Moving Object Detection and Tracking Techniques for Automatic Audio Mixing
In this paper, we propose a moving object detection and tracking algorithm based on multi-frame feature point tracking information for automatic audio mixing. First, corner feature points and optical flows of multiple frames are calculated to compensate camera movement and track objects. To compensate camera movement, we convert the traced corner feature points byhomography and RANSAC algorithm. Among the transformed corner feature points, the outlier points removed by the RANSAC are clustered and the outlier cluster region of a certain size is classified as the moving object candidate group. Objects classified as moving object candidates are assigned and tracked according to data correlation analysis based on label tracking. In this paper, we have experimentally confirmed that the proposed algorithm can detect and track moving objects such as moving human invideo sequence and that automatic audio mixing or mapping is possible by using the tracking result.
Keywords - Object Detection, Object Tracking, Data Association, Multi-Channel, Automatic Audio Mixing