شناسايي انسان از روي نحوه راه رفتن با كمك توصيف گر بافت پويا

چكيده انگليسي :

Gait Recognition using Dynamic Texture Descriptor Behnaz Abdolahi b abdolahi@ec iut ac ir Date of submission Department of Electrical and Computer Engineering Isfahan University of Technology Isfahan IranDegree M Sc Language FarsiSupervisors S Sadri s sadri@cc iut ac irN Gheissari n gheissari@cc iut ac irAbstract The visual analysis of human movements is one of the recent attractive topics in biometricresearch Psychological studies indicate that people have statistically significant ability to recognize peopleby the way they walk Therefore gait recognition has recently become a topic of great interest in computervision Authentication is the process of accepting or rejecting the claimed identity Recent efforts propose away for quick detection of threats without attracting the attention of people in public places like airports banks and subway stations Compared to other biometrics gait has some unique characteristics Commonbiometrics is usually limited and time consuming but gait analysis is unobtrusive The other majoradvantages of gait are that it requires no contact or cooperation it can be measured at a distance and isdifficult to conceal or replicate It has gained great interest because of its many applications such as humanmovement analysis surveillance video indexing sport video analysis and gait recognition Typicalapproaches for human gait recognition have used either motion or shape information However it may notbe a good idea to rely on a single modality Therefore recent researches have gained very goodperformance by using spatiotemporal analysis that combines both motion and shape information Thisthesis is intended to present a bag of video words method for the analysis of human walking based ondynamic textures Dynamic texture descriptors naturally encode motion information Applying them ontextured regions encodes appearance information as well Therefore dynamic texture descriptors can beused to describe human motion in both spatial and temporal domains In this thesis the Local BinaryPatterns from Three Orthogonal Planes LBP TOP as a dynamic texture descriptor is applied to localfeatures for describing human movements in a spatiotemporal way This dynamic texture descriptor isrobust to rotation and scale changes Since the main prerequisite to gain the best description is extracting asmuch discriminative features as possible we use local representation for feature extraction Localrepresentation is invariant to changes in viewpoint person appearance and partial occlusions Thisrepresentation describes the observation as a collection of local descriptors or patches Patches are sampledat space time interest points that have mutation in spatial and temporal domains Then each patch isdescribed by using LBP TOP descriptor Since human walking has statistical variations in both spatial andtemporal spaces Therefore a walking video sequence can be represented as a collection of video wordsafter extracting spatiotemporal interest points and describing them by a dynamic texture descriptor Thenhierarchical K means algorithm as a clustering algorithm is applied to obtain the initial visual dictionary ofvideo words Afterwards by using this visual dictionary each learning sequence is defined by a featurevector At last in pattern matching step we use a classifier to compare feature vectors extracted from testsequence with training models The performance of our method is studied on two public datasets KTH andIXMAS multiview datasets We are the first to test gait recognition method on these datasets and admirableresults are achieved on both datasets Key words human motion analysis gait recognition interest point dynamic texture local binary patternfrom three orthogonal planes clustering

عبدالهي، بهناز

شناسايي انسان از روي نحوه راه رفتن با كمك توصيف گر بافت پويا

تشخيص هويت , رديابي , آناليز حركت انسان , بينايي ماشين , نمايش تصاوير , الگوي دودويي محلي يا سه صفحه متعامد , خوشه بندي