatov et al. Significant successes have been achieved today with video warning systems, systems for authenticating individuals behind images, and so on. However, there are unresolved problems in recognizing dynamic images through the folding and variability of the behavior of objects in the real world. In this way, Denmark will directly require thorough models, methods and algorithms for recognizing dynamic objects and image sequences in different ranges of electromagnetic imaging, which will allow the video imaging system to be developed on a fairly new level.

The method of dissertation work is to improve the efficiency of recognition of dynamic objects, their active roles and subdivisions in folding scenes for image sequences for systems of external and internal video warning.

The meta was set to signify the need to complete the upcoming tasks:

To carry out an analysis of the methods for evaluating the movement and significance of the recognition of the movement of objects behind the set of subsequent images, methods for segmenting dynamic objects and semantic analysis of folding scenes, as well as the steps to the recognition systems for recognition of the importance of dynamic objects of a different purpose.

Expand the model of recognizing static and dynamic images, based on the hierarchical procedure for processing hour rows, zoning, image sequences.

To develop a method for evaluating the dynamics of dynamic structures using space-hourly information taken in different ranges of electromagnetic vibration, which allows you to choose the method of segmentation fallow according to the nature of the disruption, by the same token, to adapt adaptively to the distribution of dynamics.

To create a model of dynamic movement of dynamic structures in a folding scene, which allows, on the basis of taking away odometric data, the trajectory of movement of dynamic structures and the development of hypotheses about the basis of visual objects on the basis of analysis of the prehistory of disturbances.

To develop a complex algorithm of segmentation, sho vrakhovuє sukupnіst vyavlenih znaka dynamіchnyh struktіv pri vіlnymi prіklіnіh svіlіnі prіktіv v proektіy ob'єktіv, ґntuyuyuchis іn model і bagatoriіvnevoy ruhu u folded scenes.

Develop a method for recognizing dynamic images represented in terms of formal grammar and scene videographer, based on the method of collective decision making, as well as a method for recognizing active characters and subdivisions in a folding scene Bayesian measure.

On the basis of the developed methods and models, to design experimental systems of various recognition; are used for processing sequences of images of objects, which are characterized by fixation and sufficient set of 2£>-projections, i-recognition of dynamic images. folding scenes.

Methods, follow-up. For the dissertation work, the methods of the theory of pattern recognition, descriptive theory of image recognition, the theory of signal processing, methods of vector analysis and tensor calculation, as well as the theory of groups, the theory of formal grammars were awarded.

The scientific novelty of dissertation work is in the offensive:

1. The new model is motivated by the dinemino, the same, the rodshire of the Irarchychny riwanes of the segment (behind the local thorough vectors of the rush), the roseshnes (o'clock that are perceptions). scenes based on understanding the maximum dynamic invariant.

2. Розширена дескриптивна теорія розпізнавання зображень запровадженням чотирьох нових принципів: облік мети розпізнавання на початкових стадіях аналізу, розпізнавання поведінки динамічних об'єктів, оцінка передісторії, змінна кількість об'єктів спостереження, що дозволяє підвищити якість розпізнавання об'єктів, що рухаються за рахунок promotion of informativeness result.

3. First, an adaptive space-hour method for evaluating movement in synchronous sequences of the visible and infrared ranges of electromagnetic vibration was developed, which allows one to recognize the signs of movement on various hierarchical levels, judging by the importance of both types of sequences.

4. A new model of bagator_vnevogo rush has been developed; that allows you to decompose the scene on the edge of the river; do not > get married; by a globally accepted footing to the foreground and background, which allows you to win more reliable segmentation of the image of objects; folding perspective scenes.

5: Chi priming? that prompting; new; zagalneniya algorithm segmentation of dynamic objects; h, zastosuvannyam, impersonal sign that includes the history of behavior; and allows you to see the dynamics of other visual objects, and the interaction of objects near the scene (overshooting projections; the appearance / appearance of objects from the field of view of the video sensor) with the improvement of group transformations; and to the previously proponated analysis of the main part of the projections of the object (from two suicidal frames) from the assessment of integral and invariant estimates.

6. Модифікований метод колективного прийняття рішень, що відрізняється знаходженням ознак міжкадрових проекцій об'єкта і що дозволяє враховувати передісторію спостережень для розпізнавання активних дій та подій на основі байєсівської мережі, а також запропоновані чотири види псевдо-відстаней для знаходження міри подібності v динамічних образів з reference speakers

practical significance. Approved in dissertation robotic methods, these algorithms are recognized for practical application in the monitoring of motor transport accidents in rich and smog Russia within the framework of the sovereign project "Safe Place", in systems for automated control of various technological processes and video sequences, in systems for monitoring the monitoring of traffic control. as well as in systems for identifying objects on aerial photographs and recognizing landscape images

Realization of results of work. Distributed programs for registration with the Russian registry of programs for EOM: the program “Segmentation of the handwritten text image (SegPic)” (certificate No. 2008614243, Moscow, 5 September 2008); “Motion Estimation” program (certificate No. 2009611014, Moscow, February 16, 2009); the program "Localization of an individual (FaceDetection)" (certificate No. 2009611010, Moscow, February 16, 2009); the program "System for superimposing visual natural effects on a static image (Natural effects imitation)" (certificate No. 2009612794, Moscow, 30 April 2009); program "Visual detection of smoke (SmokeDetection)" (certificate No. 2009612795, Moscow, April 30, 2009); "Program for visual registration of sovereign license plates of motor vehicles in the case of rich traffic Russia (FNX CTRAnalyzer)" (certificate No. 2010612795, Moscow, March 23, 2010) Moscow, February 31, 2010

An act was taken on the transfer of that algorithmic and software support for recognizing cases in refrigerators on a warehouse line (VAT KZG "Biryusa", Krasnoyarsk), for identifying objects on landscape images (Concern radio equipment "Vega", VAT KB "Promіn" , м. Рибінськ Ярославської області), для сегментації лісової рослинності за набором послідовних аерофотознімків (ТОВ «Альтекс Геоматика», м. Москва), для виявлення пластин державних реєстраційних знаків автотранспортних засобів у відеопослідовності при багатопоточному русі та підвищення якості їх відображення^ (УГИБД Красноярському edge, m. Krasnoyarsk).

The developed algorithms and software security are developed in the initial process during the study of the disciplines "Intellectual Data Processing", "Computer Technologies in Science and Light", "Theoretical Fundamentals of Digital Image Processing", "Image Recognition", "Neural Measurements", Algorithms for processing images”, “Algorithms for processing video sequences”, “Analysis of scenes and machine vision” at the Siberian State Aerospace University named after Academician M.F. Reshetnov (SibDAU).

The reliability of the results obtained from the dissertation work is ensured by the correctness of the victorious methods and the mathematical rigor of their revisions, as well as the validity of the formulation of the position of the work in the results of their experimental verification.

The main provisions for blaming the Zakhist:

1. The model for processing and recognizing dynamic images in folding scenes, is completely expanded by ієrarchical equals of segmentation and recognizing not only objects, but also active ones.

2. Expansion of the descriptive theory of recognition of images for time series (image sequences) with additional advancement of the information content of data analysis in the space area, and in the time zone of the warehouse.

3. Adaptive space-hour method for evaluating movement. on the basis of tensor manifestations of local 31 effects in synchronous sequences of the visible and infrared ranges of electromagnetic vibration.

4. The model of the bagatory movement in folding scenes, which expands the decomposition of perspective scenes on the edge for a reliable analysis of the trajectory of the movement of objects.

