In this paper, an object re-identification method based on fusion of local features is proposed. Video image is obtained through visual sensor network and objects in scenes are concerned. Firstly, object is divided into multiple regions based on color feature and shape feature. Then, feature descriptors are designed to extract color and shape information of key regions. Finally, object re-identification is implemented by evaluating the similarity among the objects. The effectiveness of the proposed method is verified by experiments.
修改评论