Automated Multimedia Information Retrieval Using Color and Texture Feature Techniques

CBVR is a technique that utilizes the visual content of video, to search for similar videos in large-scale video databases, according to a user’s interest. The CBVR problem is motivated by the need to search the exponentially increasing space of image and video databases efficiently and effectively. The visual content of video is analyzed in terms of low-level features extracted from the video. These primarily constitute color, shape and texture features. In this paper, we propose the novel and interactive systems based on visual paradigm in which low level feature plays an important role in video retrieval. We implement and compare five color feature extraction algorithms, three texture feature extraction algorithms. For color feature extraction, the conventional Color Histogram (GCH, LCH), Color Moment, Color Coherence Vector, RGB Averaging method are implemented and compared. For texture feature extraction, Autocorrelation, Grey Level Co occurrence matrix, Edge Frequency are implemented and compared.