Name: Arshiya KhanFEATURE EXTRACTIONHuman Detection and TrackingSilhouette RepresentationRECOGNITIONCONCLUSIONSName: Arshiya KhanClass: Cs 265Semester: Spring 2004 Report on Gait RecognitionGait RecognitionAbstract: In today’s world, there is a growing need to authenticate and identify individuals automatically. Securing personal privacy and deterring identity theft are national priorities. Biometrics is a technology that makes use of the physiological or behavioral characteristics of people to authenticate their identities. It can accurately identify or verify individuals based upon their unique physical or behavioral characteristics. Thispaper is a brief report about gait recognition Introduction: Vision –based human identification at a distance is in higher demand in computer vision community. This demand is strongly driven by the need for automated person identification systems for visual surveillance and monitoring applications. The art and science of biometrics is all about coming up with an all-purpose personal identifier. Biometric authentication is the "automatic", "real-time", "non-forensic" subset of the broader field of human identification. Humans recognize each other according to their various characteristics. For example, people recognize others by their face when they meet them and by their voice as they speak to each other. Identity verification in computer systems has traditionally been based on something that one has or one knows. A biometric system is essentially a pattern recognition system that establishes a person’s identity by comparing the binary code of a uniquely specific biological or physical characteristic to the binary code of the stored characteristic. This is accomplished by acquiring a live sample from an individual who is requesting access. The system then applies a complex and specialized algorithm to the live sample and converts it into a binary code and then compares it to the reference sample to determine the individual's access status. A profile or template containing the biometrical properties of a person is stored in the system (generally after data compression), by recording his characteristics. These characteristics are scanned several times during enrolment in order to get a profile that corresponds most with reality. A scan of the biometrics of a person is made and compared with the characteristics that are stored in the profileClassifications:There are various ways to classify biometric systems and devices. Biometric systems: These systems can be used in two different modes: - Verification - Identification.Identity verification occurs when the user claims to be already enrolled in the system, and the biometric data obtained from the user is compared to the data stored in the database. Identification, on the other hand,occurs when the identity of the user is a priori unknown. The user’s data is matched against all the records in the database. Identification is technically more challenging and costly [5]. Biometric devices: These are classified according to two distinct functions: - Positive Identification: To prove a person is enrolled in the system. - Negative Identification: To prove a person is not enrolled in the system. These functions are "duals" of each other. In the first function, the present person is linked with an identity previously registered, or enrolled, in the system. The second function, establishes that the present person is not already present in the system. The purpose of this negative identification system is to prevent the use of multiple identities by a single person. If a negative identification system fails to find a match between thesubmitted sample and all the enrolled templates, an "acceptance" results. A match between the sample and one of the templates results in a "rejection"[5].Examples of Biometrics are Fingerprint, Handwritten signature, Facial recognition, Speech recognition, Gait recognition etc.Gait recognition aims to discriminate individuals by the way they walk. In comparison with other first-generation biometric features such as fingerprint and iris, gait has the advantage of being unobtrusive, i.e., it requires no subject appearance and the dynamics of human walking motion. i.e. recognizing people by gait depends greatly on how the silhouette shape of an individual changes over time in an image sequence. Gait motion is composed of a sequence of static body poses and it is expected that some distinguishable signatures with respect to these static body poses can be extracted and used for recognition. Each person seems to have a distinctive and idiosyncratic way of walking, which can be easily understood from a biomechanics viewpoint. In this paper we aim to discuss about an automatic gait recognition method based upon spatiotemporal silhouette analysis measured during walking Intuitively, recognizing people by gait depends greatly on howthe silhouette shape of an individual changes over time in an image sequence. So, we may consider gait motion to be composed of a sequence of static body poses and expect that some distinguishable signatures with respect to those static body poses can be extracted and used for recognition by considering temporal variations of those observations. Also, eigenspace transformation based on PCA has actually been demonstrated to be a potent metric in face recognition (i.e., eigenface) and gait analysis. Based on these observations, this paper discusses about a proposal made by some people on a silhouette analysis-based gaitrecognition algorithm using the traditional PCA. The algorithm implicitly captures the structural and transitional characteristics of gait [2] Background ModelingMotion SegmentationHuman TrackingSilhouette Extraction2D Silhouette Unwrapping1D Signal NormalizationProjection in EigenspaceRecognitionEigenspace Computation Human Detection &TrackingFeature ExtractionTraining orClassificationThe overview of the proposed algorithm is shown in Figure above. It consists of three major modules, namely, human detection and tracking, feature extraction, and training or classification. The first module serves to detect and track the walking figure in an image sequence. A background subtraction procedure is performed to segment motion from the background, and the moving region corresponding to the spatial silhouette of the walking figure is successively tracked through a simple correspondence method. The second module is used to extract the binary
View Full Document
Unlocking...