5. A more advanced algorithm for segmenting dynamic objects, which allows, on the basis of group transformations and proponation of integral and invariant estimates, to reveal the overlap of projections of objects, the appearance / appearance of objects from the field of view of the video sensor.

6. Methods for recognizing dynamic images based on the modified method of collective acceptance of the solution and recognition of pseudo-views in metric spaces, as well as active scenes and subdivisions in folding scenes.

Approbation of robots. The main provisions of the results of the dissertation studies were discussed and discussed at the 10th international conference "Pattern Recognition and Image Analysis: Modern Information Technologies", (S.-Petersburg, 2010), the international congress "Ultra Modern Telecommunications and Control Systems ICUMT2010"; XII international symposium on non-parametric methods in cybernetic and system analysis (Krasnoyarsk, 2010), II international symposium "Intelligent Decision-Technologies - IDT 2010" (Baltimore, 2010), III international conference. "Automation, Control? and Information Technology - AOIT-ICT"2010" (Novosibirsk, 2010), 10th, 11th and 12th international conferences and exhibitions "Digital processing of signals and її zastosuvannya" (Moscow, 2008 - 2010), X international scientific and technical conference "Theoretical and applied power of modern information technologies" (Ulan-Ude, 2009), IX international scientific and technical conference "Cybernetics and high technologies of the XXI century" (Voronіzh, 2008), all-Russian conference image processing” (Krasnoyarsk, 2007), at the X, XI and XIII international scientific conferences “Reshetnevskiy reading” (Krasnoyarsk, 2006, 2007, 2009), as well as at the scientific seminars of the State University of Aerospace Appliances - Petersburg, 2009) , Institute for Calculating Modeling of CO

RAS (Krasnoyarsk, 2009), Institute of Image Processing Systems RAS (Samara, 2010).

Publications. Based on the results of the dissertation research, 53 other papers were published, including 1 monograph, 26 articles (of which 14 articles were published by scholars, included to the list of the Higher Attestation Commission, 2 articles were published by scholars, refurbished by Thomson Reuters: Science Citation Index Expanded / Conference Proceeding Index ”), 19 abstracts of additional notes, 7 certificates registered with the Russian Register of Programs for EOM, as well as 3 stars from NDR.

Special entry. These are the main results, contributions to the dissertation, including the formulation of the task of both mathematical and algorithmic solutions, taken by the author in a special way, or vikonated by scientific curiosity and for an uninterrupted participation. For the materials of the work, two dissertations on the scientific level of a candidate of technical sciences were stolen, and the author was an official scientific kerivnik.

Robotic structures. The work is composed of an entry, six divisions, visnovkiv, a bibliographic list. The main text of the dissertation contains 326 pages, the summary is illustrated with 63 figures and 23 tables. Bibliographic list contains 232 names.

6.7 Wisnovki on choli

In this section, the structure and main functions of the experimental software complex "ZROEL", v.1.02, are reviewed; vykonuє system ієarkhіchnu ієrkhіchnu obrobkі sequentі izobrazhen аѕ up to vyshchih іvnіv rіvnіv rіvnіvnіnі ob'єktіv i podіy. graphs, measures and classifiers A number of low-level modules of the system are operated in automatic mode

Experimental studies on the basis of this software package were carried out on a number of video sequences and infrared sequences from the test base "OTCBVS^07", on test video sequences "Hamburg taxi", "Rubik cube". "Silent", as well as on the original video material. Five methods of evaluating movement were tested. It has been experimentally shown that the blocking method and the proponation method for the infrared sequence show similar values ​​and are the least accurate. The proponation method for video sequence and the method of stepping behind point features demonstrate similar results. In this case, the divisions of tensorial pidkhіd have less obligatory computer calculation against the method of stitching with point singularities. Composite variable of synchronization of video sequences and infrared sequences to the extent of variation for significance of the modulus of the speed vector and in the minds of reduced illumination of the scene.

Для розпізнавання візуальних об'єктів застосовувалися чотири види псевдо-відстаней (псевдо-відстань Хаусдорффа, Громова-Хаусдорффа, Фреше, природна псевдо-відстань) для знаходження міри подібності вхідних динамічних образів з еталонними динамічними образами (залежно від уявлення динамічної ознаки - множини безлічі векторів , impersonal functions). Voni showed their ability to create images with acceptable morphological transformations. The integration of the normalization of the assessment of the shape to the contour Kc of the upper part of the projection of the object between mentally significant frames and the area of ​​the upper part of the 5e and the invariant assessment - the correlation function of the upper parts of the projections Fcor. Zastosuvannya modified method of collective acceptance of the solution allows "to see" in the distance warning of incoming images (falls of overlapping projections of objects, creating a scene in the light of the fire) and choose the most appropriate caution. Experiments have shown that using a modified method of collective decision making increases the accuracy of recognition by an average of 24-29%.

Experimental results of traffic evaluation, segmentation and recognition of objects were rejected on test sequences of images ("Hamburg taxi", "Rubik cube". "Silent", video sequences and infrared sequences from the test base "OTSVVS" 07") people were tested from test databases PETS, CAVIAR, VACE The best results show recognition for two sequences, as well as the best experimental results were achieved when recognizing periodic active people, which do not change in groups (walking, big, raising hands).

At the completion of the sixth division, such applied projects were reviewed, such as “Visual registration of sovereign license plates in motor vehicles in a rich flow of Russia”, “System for identifying refrigerator case models behind images”, “Algorithms. processing of segmentation, landscape images. Identification of objects. Algorithmic and software security was transferred to the applicants, organizations: The results of test operation showed the practicality of software security, developed on the basis of models and methods adopted by dissertation robots.

WISNOVOK

At the dissertation robot, the most important scientific and technical problem of processing space-time data, removing from the sequences of the visible and infra-red ranges in the electromagnetic visualization, and the recognition of dynamic images in folding scenes was posed. The system of hierarchical methods of processing and observation is a sign of space-time data and the methodological basis for the implementation of applied tasks at the gallery of video surveillance.

In the introduction, the actuality of the dissertation work is outlined, the metaphor is formulated and the task of research is set, the scientific novelty and practical value of the results are shown, the main provisions are presented, which are to be blamed for the zahist.

In the first section, it was shown that visual objects in the video sequence are characterized by a more rich vector sign, lower images in the classical setting of the recognition of static images. At dissertation robots, clarifying stages are introduced at the middle and highest levels of processing, which may be important for dynamic images.

The classification of the main types of recognition for static images, static scenes with elements of chaos and sequences of images was motivated, which reflects the historical nature of the development of mathematical methods in this gallery. A report analysis of methods for evaluating movement, algorithms for segmenting objects that collapse, and methods for interpreting sub-folding scenes has been carried out.

Розглянуто існуючі комерційні апаратно-програмні комплекси в таких галузях, як моніторинг транспортних засобів різного призначення, обробка спортивних відеоматеріалів, забезпечення безпеки (розпізнавання осіб, несанкціоноване проникнення людей на територію, що охороняється), Також аналізуються дослідницькі розробки для систем відеоспостереження.

At the end of Chapter 1, the statement of the problem of space-hour processing of image sequences is introduced, it is presented in three equal and five stages of processing and recognition of visual information from image sequences.

In another dissertation, the formal model of processing and recognizing objects for their static images and sequences of images was broken down. Promising admissible visualization at the expanse of the image and the expanse is a sign for the direct task and the return task. Rules for invariant virtual functions and a specified maximum dynamic invariant are introduced. When recognizing the trajectories of various images in the rich space, the sign can be overshadowed. When the projections of objects change, the significance of the narrowed maximum dynamical invariant becomes more collapsible, and in such situations it becomes impossible to achieve.

