Local Binary Pattern Based Shot Boundary Detection for Video Summarization

Video Summarization plays a vital role in the internet user's life, especially for those searching for user specified video of interest for a long time. In order to provide support for users in terms of searching and retrieving video content, it is necessary to segment the video into shots and extract representative frame of each shot which acts as a summary of the video. So, in this paper, an approach for shot boundary detection and keyframe extraction is proposed. To provide a better summarization of video, Local Binary Pattern (LBP) based feature extraction for detecting video shots is performed in order to identify objects motion information. Based on local maxima analysis, an efficient keyframe, which contains salient information of a shot is proposed. The performance of the proposed work is evaluated using evaluation metrics like Precision, Recall, and F1-Measure.