Content Based Video Retrieval

Content-based retrieval allows finding information by searching its content rather than its attributes. The challenge facing content-based video retrieval (CBVR) is to design systems that can accurately and automatically process large amounts of heterogeneous videos. Moreover, a content-based video retrieval system requires in its first stage to extract the video into separate frames. Afterward, features are extracted for video frames. And finally, choose a similarity/classifier metric and a machine learning algorithm that is efficient enough to retrieve query–related videos results. Histogram of Oriented Gradients (HOG) features are extracted for a video frame and Random Forest Classifier, a machine learning algorithm is used for the classification of the video frame.

Keywords: Content-Based Video Retrieval (CBVR), Histogram of Oriented Gradients, Machine Learning, Random Forest Classifier.