AN IMPROVED HYBRID METHOD FOR MOVING OBJECT DETECTION

摘要

Recently, computer vision has been used to perform several everyday tasks. The detection of moving objects is a basic task for computer vision applications. In this paper, existing solutions are reviewed, and then novel hybrid approach to detect moving objects is proposed. The approach combines the inter-images difference based on entropy image and optical flow computed by a local method with a hierarchical coarse-to-fine optical flow estimation. As illustrated experimentally, the proposed method can improve detection performances.