Local Adaptive Clustering Based Image Matching for Automatic Visual Identification

In industrial production, monitoring cameras are widely used to pay attention to the running status of equipments. With the development of computer vision, it is possible to use image features to recognize devices. In this paper, a vision-aided identification system is designed to realize real-time automatic labeling of equipments by image matching in surveillance videos. The ORB algorithm is used to extract the image features, and the GMS algorithm is used to eliminate the wrong matching points. Then, according to the clustering locality and template locality, a Local Adaptive Clustering (LAC) method is proposed to optimize the position of labels. Matching templates are segmented with the center of clusters, which improves the efficiency and stability of labels. Experimental results show that LAC significantly suppresses the drift of labels.