AN ENHANCED HYBRID CUCKOO SEARCH AND GENETIC ALGORITHM USING HAAR-LIKE FEATURE FOR SURVEILLANCE VIDEOS

Abstract

This paper is mainly focused on image processing concept which is used in surveillance videos for security purpose. Image processing technique involves dispensation of image using a mixture of mathematical functions in which image or video is given as input and the extent to which the image is matched is given as output. The main problem found in this paper are edge detection and noise removal in images. Initially this research paper is compared with cuckoo search and genetic algorithm. Then these two algorithms are combined together as an hybrid CS - GA Algorithm and finally the result of this hybrid algorithm is compared with enhanced hybrid CSGA using Haar-like feature for real time surveillance video. The result is compared in terms of Recognition rate accuracy and time complexity