The main principles of the descriptive theory of image recognition are examined, which form the basis of regular methods for choosing the synthesis of algorithmic procedures for processing information in image recognition. Запропоновано додаткові принципи, що розширюють дескриптивну теорію для динамічних зображень: облік мети розпізнавання на початкових стадіях обробки послідовності зображень, розпізнавання ситуацій динамічних об'єктів, оцінка передісторії динамічних об'єктів, змінна кількість об'єктів спостереження в складних сценах.

Reportedly, the problem of looking for the main signs for the analysis of sequences of images is examined depending on the type of capture (in times of single-angle capture), the movement of the video sensor and the visibility of objects that collapse in the visibility zone. An inventory of several situations in the open space has been given, a sign of a complicated task for the world.

At the third stage, the stages of processing the sequences of the image and the recognition of objects, active actions, similar to the genre of the scene were formulated. Steps to achieve the last hierarchical nature of the processing of visual information. Also, think about the exchange of hierarchical methods in the space-hour processing of image sequences.

The classification of dynamic regions of the image is carried out by means of analysis of the internal values ​​31) of the structural tensor, the property vectors of which are assigned to the local effects of the intensities of the images of the suicidal frames and are scored for the assessment of the local orientations of the dynamic regions. A new method for evaluating the movement of the space-hour communication of visible and infrared ranges of variation based on the tensor approach has been grounded. The possibility of stowing a spaciously changing core, adaptive to the expansion of that orientation of sharpening of a point, was examined. Adaptation sharpening, on the back of the head, the shape of the stake, and then reshaping after 2-3 iterations on the shape of the oriented ellipse allows polypsity assessment of the orientation of the structures on the image. Such a strategy improves the estimates of gradients in the space-time data set.

Evaluation of local parameters in a rush to go through the way of calculation of geometric primitives and special points of a local region. In this manner, the assessment of local signs of the development of regions is the basis for the development of further hypotheses of the presence of visual objects of that third class. The choice of synchronous video sequences and infrared sequences allows us to improve the results of segmentation of the regions on the image of the local vectors in the region.

It is shown that it is possible to evaluate inter-color images on the basis of multi-colored gradient methods, prompted by all directions at the skin points of the cordon, by vector methods from multiple ordinal statistics about color images, as well as the development of tensor approaches within the framework of multi-color methods. Methods for clarifying contour information may be important for regions due to the large number of permissible projections.

In the fourth division, a rich model of movement was inspired based on the structures of movement, which reflects the dynamics of objects in real scenes and expands the scene's scenery, which is influenced by objects of interest and unruly decay.

There are further models of dynamic objects on the plane, which are based on the theory of compact groups Li. Presented models for design transformation and various models of Athenian transformation. Such transformations are good at describing the structure of the world with a large number of projections (technogenic objects). Submission of structures from an uncircumscribed number of projections (anthropogenic objects) and affinity chi projective transformations are accompanied by low additive minds (zocrema, perhaps in the distance of objects from the video sensor, small-sized objects too). This theorem, brought by L. S. Pontryaginim, is suggested, on the basis of which it was possible to know the internal automorphism of group coordinates, which describes the object with accuracy to the point of destruction between the substantive frames. The size of the damage depends on the method of assessing the turnover of the inter-personnel cost, divided into 3" divisions.

The expansion of the permissible transitions between groups of transformations through the duality of nature 2£)-image (introduction of a change in the projection of the object and the visual transformation of a number of objects: (vzaєmodiya of objects)) has been motivated. Knowed the criterion, Yaki, when snake groups, the rewinding of Active, in the scenes, and the male, іnthegovani, the contour of the KS of the KSAMI POSITIONS STRIMICAL PARTICS TO ONELARITIA ROSENIS - Corleta of the Korletovyna - Corleta groups Li s'd, yakі allow estimating the steps of indolence and revealing the nature of the disruption of objects that are guarded.

Also, a model of the rehistory of the movement of objects in sequences of images was motivated, which includes timing rows of trajectories of movement, change of the form of the object in the case of yoga in space, as well as change of the form of the object, showing the appearance of the scene / recognition of the object from the field of the sensor gap (it is used to recognize active actions and sub-scenes). one

The algorithm of segmentation of objects in folding scenes has been broken down, which includes the reverse folding of segmentation (redrawing the image, appearing that object from the field of the camera view, moving to the camera), which includes three steps: pre-segmentation, segmentation and post-segmentation. For the dermal subdivision, the formulation of tasks, visual and visual data, development of block diagrams of algorithms that allow segmentation of folded scenes, victorious transcendence of synchronous sequences of different ranges of visualization.

At the fifth division, the process of recognizing dynamic images is considered, which is based on formal grammar, the videographer scene modifications and the method of collective decision making. A dynamic scene with a rich movement creates a structure that changes in the hour, it is enough to win the structural methods of recognition. Tririval contextual grammar of recognizing folded scenes has been propagated from a bagatarial movement of objects realizing two tasks: the task of parsing the sequence of the image and the task of syntactic analysis of the scene.

The most basic way of semantic description of the scene is videograph, prompting by the method of hierarchical grouping. On the basis of complex signs of the lower level, local expanses of structures are formed, stations in hours, local expanses of objects and a videograph of the scene will be formed, which includes the recognition of expanses of objects, the collection of authorities in them, as well as space-hours links between them.

Modifications of the method of collective adoption of the solution are based on the domestic procedures of recognition. On the first level, there is a recognition of the reliability of the image of your and other areas of competence. On the other side of the river, it is gaining chivalry of the highest rule, the competence of which is maximum in the assigned area. Incited to develop for pseudo-visions with the knowledge of the world of similarity of input dynamic images with reference dynamic images in the fallow in the representation of dynamic signs - multiplies of numerical characteristics, multiplies of vectors, multiplies of functions.

When recognizing a sub-videograph of a folding scene, it expands to a sub-videograph: An object-deposit model of a dynamic object is prompted. As a function of consistency, the simplest classifiers are found in the space of signs (for example, after the method of ^-average), so that the setting is established for a large number of templates associated with a previously known object. The method and formation of templates in the projections of visual objects is reviewed.

The videographer will be based on the Markov measure. The ways of revealing active media agents, as well as the procedure for encouraging that videographer to be identified for recognition, near the stage, are examined. Each skin suit will have its own model, as it is trained on test stocks. Iyavlennya podіy zvoditsya to clusterization of active dіy, scho subsequently vykonuyutsya, on the basis of the Bayesian approach. Recursively recursively split-matrices of your coefficients at the input video sequence and consistency with the references, subdivisions, and otrimanimy at the stage of training. This information is intended for the genre of the scene and, if necessary, indexing the video sequence in the data base. The scheme for understanding and interpreting images and video materials for indexing in multimedia Internet databases has been decomposed.

A description of the experimental software complex "SPOER", v.l.02, with a summary of image sequences and recognition of rough objects and pods, is presented at the sixth division. Vіn vykonuє system ієrarchіchnu ієrkhіchnu ієrkhіchnu obrobkі sequentі izobrazhen аѕ up to іnіvіshіh rіvnіv rіvnіv rіvnіvnіnі ob'єktіv і podіy. Vіn є avtomatizirovannoy system, scho vimagaє participation of people for nauchannya that nalashtuvannya graphіvіn, mérezh і klаsifikatorіv. A number of low modules of the system work automatically.

