The application research of computational intelligence in content video retrieval technology

Along with the unceasing development of computer, communication and multi-media technology and rapid growth of video information, How to effectively depending on the video data to organize, manage, expression and retrieval that becomes the hot issue in video retrieval research field. The paper discusses the feature extraction of video shot boundary detection, researches and analyzes the general method of shot cut detection, then propose the support vector machine based on shot cut detection method. This method makes use of support vector machine Good generalization ability of learning and the characteristics of the lens shear, structural support vector machine classifier, to achieve shot cut detection. It reduced the number of the experience reading and the resulting influence of human factors. Simulation experiments test video clips with different characteristics have achieved good results, through comparison of test results with other methods proved the validity of research methods.