Recognition of Human Actions in Video

Abstract

Recognition and analysis of Human actions is an important task in the area of computer vision. There are many applications of this research which include surveillance systems, patient monitoring systems, human performance analysis, con tent - based image/video retrieval/storage, virtual reality and a variety of syst ems that involve interactions between persons or interactions between person and devices, etc. The need for such system is increasing day - by - day, with the increase in number of surveillance cameras deployed in public spaces. Automated systems are required that can detect, categorize and recognize human activities and request the human attention only when necessary. In this paper, important steps of such a system are described that can robustly tracks human in various environments and recognizes their actions through image sequences acquired from a single fixed camera. The overall system consists of major th ree steps: blob extraction, feature extraction, and human action recognition. Given the sequence of images, a statistical method is demonstrated to extract the blobs and to remove the shadows and highlights in order to obtain a more accurate object silhouet te. Shape context is used to extract features in next step and at - last human action is recognized using neural network