In experimental investigations carried out with the help of the software complex "SPOER", v.l.02, video sequences and infrared sequences of images from the test base "OTCBVS" 07, tests of video sequences "Hamburg taxi", "Rubik cube" were tested. Five methods of evaluating the speed were tested. The proponation method for video sequence demonstrating the most accurate results and using the least amount of computer calculation in comparison with other methods.

For the Rospіznovannaya of the Obzalny, the permissions of the morphologists of the projectoviloviloviloviloviloi, Zastosuvannya modified method of collective acceptance of the decision allows "to see" in the distance warning of incoming images (falls of overlapping projections of objects, visual creation of scenes in the light of the fire) and choose the most appropriate ones. Experiments have shown that using a modified method of collective decision making increases the accuracy of recognition by an average of 24-29%.

Experimental results of evaluation-ruhu; segmentation and recognition of objects were removed on the test image sequences (Hamburg taxi, Rubik cube. Silent, video sequences and infrared sequences from the OTCBVS*07 test database). For the recognition of active diy people, butts from the test bases PETS, CAVIAR, VACE were tested. The best results show recognition for two sequences. Also, the best experimental results were achieved when recognizing periodic active children who did not change in groups (walking, big, raising hands). Hibni spratsovuvannya obumovleniya zasvіchennyam and the presence of shadows near a number of scenes.

На базі експериментального комплексу «ЗРОЄЯ», V. 1.02 були розроблені системи обробки відеоінформації різного цільового призначення: «Візуальна реєстрація державних номерних знаків автотранспортних засобів при багатопотоковому русі», «Система ідентифікації моделей корпусів холодильників за зображеннями», «Алгоритми обробки та сегментації ландшафтних зображень . Identification of objects. Algorithmic and software security was transferred to interested organizations. The results of the test operation showed the practicality of the software, developed on the basis of proponing in dissertation robot models and methods.

In this rank, the following results were taken from the dissertation robot:

1. A formal model for the processing and recognition of space-hour structures based on an adaptive hierarchical procedure was suggested. processing sequences of images, which are considered by them, that in them isomorphic and homomorphic transformations are enshrined and the functions of static and dynamic invariants are enshrined. Also, the model was motivated to search for static and dynamic signs of objects for some tasks to analyze sequences of images in the presence of a video sensor that is collapsing, and objects that are collapsing in the scene.

2. Розширені- основні положення дескриптивного підходу до розпізнавання послідовностей зображень, що дозволяють враховувати цілі розпізнавання на початкових стадіях обробки послідовності зображень з подальшою сегментацією областей інтересу, будувати траєкторії руху і розпізнавати поведінку динамічних об'єктів, враховувати передісторію руху об'єктів об'єктів caution.

3. The hierarchical method of processing and recognizing the space-time structures, which is composed of three equal and five stages, and transferring the normalization of object projections, which allows speeding up the number of references for one class with different dynamic characteristics of folding, has been developed.

4. The method of estimating the speed for image sequences from the visible and infrared ranges of electromagnetic vibrational reproduction has been developed, which is examined by the fact that space-hour sets of data are victorious, presented in the form of structural tensors and ЪB tensors. the flow is clear. Otriman's assessment of movement allows you to choose the most efficient method of segmentation of dynamic visual objects, which considers the number of acceptable projections.

5. A model of the variegated movement of the regions of the image based on local vectors of speed was suggested, which allows us to split the scene not only on the objects of the foreground and the background, but also on the level of the movement of the objects, in the distance in the poster. This is especially true for folding scenes, which are recorded by a handy video sensor, if all the objects of the scene are located in a visual Russian.

6. An adaptive algorithm for segmenting dynamic objects has been developed: a) for objects with a limited number of projections, based on the analysis of the history of the dynamics of local dynamic regions, which is inspired by the fact that in case of overlaps, the image will be processed after the template, to get the correct shape stagnation of the Kalman filter is predicted, in-line, trajectory; b) for objects with a sufficient number of projections based on complex analysis, color, textural, statistical, topological signs and a sign of movement, which is suspected that when the image overlaps, the shape of the region is obtained using the active contour method.

7. A method of creating a dynamic videograph of a folding scene was proposed for the method of ієrarchical grouping of complex signs of the lower level of the local expanse of the structure, the hours, and the distance of the local expanse of the object. The formation of the videographer sets the timing of the hours between the objects and takes all the important signs for recognizing the signs near the stage. Expanded the two-world grammar of M.I. Schlesinger at the borders of the structural method of recognition to the triple contextual grammar.

8: For the recognition of dynamic objects of modifications, the collective method adopts a solution, first of all, the recognition of the reliability of the image of the area of ​​competence, and then chooses the virishal rule, the competence of which is maximum in the given area. It was motivated to see pseudo-visions for the recognition of the world of the similarity of the input dynamic images from the fallow images of the manifestation of dynamic signs.

9. The method of recognizing the pods on the basis of the Bayesian measure, which is recursively splitting the matrix of the main coefficients from the input video sequence and the poring from the reference pods, is broken down at the stage of learning. Tsya іnformatsija є vihіdnoy for vyznachennya scene genre and іndexuvannya video sequences in multimedia Internet databases.

10. Practical task of processing and recognizing the sequences of the images as a supplement to the adaptive-archarchical method of space-hour processing, the practicality of the method is shown, the effectiveness of the system of hierarchical processing methods is demonstrated. recognition of visual information with the possibility of adaptive selection of a sign. processes of rozvyazannya tasks. Take away the results from looking at the design of experimental systems and transfer them to interested organizations.

In this way, this dissertation robot has violated the important scientific and technical problem of information security of video warning systems and the development of a new one directly at the gallery of the space-hour processing and recognition of dynamic images.

1. Automatic analysis of folding images / Ed. EAT. Braverman. M.: Svіt, 1969. - 309 p. Bongard M.M. Problems of recognition. - M: Nauka, 1967.-320 p.

2. Alpatov, B.A., The manifestation of the object that is collapsing, in the sequence of the image for the presence of the border on the area and the movement of the object / B.A. Alpatov, A.A. China// Digital processing of the image, No. 1, 2007. p. 11-16.

3. Alpatov, B.A., Vision of objects that collapse in the minds of geometric images / B.A. Alpatov, P.V. Babayan // Digital processing of signals, No. 45 2004. p. 9-14.

4. Alpatov, B.A., Babayan P.V. Alpatov B.A., Babayan P.V. Digital signal processing, No. 2, 2006. 45-51 p.

5. Bolshakov, A.A., Methods for processing rich data and time series: Manual for universities / A.A. Bolshakov, R.I. Karimov/M: Garyacha liniya-Telecom, 2007. 522 p.6: Bongard, M.M. Problems of recognition / M.M. Bongard/M.: Nauka, 1967.-320 p.

6. Bulinsky, A.V. Theory of vipadical processes1/A.B. Bulinsky, O.M. Shiryaev/M.: FIZMATLIT, 2005. 408 p.

7. Weinzweig, M.M. The architecture of the representation system of dynamic scenes in terms of understanding / M.N. Vaintsvaig, M.M. Polyakova // Zb. tr. 11th All-Russian conf. "Mathematical Methods for Pattern Recognition (MMRO-11)", M., 2003. pp. 261-263.

8. Vapnik, V.M. The head of the apprenticeship of recognition of images / V.M. Vapnik / M .: Knowledge, 1970. - 384 p.

9. P.Vapnik, V.M. Theory of pattern recognition (statistical problems of learning) / V.M. Vapnik, A.Ya. Chervonenkis/M.: Nauka, 1974. 416 p.

