Research on Advanced Performance Evaluation of Video Digital Contents

Abstract

The purposes of this study are to present performance evaluation improvements of filtering technology needed for technological protection measures as a means of copyright protection technology of a variety of contents distribution environment and to draw test experiment evaluation results targeting P2P· Web-hard· Torrent service etc. Though the filtering techniques different by contents genre such as a sound source stabilized image· video· text· games· SW etc. have been proposed, this study attempts to propose improved performance evaluation measurement items for performance evaluation of existing video filtering technology, limiting to the field of video. This study attempts to define performance evaluation items of existing filtering such as tenacity and reliability, time taken for feature point extraction and recognition, consistency on providing same results, minimum unit needed for recognition, filtering on random part, accuracy for time information and time precision, media type classification, size of feature information, size of feature DB according to evaluation item by improved evaluation method and to conduct a performance evaluation experimental test for technical measurement against data set needed for evaluation, methods for quantization of evaluation results.