An Intensified System for Identifying Distinct Clusters over the Video Using Improved Semi-Supervised Clustering Technique

Abstract

The improved video clustering technique called Automatic Video Frame Separation System. The proposed (AVFS) system aims to automatically separate the video frames into finite number of distinct cluster with higher similarity using improved semi-supervised clustering technique. The proposed (AVFS) system contains five stages, in the first stage, the automatic video separation system gets the different input videos with different size from different data store. Next stage, the (AVFS) system converts each and every individual video into sequence of frames using standard video to frame converter tool. The third stage has a proposed (AVFS) system pre-processing the video frames to improve the frame quality and optimize the size of the frames using improved arithmetical operation. Similarly, in the fourth stage the proposed system extract the predetermined distinct number of different feature from each and every individual frame over the individual video using standard statistical operation. Following the previous stage; the fifth stage has the proposed system following the previous stage result and split the feature set of input video frames into distinct number of highly relative clusters using optimum N means technique (ONM). The proposed Automatic Video Frame Separation System is well suitable to robotically separate the video frames into distinct clusters with high intra resemblance and higher inter diversity among the clusters.