Space target detection in optical image sequences for wide-field surveillance
Dan Liu,Xiaodong Wang,Yunhui Li,Zeming Xu,Jianing Wang,Zhonghui Mao
DOI: https://doi.org/10.1080/01431161.2020.1782508
IF: 3.531
2020-08-12
International Journal of Remote Sensing
Abstract:A wide-field surveillance system with a long exposure time has a stronger detection capability for faint space targets. However, some of the complexities it generates also pose difficulties for space target detection; a large amount of image data, numberless object points, some stars manifesting as streak-like sources, and possible discontinuous or nonlinear target trajectories. This paper presents a high precision and low computational-cost space target detection method to overcome these obstacles. Firstly, the minimum external rectangle method is implemented to effectively remove stars and noise. Secondly, the motion velocity of the targets is calculated as the basis for predicting the allowed state transition region in each image of the frame set. Finally, a dynamic programming sliding window method is proposed to detect space targets with continuous, discontinuous, linear or nonlinear trajectories. The experimental results show that this method can effectively detect faint space targets in wide-field surveillance under a long exposure time. This method also has the advantage of a low computational cost.
imaging science & photographic technology,remote sensing