10. Vasiliev, V.I. Recognition of rukhomih tіl / V.I. Vasiliev, A.G. Ivakhnenko, V.Є. Reutsky and in. // Automation, 1967 No. 6, p. 47-52.

11. Vasiliev, V.I. Recognition systems / V.I. Vasiliev / Kiev: Nauk. Dumka, 1969. 292 p.

12. Vasiliev, V.I. Recognition systems. Dovidnik/V.I. Vasiliev / Kiev, Nauk, Dumka, 1983. 422 p.

13. Vizilter, Yu.V. Implementation of the method of analysis of morphological data in machine learning problems>/Yu.V. Vizilter // Bulletin of computer and information technologies, No. 9, 2007, p. 11-18.

14. Vizilter, Yu.V. Projective morphology with improved interpolation / Yu.V. Vіzіlter // Bulletin of computer and information technologies, №4, 2008.-p. 11-18.

15. Vizilter, Yu.V., Projective morphology and its development in the structural analysis of digital images / Yu.V. Vizilter, S.Yu. Zhovtiv // Izv. RAN. TISU, No. 6, 2008. p. 113-128.

16. Vizilter, Yu.V. Research on the behavior of autoregression filters in the task of seeing and analyzing traffic on digital video sequences / Yu.V. Vizilter, B.V. Vishnyakov // Bulletin of computer and information technologies, No. 8, 2008. - p. 2-8.

17. Vizilter, Yu.V. Projective morphology of images based on models that are described by structured functionals /Yu.V. Vizilter, S.Yu. Zhovtov // Bulletin of computer and information technologies, No. 11, 2009.-p. 12-21.

18. Vishnyakov, B.V. The use of the modified method of optical flows in the problem of revealing that inter-frame quilting movement.

19. Ganebnikh, S.M. Analysis of scenes on the basis of stagnation of wood-like images / S.N. Ganebnikh, M.M. Lange // Zb. tr. 11th all-grown. conf. "Mathematical Methods for Pattern Recognition (MMRO-11)", M., 2003.-p. 271-275.

20. Glushkov, V.M. Introduction to cybernetics / V.M. Glushkov/Kiev: type of AN URSR, 1964. 324 p.

21. Gonzalez, R., Woods R. Digital image processing. Translation from English. for red. P. A. Chochia / R. Gonzalez, R. Woods / M.: Technosfera, 2006. 1072 p.

22. Goroshkin, A.N., Image segmentation of handwritten text (SegPic) / O.M. Goroshkin, M.M. Favorska // Certificate No. 2008614243. Registered with the Registry of Programs for EOM, Moscow, September 5, 2008.

23. Grenander, W. Lectures on the theory of images / W. Grenander / In 3 volumes / Translated from English. For red. Yu.I. Zhuravlova. M: Mir, 1979-1983. 130 s.

24. Gruzman, I.S. Digital processing of images in information systems: Navch. Posibnik / I.S. Gruzman, B.C. Kirichuk, V.P. Kosikh, G.I. Peretyagin, A.A. Spektor/Novosibirsk, type NDTU, 2003. p. 352.

25. Reliable and Plausible Visnovok in Intellectual Systems, Ed. V.M. Vagina, D.A. Pospelova. 2nd view., corrected. that dod. - M.: FIZMATLIT, 2008. - 712 p.

26. Duda, R. Image recognition and analysis of scenes / R. Duda, P. Hart/M.: Svit publishing house, 1978. 512 p.

27. Zhuravl'ov, Yu.I. About the algebraic step to complete the task of recognizing and classifying / Yu.I. Zhuravlyov // Problems of Cybernetics: Zb. st., vip. 33, M: Nauka, 1978. p. 5-68.

28. Zhuravl'ov, Yu.I. About Algebraic Correction of Processing Procedures (Reworking) of Information / Yu.I. Zhuravlyov, K.V. Rudakov // Problems of applied mathematics and informatics, M.: Nauka, 1987. p. 187-198.

29. Zhuravl'ov, Yu.I. Recognition of images and recognition of images / Yu.I. Zhuravlov, I.B. Gurevich// Shhorichnik “Recognition. Classification. Forecast. Mathematical methods and їх zastosuvannya”, VIP. 2, M: Nauka, 1989.-72 p.

30. Zhuravlov, Yu.I. Recognition of images and analysis of images / Yu.I. Zhuravlyov, I.B. Gurevich / Piece intelligence in 3 books. Book. 2. Models and methods: Dovіdnik / Ed. YES. Pospelova, M.: type-in ​​"Radio and call", 1990. - p.149-190.

31. Zagoruiko, N.G. Methods of recognition and identification / N.G. Za-goruiko/M.: Glad. radio, 1972. 206 p.

32. Zagoruiko, N.G. Piece intelligence and empirical transfer / N.G. Zagoruiko / Novosibirsk: view. NSU, ​​1975. 82 p.

33. Ivakhnenko A.G. About the development of the theory of invariance and combined control before the synthesis and analysis of primary systems / O.G. Ivakhnenko // Automation, 1961 No. 5, p. 11-19.

34. Ivakhnenko, G.I. Self-learning system of recognition and automatic control / A.G. Ivakhnenko/Kiev: Tekhnika, 1969. 302 p.

35. Kashkin, V.B. Remote sensing of the Earth from space. Digital processing of the image: Heading help / V.B. Kashkin, A.I. Su-hinin/M.: Logos, 2001. 264 p.

36. Kobzar, A.I. Applied mathematical statistics. For engineers and scientists / O.I. Kobzar / M: FIZMATLIT, 2006. 816 p.

37. Kovalevsky, V.A. Correlation method of image recognition / V.A. Kovalevsky // Zhurn. calculated Mathematics and Mathematical Physics, 1962, 2, no. 4, p. 684-690.

38. Kolmogorov, A.N.: Epsilon-entropy and epsilon-multiplier in functional spaces / O.M. Kolmogorov, V.M. Tikhomirov // Theory of Information and Theory of Algorithms. M: Nauka, 1987. p. 119-198.

39. Korn, G. Dovіdnik z mathematici dlya naukovtsіv i іnzhenerіv / G. Korn, T. Korn // M.: Nauka, Gol. ed. fiz.-mat. lit., 1984. 832 p.

40. Kronover, R. Fractals and chaos in dynamical systems / R. Kronover // M.: Tekhnosfera, 2006. 488 p.

41. Lapko, A.V. Non-parametric and hybrid systems of classification of various types of data / A.V. Lapko, BlA. Lapko // Tr. 12th All-Russian conf. "Mathematical Methods and Models of Image Recognition" (MMRO-12), M., 2005.-p. 159-162.

42. Levtin, K.E. Visual detection of Dima (SmokeDetection) / K.E. Levtin, M.M. Favorska // Certificate No. 2009612795. Registered in the Register of programs for EOM Moscow, ZO Lipnya 2009

43. Lutsiv, V.R. Principles of unification of optical systems of robots / V.R. Lutsiv, M.M. Favorska // V-book. "Unification and standardization of industrial robots", Tashkent, 1984. p. 93-94.

44. Lutsiv, V.R. Universal optical system for HAP/V.R. Lutsiv, M.M. Favorska // At the book. “Dosvіd svorhennya, vprovadzhennya ta vykoristannya APCS in associations and enterprises”, L., LDNTP, 1984. p. 44-47.

45. Medvedeva, E.V. Method for evaluating vectors in motion in video images / E.V. Medvedeva, B.O. Timofiev // At the materials of the 12th international conference and exhibition "Digital processing of signals and її zastosuvannya", M .: U 2 vol. T. 2, 2010. p. 158-161.

