An Automatic Tracking Approach for Monitoring Moving Targets from Meteorological Satellite Image Sequence Based on Point-Pattern Matching

Abstract

Identification and tracking of moving targets from image sequence is an active branch of image processing and computer vision field. In meteorology, automatic tracking of the trajectories of convective systems from meteorological satellite remote sensing images remains a challenging issue. To monitor the evolution of Mesoscale Convective Systems (MCS) over the Tibetan Plateau in China in summer, an automatic tracking approach based on point-pattern matching is proposed using GMS brightness temperature data. The method tracks the trajectories of MCS by determining the spatial location relationship between sets of points representing MCSs on consecutive images. The experimental results demonstrate the capability of the proposed method to track the evolution of MCS life cycles, especially which have smaller area and faster velocity of displacement. Furthermore, the tracking results show the activity frequency and geographic distribution, and the trajectories and propagation of MCS over the Tibetan Plateau from June to August 1999-2000.