Video Filtration for Content Security based on Multimodal Features

In order to meet the urgent requirement for management of video websites,an online video classification method based on multimodal features is designed and thereby the security supervision of the videos realized. This method filters the input videos by such different features as audio,color motion and space-time features in a specific order. Based on the definition of the illegal scenes,including horror,violence and pornography,the potential illicit information in videos is detected. Experiments show that this method can effectively improve the precision rate of detection and classification.