An Efficient Clustering Mining Algorithm for Indoor Moving Target Trajectory Based on the Improved AGNES

Abstract

In recent years, with the rapid development of wireless communication technology including Wi-Fi, Bluetooth and RFID and other new types of positioning method, the indoor mobile object positioning has become possible. At present the research on indoor mobile object trajectory analysis is still in the start stage. But as people and goods stay indoor environment for most of time, the indoor positioning technology and the analysis of the indoor moving targets track will be the developing trend in the future. When deployed in real environment, the existing indoor moving target trajectory analysis methods need high equipment cost and their scalability is also very poor. In this paper we proposes an algorithm for indoor moving target trajectory analysis and data clustering based on improved AGNES algorithm. Through improving the weighted function of the algorithm, we realize the extraction and analysis of the indoor moving target trajectory. After deploying in the actual environment, we test the algorithm in practice. The results indicate that the improved algorithm greatly reduces the number of hardware and the deployment cost. And it can also effectively improve the efficiency of the moving target trajectory analysis.