Detection and Tracking of Cooperative Target for Space Exploration Task Based on Visual Measurement

Abstract

The detection and tracking of cooperative targets are crucial significance in the rendezvous and docking of space targets. In this paper, a high performance detection and tracking method is proposed based on the combination of adaptive threshold sub-pixel detective positioning with the nearest neighbor clustering analysis of feature points in visual images. In order to enhance the accuracy of detective positioning the bilinear interpolation is employed to achieve sub-pixel coordinates positioning. And then the improved particle filter is used to carry out the prediction and tracking of cooperative targets so as to overcome stochasctic disturbances from stray light and noises. A series of experiment results indicate that the proposed method is characterized with its well performances of detection accuracy and real-time tracking and its outstanding simpleness and practicability so as to play important role in the implementation of cooperative target detection and tracking by visual measurement in the RVD of space exploration.