46. ​​Methods of computer processing of images / Ed. V.A. Soifer. 2nd view., Isp. - M.: FIZMATLIT, 2003. - 784 p.

47. Methods of automatic detection and support of objects. Image processing and management / B.A. Alpatov, P.V. Babayan, O.E. Balashov, A.I. Stepashkin. -M: Radiotehnika, 2008. - 176 p.

48. Methods of computer optics / Ed. V.A. Soifer. M.: FIZMATLIT, 2003. - 688 p.

49. Mudrov, A.Є. Numerical methods for PEOM with the language Basic, Fortran and Pascal / А.Є. Mudrov/Tomsk: MP "RASKO", 1991. 272 ​​p.

50. Pakhirka, A.I. Individual localization (FaceDetection) / A.I. Pakhirka, M.M. Favorska // Certificate No. 2009611010. Registered with the Program Registry for EOM Moscow, February 16, 2009.

51. Pakhirka, A.I. Nonlinear image enhancement / A.I. Pakhirka, M.M. Favorska Certificate No. 2010610658. Registered in the Register of programs for EOM Moscow, March 31, 2010.

52. Pontryagin, L. S. Uninterrupted groups J L. S. Pontryagin // 4th kind., M.: Nauka, 1984.-520 p.

53. Potapov, A.A. Fractals in radiophysics and radar: Vibration topology / A.A. Potapov // View. 2nd, revised. that dod. - M: University book, 2005. 848 p.

54. Radchenko Yu.S. Follow-up to the spectral algorithm for the appearance of "changes in video sequences" / Yu.S. Radchenko, A.V. Buligin, T.A. Radchenko // Izv. VNZ.

55. Salnikov, I.I. Raster space-clock signals in image analysis systems / I.I. Salnikov // M.: Fizmatlit, 2009. -248 p.

56. Sergunin, S.Yu. Scheme of dynamic and bagatory image description / S.Yu.Sergunin, K.M.Kvashnin, M.I. Kumskov // Zb. tr. 11th All-Russian Conf: "Mathematical Methods for Pattern Recognition (MMRO-11)", M., 2003. p. 436-439:

57. Slinko Yu.V. Verification of the task of one-hour follow-up and contouring by the method of maximum likelihood / Yu.V. Slinko // Digital Signal Processing, No. 4, 2008. p. 7-10

58. Solso, R. Cognitive psychology / R. Solso / St. Petersburg: Peter, 6th kind., 2006. 590 p.

59. Tarasov, I.Є. Development of digital devices based on FPGA "Xi-linx" from downloads of VHDL movies / I.Є. Tarasov/M.: Garyacha liniya-Telecom, 2005. - 252 p.

60. Favorska, M.M. Development of algorithms for digital recognition of images in adaptive robotic systems / M.N. Favorska// L!, Leningrad Institute of Aviation. prilad., 1985. Manuscript deposit: VINITI 23.01.85. No. 659-85 Dep.

61. Favorska; MM. Application of spectral methods for image normalization and recognition in adaptive robotic systems / M. N. *. Favorska // L., Leningradsky, in-t aviation. prilad., 1985. Manuscript dep. at VINITI 23.01.85. No. 660-85 Dep.

62. Favorska, M.M. Dosvid of the development of algorithms for recognition of objects for stamping production / M.M. Favorska // At the book. "Stan, dosvid that directly works with complex automation on the basis of traffic police, RTK and PR", Penza, 1985. p. 64-66.

63. Favorska, M.M. Doslіdzhennya projective powers of groups of objects / М.М. Favorska, Yu.B. Kozlov // Bulletin of the Siberian State Aerospace University. Vip. 3, Krasnoyarsk, 2002. - p. 99-105.

64. Favorska, M.M. Designation of the Athenian structure of the object behind the floor / M.M. Favorska // Bulletin of the Siberian State Aerospace University, Vip. 6, Krasnoyarsk, 2005. - p. 86-89.

65. Favorska-M.M. Zagalna classification p_dkhod_v to rezp_znavannya izobrazhen / M-.N. Favorska // V< материалах X междунар. научн. конф. «Решетневские чтения» СибГАУ, Красноярск, 2006. с. 54-55.

66. Favorska M.M. Invariant virtual functions for the recognition of static images / M.N. Favorska // Bulletin of the Siberian State Aerospace University. Vip. 1 (14), Krasnoyarsk, 2007. p. 65-70.

67. Favorska, M.M. Imovirnіsnі methods and segmentation of the video stream as a task of daily data / М.М. Favorska // Bulletin of the Siberian State Aerospace University. Vip. 3 (16), Krasnoyarsk, 2007. p. 4-8.

68. Favorska, M.M. Selecting the number of informative signs in image recognition systems / M.M. Favorska // In the materials of the XI international. Sciences. conf. "Reshetnivsky reading" SibDAU, Krasnoyarsk, 2007 p. 306-307.

69. Favorska, M.M. Strategies for segmentation of two-dimensional images / M.M. Favorska // At the materials of the All-Russian scientific conference “Models and methods for processing images MMOI-2007”, Krasnoyarsk, 2007. p. 136-140.

70. Favorska, M.M. Segmentation of landscape images based on the fractal approach / M.M. Favorska // At the materials of the 10th international conference and exhibition "Digital processing of signals and її zastosuvannya", M., 2008. p. 498-501.

71. Favorska, M.M. The model for recognizing the image of the handwritten text/M.M. Favorska, O.M. Goroshkin // Bulletin of the Siberian State, Military Aerospace University. Vip. 2 "(19), Krasnoyarsk, 2008. pp. 52-58.

72. Favorska, M.M. Algorithms for the implementation of traffic assessment on video warning systems / M.N. Favorska, A.S. Shilov // Management systems and information technologies. Perspective research / IPU RAS; VDTU, No. 3.3 (33), M.-Voronizh, 2008. p. 408^12.

73. Favorska, M.M. Prior to the development of formal grammars in recognizing objects in folding scenes // M.N. Favorska / Proceedings of the XIII International Science Conf. "Reshetnevsky reading". About 2 year. 4.2, Krasnoyarsk, 2009. p. 540-541.

74. Favorska, M.M. Recognition of dynamic images based on prophetic filters / M.M. Favorska // Bulletin of the Siberian State Aerospace University. Vip. 1(22) about 2 years. 4f. 1, Krasnoyarsk, 20091 p. 64-68.

75. Favorska, M.M. Favorska, A.I. Pakhirka, A.C. Shiliv; M.V. Zhіnok // Bulletin. Siberian State Aerospace University. Vip. 1 (22) about 2 years. Part 2, Krasnoyarsk, 2009. p. 69-74.

76. Favorska, M.M. Knowledge of Rough Video Objects, from Implanted-Local 3D Structural Tensors / М.М. Favorska // Bulletin of the Siberian State Aerospace University. Vip. 2 (23), Krasnoyarsk, 2009. p. 141-146.

77. Favorska, M.M. Evaluation of object movement in folding scenes based on the tensor approach / M.M. Favorska // Digital processing of signals, No. 1, 2010.-p. 2-9.

78. Favorska, M.M. Comprehensive analysis of the characteristics of landscape images / M.M. Favorska, N.Yu. Piven // Optical journal, 77, 8, 2010.-p. 54-60.

79. Fine, B.C. Image identification/BC. Fine // M: Nauka, 1970.-284 s.

80. Forsythe, D.A. Computer sir. Modern pidkhid / D.A. Forsyth, J. Pons // M: vidavnichiy dіm "William", 2004. 928 p.

81. Fu, K. The latest methods in the recognition of images and the learning of machines / K. Fu / M .: Nauka, 1971. 320 p.

