Histogram difference with Fuzzy rule base modeling for gradual shot boundary detection in video cloud applications

In the field of shot boundary detection the fundamental step is video content analysis towards video indexing, summarization and retrieval as to be carried out for video cloud based applications. However, there are several beneficial in the previous work; reliable detection of video shot is still a challenging issue. In this paper the focus is carried out on the problem of gradual transition detection from video. The proposed approach is fuzzy-rule based system with gradual identification and a set of fuzzy rules are evaluated with dissolve and wipes (fad-in and fad-out) during gradual transition. First, extracting the features from the video frames then applying the fuzzy rules in to the frames for identifying the gradual transitions. The main advantage of the proposed method is its level of accuracy in the gradual detection getting increased. Also, the existing gradual detection algorithms are mainly based on the threshold component, but the proposed method is rule based. The proposed method is evaluated on variety of video sequences from different genres and compared with existing techniques from the literature. Experimental results proved for its effectiveness on calculating performance in terms of the precision and recall rates.