Assessment of Combined Shape, Color and Textural Features for Video Duplication
Gottumukkala HimaBindu, Chinta Anuradha, P.T. Satyanarayana Murty · Traitement du signal · 2019
The aim of the work is to identify the duplication in a video database with the aid of feature extraction techniques.The process includes extraction of image features (shape, color, and texture) for duplicate identification.The color contains 256 features, shape contains 200 features, the texture contains two different features namely gray-level co-occurrence matrix (GLCM) (22 features in 4 degrees) and grey-level run length matrix (GLRLM) (11 features) are extracted.In this paper, the preliminary work is to convert video into frames and then each frame into blocks subsequently including feature extraction.A query video is then considered for the same process of feature extraction and compared with the normal video.The distance between query video and normal video if found to be similar then the video identified as duplicate video.The results are performed for various evaluation matrix and plotted graphs are shown.The sensitivity value for whole feature extractions is 0.88, the specificity value for whole feature extractions is 0.83 and the accuracy value for whole extractions is 0.86.The entire process implemented in the working platform of MATLAB.