Image-based clustering algorithm in high precision real time location system 1

An image-based clustering algorithm is presented for high-precision real time location system (RTLS). In RTLS, it is necessary to make each cluster consisting of two or more objects when these objects are close enough to meet a specific distance threshold for the minimum period of time. These clusters can show the user of the specific object. The proposed algorithm is based on comparisons between the images generated from the tracking data in a pre-defined area. The processing time in the RTLS using the proposed algorithm can decrease with the almost same detection performance, compared with the previous distancebased algorithm.