Ying Xie Noyes

Real-time scene change detection assisted with camera 3A: auto exposure, auto white balance, and auto focus

Many scene change detection techniques have been developed for scene cuts, fade in and fade out by analyzing video encoder input signals. For real time scene change detection, sensor input signals provide first-hand information which can be used for scene change detection. In this paper, by analyzing camcorder front end sensor input signals with our proposed algorithms based on camera 3A (auto exposure, auto white balance and auto focus), a novel scene change detection technique is described. Camera 3A based scene change detection algorithm can detect scene changes in a timely manner and therefore fits well for real time scene change detection applications. Experimental results show that this algorithm can detect scene changes with good accuracy. The proposed algorithm is computationally efficient and easy to implement.