82. Fu, K. Structural Methods in Image Recognition/K. Fu/M.: Svіt, 1977.-320 p.

83. Fukunaga, K. Introduction to the statistical theory of pattern recognition / K. Fukunaga/M.: Nauka, 1979. 368 p.

84. Shelukhin, O.I. Self-similarity and fractals. Telecommunication programs / O.I. Shelukhin, A.V. Osin, S.M. Smolsky / For red. O.I. Shelukhina. M: FIZMATLIT, 2008. 368 p.

85. Shilov, A.S. Movement Estimation / O.S. Shilov, M.M. Favorska // Certificate No. 2009611014. Registered with the Program Registry for EOM Moscow, February 16, 2009.

86. Sh.Shlezinger, M.I. Correlation method for recognizing image sequences / M.I. Schlesinger / At the book: automatic, what to read. Kiev: Nauk.dumka, 1965. p. 62-70.

87. Shlesinger, M.I. Syntactic analysis of two-world sound signals for the minds of a shift / M.I. Schlesinger // Cybernetics, No. 4, 1976. - pp. 76-82.

88. Stark, G.-G. Zastosuvannya wavelets for TsGZ / G.-G. Shtark/Ml: Technosfera, 2007. 192 p.

89. Shup, T. Applied numerical methods in physics and technology: Per. from English / T. Shup / Ed. S.P. Merkur'eva; M: Vishcha. School, 19901 - 255 p.11 "5. Electric, resource: http:// www.cse.ohio-state.edu/otcbvs-bench

90. Electric, resource: http://www.textures.forrest.cz/ electronic resource (database of texture images textures library forrest).

91. Electric, resource: http://www.ux.uis.no/~tranden/brodatz.html Electronic resource (Brodatz texture image database).

92. Allili M.S., Ziou D. Active contours for video object tracking using region, boundary and shape information // SIViP, Vol. 1, no. 2, 2007.pp. 101-117.

93. Almeida J., Minetto R., Almeida T. A., Da S. Torres R., Leite N. J. Robust estimation of camera motion, optical flow models // Lecture Notes in

94. Computer Science (including subseries Libraries and Literary Sciences) 5875 LNCS (PART 1), 2009. pp. 435-446.

95. Ballan L., Bertini M., Bimbo A. D., Serra G. Video Event Classification using String Kernels // Multimed. Tools Appl., Vol. 48, no. 1, 2009.pp. 6987.

96. Ballan L. Bertini M. Del Bimbo A., Serra G. Action categorization in soccer videos using string kernels // In: Proc. IEEE Int'l Workshop on Content-Based Multimedia Indexing (CBMI). Chania, Crete, 2009. pp. 13-18.

97. Barnard K., Fan Q. F., Swaminathan R., Hoogs A., Collins R, Rondot P., and Kaufhold J. Evaluation of localized semantics: Data, methodology, and experiments // International Journal of Computer Vision, IJCV 2008, Vol. 77, no. 1-3,2008.-pp. 199-217.

98. Bertini M., Del Bimbo A., Serra G. Development of rules for semantic video event annotation // Lecture Notes In Computer Science; In: Proc. Int'l Conference on Visual Information Systems (VISUAL), Vol. 5188, 2008.pp. 192-203.

99. Bobick A.F., Davis J.W. Recognition of human-movement using temporal templates // IEEE Transactions on Pattern Analysis and Machine Intelligence, Vol. 23, no. 3, 2001.pp. 257-267.

100. Boiman O., Irani M. Designation of irregularities in images and in video // International Journal of Computer Vision, Vol. 74, no. 1, 2007.pp. 17-31.

101. Bresson X., Vandergheynst P., Thiran J.-P. A Variational Model for Object Segmentation For Help Boundary Information and Shape Prior Driven4 by Mumford-Shah Functional // International Journal of Computer Vision, vol. 68, no. 2, 2006.-pp. 145-162.

102. Cavallaro A., Salvador E., Ebrahimi T. Shadow-aware object-based video processing // IEEE Vision; Image and Signal Processing, Vol. 152, no. 4, 2005.-pp. 14-22.

103. Chen J., Ye J. Training SVM with indefinite kernels // In: Proc. of the 25th international conference on Machine learning (ICML), Vol. 307, 2008.pp. 136-143.

104. Cheung S.-M., Moon Y.-S. Detection of Approaching Pedestrians as Distance Using Temporal Intensity Patterns // MVA2009, Vol. 10, no. 5, 2009.-pp. 354-357.

105. Dalai N., Triggs B., and Schmid G. Human detection using oriented histograms of flow and appearance // In ECCV, vol. II, 2006. pp. 428^141.

106. Dalai N., Triggs B. Histograms of Oriented Gradients for Human Detection // IEEE Conference on Computer Vision and Pattern Recognition (CVPR), vol. II, 2005-pp. 886-893.

107. Dani A.P., Dixon W.E. Single camera structure and motion estimation // Lecture Notes in Control and Information Sciences, 401, 2010. pp. 209-229.

108. Datta Ri, Joshi D;, Li J., and Wang J. Z1 Image retrieval: Ideas, influences, and trends of new age // ACM"-Computing Surveys, Vol. 40:, no: 2, 2008. ■ -pp 1-60.

109. Dikbas S., Arici T., Altunbasak Y. Fast motion estimation with interpolation-free sub-sample accurracy // IEEE Transactions on Circuits and Systems for Video Technology 20(7), 2010. -pp. 1047-1051.

110. Dollar P., Rabaud V., Cottrell G., Belongie S. Behavior recognition via sparse spatio-temporal features // In: Proc. 2nd Joint IEEE International Workshop on Evaluation of Tracking and Surveillance, VS-PETS, 2005. pp. 65-72.

111. Donatini P. and Frosini P. Natural pseudodistances між closed surfaces // Journal of European Mathematical Society, Vol. 9, no. 2, 2007pp. 231-253.

112. P. Donatini and P. Frosini, “Natural pseudodistances and closed curves,” Forum Mathematicum, Vol. 21, no. 6, 2009.pp. 981-999.

113. Ebadollahi S., L., X., Chang S.F., Smith J.R. Visual event detection using multi-dimensional concept dynamics // In: Proc. IEEE Int'l Conference on Multimedia and Expo (ICME), 2006. pp. 239-248.

114. Favorskaya M., Zotin A., Danilin I., Smolentcheva S. Realistic 3D modeling of nature with natural effect // Proposals for another KES International Symposium IDT 2010, Baltimore. USA. Springer-Verlag, Berlin, Heidelberg. 2010.-pp. 191-199.

115. Francois A.R.J., Nevatia R., Hobbs J.R., Bolles R.C. VERL: An ontology framework for representing and annotating video events // IEEE Multimedia, Vol: 12; no. 4, 2005.pp. 76-86.

116. Gao J., Kosaka A:, Kak A.C. Multi-Kalman Filtering Approach for Video Tracking of Human-Delineated Objects in Cluttered" Environments // IEEE Computer Vision and Image Understanding, 2005, V. 1, no. 1. pp. 1-57.

117 Gui L., Thiran J.-P., Paragios N. Joint Object Segmentation and Behavior Classification in Image Sequences // IEEE Conf. on Computer Vision and Pattern Recognition, 17-22 June 2007. pp. 1-8.

118. Haasdonk B. Feature space interpretation of SVMs with indefinite kernels // IEEE Transactions on Pattern Analysis and Machine Intelligence. Vol. 27, no. 4, 2005.pp. 482-492.

