EFFICIENT MOTION DETECTION TO BACKGROUND SUBTRACTION ? SUPPORT SECURITY

Detection of moving objects in video streams is the first significant step of information extraction in many computer vision applications. Even there is usefulness to segment video streams into moving and background components, moving objects detection concentrates on recognition, classification, and activity analysis by making the latter steps efficient. We propose a technique through artificial neural networks which is applied in image processing of human and more commonly in cognitive science. The technique which is proposed can handle scenes containing moving backgrounds, steady illumination variations, has no bootstrapping restrictions, can include into the background model shadows cast by moving objects, and achieves strong detection for different types of videos taken with stationary cameras. We compare our method with other modelling techniques and report results, in terms of processing speed and accuracy for the color video sequences that represent typical situations critical for video surveillance systems. Facial features are segregated at different resolutions to provide noise information, edge, smoothness, and blurriness present in a face. WLD descriptor represents an image as a histogram of differential excitations and gradient orientations and also it has various interesting properties like robustness to noise and smart detection of edges, illumination changes and powerful image representation in feature extraction stage. The Gabor filter bank is used to extract the features from face regions to discriminate the illumination changes. These collective features are useful to distinguish the maximum number of samples accurately and it is matched with already stored original face samples for identification. The simulated results will be shown that used granulation and hybrid spatial features descriptors has better discriminatory power and recognition accuracy in the process of recognizing different facial appearance.