Particle Filter used to Recognize Object in Video

Recognition of the objects in video can offers significant benefits to video retrieval including automatic annotation and content based queries based on the object characteristic, detecting particular object in video is an important stop toward semantic understanding of visual imagery. Thus this paper will give the basic information about object recognition and particle filters. It estimates the probability of object presence in current image given the history of observations up to current time to do this the object presence is modeled by two state markov chain and the problem is translated into sequential Bayesian estimation which can be solved by particle filter. The observation density required by the particle filters is based on selected discriminative Haar like feature. This paper is related video segmentation for moving objects followed by the implemented tracking and classification algorithm. 1. Introduction Object detection in images has received considerable attention in the past decades, probably because reliable object detection systems are required as a front-end in numerous applications. Object detection deals with determining if an instance of a given class of objects (for examples cars, faces, etc.) is present or not in an image. Successful object detection systems are based on the learning of object appearance using large collections of exemplars. Object recognition is increasingly concerned with the ability to recognize generic classes of objects rather than just specific instances. It has been dominated by approaches that separate processing into distinct stages of feature extraction and matching. In the first stage, discrete primitives, or features are detected. In the second stage, stored models are matched against those features. For instance, in the pioneering work of Roberts children's blocks were recognized by first extracting edges and corners from images and then matching these features to polyhedral models of the blocks. The model-based recognition paradigm of the 1980's similarly followed this approach. These methods focus largely on the problem of efficiently searching for correspondences between features that have been extracted from an image, and features of a stored model. Target tracking is an important element of surveillance, guidance or obstacles avoidance systems whose role is to determine the number, position and movement of the targets the fundamental building block of a tracking systems is a filter for recursive target applications. Visual object tracking is a difficult problem, but in recent years, particle filter-based object trackers have proven to be very effective. Conceptually, a particle Mr. A. M. Borkar, Mr. N. S. Panchbudhe, Mr.U.S.Ghate, Mr. A. K. Sharma, The International Journal of Computer Science & Applications (TIJCSA), Vol. 1 No.2 April 2012