119. Harris C. and Stephens M. Combined corner and edge detector // In Fourth Alvey Vision Conference, Manchester, UK, 1988. pp. 147-151.

120. Haubold A., Naphade M. Classification of video events using 4-dimensional-time-compressed motion features // In CIVR "07: Proceedings of the6th ACM International Conference on Image and Video retrieval, NY, USA, 2007. -pp 178-185.

121. Haykin S. Neural Networks: A Comprehensive Introduction. / N.Y.: Prentice-Hall, 1999; .- 658 pi.

122. Hoynck M., Unger M., Wellhausen J. and Ohm J.-R. A Robust Approach to Global Motion Estimation for Content-based Video Analysis // Proceedings of SPIE Vol. 5601, Bellingham, WA, 2004. pp. 36-45.

123. Huang Q., Zhao D., Ma S., Gao W., Sun H. Deinterlacing for help with hierarchical motion analysis // IEEE Transactions on Circuits and Systems for Video Technology 20(5), 2010. pp. 673-686.

124. Jackins C.L., Tanimoto S.L. Quad-trees, Oct-trees and K-trees: Generalized Approach to Recursive Decomposition of Euclidean Space // IEEE Transactions onPAMI, Vol. 5, no. 5, 1983.-pp. 533-539.

125. Ke Y., Sukthankar R:, Hebert Mi. Efficient visual event detection using volumetric features // In: Proc. of Int'l Conference on Computer Vision (ICCV), vol.1, 2005.-pp. 166-173.

126. Klaser A., ​​Marszalek M., and Schmid C.A Spatio-Temporal Descriptor Based on 3D-Gradients // In BMVC, British Machine Vision, Conference, 2008. -pp. 995-1004.

127. Kovashka, A., Grauman, Before Learning a hierarchy of discriminative space-time neighborhood features for human action recognition // Proceedings of IEEE Computer Society Conference on Computer Vision and Pattern Recognition, 2010. pp.2046-2053.

128. Kumskov M.I. Calculation Scheme of the Image Analysis Models of the Objects to be Recognized // Pattern Recognition and Image Analysis, Vol. 11, no. 2, 2001.p. 446-449:

129. Kwang-Kyu S. Content-based image retrieval by combining genetic algorithm and support vector machine // In ICANN (2), 2007. pp. 537-545.

130. Lai C.-L., Tsai S.-T., Hung Y.-P. Study on three-dimensional coordinate calibration using fuzzy system // International Symposium on Computer, Communication, Control and Automation 1, 2010. - pp. 358-362.

131. Laptev I. On space-time interest points // International Journal of Computer Vision, Vol. 64, no. 23, 2005.pp. 107-123.

132. Leibe B., Seemann E., Schiele B. Pedestrian Detection in-Crowded* Scenes // IEEE Conference on Computer Vision and "Pattern Recognition, Vol. 1, 2005.-pp. 878-885.

133. Lew M. S., Sebe N., Djeraba C., and Jain R. Content-based multimedia-information1 retrieval: State of the art and challenges // ACM Transactions on Multimedia Computing, Communications, and Applications, Vol. 2, no. 1, 2006.pp. 1-19.

134. Li J. and Wang J. Z. Real-time computerized annotation of pictures // IEEE Trans. PAMI, Vol. 30, 2008.pp. 985-1002.

135. Li L., Luo R., Ma. 9 IEEE Intern. Workshop on PETS, New York, 2006. pp. 91-98.

136. Li L., Socher R., and Fei-Fei L. Towards Total Scene Understanding: Classification, Annotation and Segmentation in Automatic Framework // IEEE Conference on Computer Vision and Pattern Recognition, CVPR, 2009. pp. 2036-2043.

137. Li Q., ​​Wang G., Zhang G.) Chen S. Global motion estimation based on pyramid with mask // Jisuanji Fuzhu Sheji Yu Tuxingxue Xuebao/Journal of Computer-Aided Design and Computer Graphics, Vol.: 21, no. 6, 2009.pp. 758-762.

138. Lindeberg T., Akbarzadeh A. and Laptev I. Galilean-diagonalized spatio-temporal interest operators // Proceedings of the 17th International Conference on Pattern Recognition (ICPR "04), 2004. pp. 1051-1057.

139. Lim J., Barnes, N. Recognizing the epipolis for assisting septic flow at antipodal points // Computer Vision and Image Understanding 114, no. 2, 2010.pp. 245-253.

140. Lowe D. G. Distinctive Image Features from Scale-Invariant Keypoints // International Journal of Computer Vision, Vol. 60, no. 2, 2004.pp. 91-110.

141. Lucas B.D., Kanade T. An Iterative Image Registration Technique with Application to Stereo Vision // International Joint Conference on Artificial Intelligence, 1981. pp. 674-679.

142 Mandelbrot B; B. The Fractal Geometry of Nature / N.Y.: Freeman ^ 1982. 468 p.; Russian, trans.: Mandelbrot B. Fractal, geometry of nature: Per. from English / M.: Institute of Computer Studies, 202. - 658 p.

143. Mandelbrot V.V., Frame M.L. Fractals, Graphics, and Mathematics Education/N. Y.: Springer-Verlag, 2002. 654 p.

144. Mandelbrot B.B. Fractals and Chaos: Mandelbrot Set. and Beyond / N.Y.: Springer-Verlag, 2004. 308 p.

145. Memoli F. For help Gromov-Hausdorff help me for comparison // Proceedings of Eurographics Symposium on Point-Based Graphics. Prague, Czech Republic, 2007. pp. 81-90.

146. Mercer J. Functions of positive and negative typ and their connection with theory of integral equations // Transactions of the London Philosophical Society (A), vol. 209, 1909. pp. 415-446.

147. Mikolajczyk K. Detection of local features invariant to affine transformations, Ph.D.thesis, Institut National Polytechnique de Grenoble, France. 2002.171 p.

148. Mikolajczyk K. and Schmid G. An Affine Invariant Interest Point Detector // Proceedings of ECCV. Vol. 1. 2002. pp. 128-142.

149. Minhas R., Baradarani A., Seifzadeh S., Jonathan Wu, Q.M. Human action recognition using extreme learning machine based on visual vocabularies // Neurocomputing, Vol. 73 (10-12), 2010. pp. 1906-1917.

150. Mladenic D., Skowron A., eds.: ECML. Vol. 4701 of Lecture Notes in Computer Science, Springer, 2007. pp. 164-175.

151. Moshe Y., Hel-Or H. Video block motion estimation based on gray-code kernels // IEEE Transactions on Image Processing 18(10), 2009. pp. 22432254.

152. Nakada T., Kagami S;, Mizoguchi H. Pedestrian Detection using 3D Optical Flow Sequences for-afMobile Robot // IEEE Sensors, 2008. pp: 116-119:

153. Needleman, S. B:,. Wunsch C.D.; As a wild method zastosovuetsya to the point of asking for similarity in the aminoacid sections of two articles // Journal of Molecular Biology Vol. 48, no: 3, 1970. pp. 443-453.

154. Neuhaus M., Bunk H. Edit distance-based kernel functions for structural pattern classification // Pattern Recognition. Vol. 39, no. 10, 2006. pp: 1852-1863.

155. Nevatia R., Hobbs J., and Bolles B. An anthology for video recording // In Workshop on Event Detection and Recognition. IEEE, Vol.12, no. 4, 2004.pp. 76-86.

156. Nguyen.N.-T., Laurendeau D:, Branzan-Albu A. A broad method for camera monitoring in vehicles based on optical flow // The 6th International