空间目标快速轮廓特征提取与跟踪技术

为满足空间目标交会对接任务中高精度、快速的测量要求,提出了一种空间目标快速轮廓特征提取与跟踪技术。该算法首先从初始帧图像中分割定位目标所在局部区域,作为目标连续跟踪的初始值；其次基于初始帧目标局部区域完成对初始帧目标边缘特征的检测及细化处理；最后采用Hough变换完成对初始帧目标边缘的检测及细化后的局部图像轮廓直线的提取,分别选取目标轮廓四方向最优的直线参数作为最终目标轮廓直线获取的效果,并采用梯度最大法则实现两两求交获取的轮廓特征的优化提取。在目标逼近过程中,结合相邻帧图像间目标尺度动态变化的关联性,根据初始帧提取目标轮廓特征的先验信息,确定目标在第二帧图像中的轮廓位置,并依次根据上一帧图像的轮廓位置信息定位目标在当前帧所在的区域,通过局部处理实现序列图像轮廓区域特征的连续跟踪。该算法无需遍历整个图像,所需处理的目标区域大幅减小,能够有效克服由目标图像较多边缘干扰导致的轮廓提取效果差及处理速度慢的缺点,具有速度快、准确性强、稳定性高等优点。

Research on Fast Contour Feature Extraction and Tracking Technology for Space Target

In order to meet the requirements of high precision and fast measurement in the space target rendezvous and docking tasks, a fast contour feature extraction and tracking technology for the space target is proposed. Firstly, the local target region is segmented and located from the initial frame image in this algorithm and it is used as the initial value of the target continuous tracking. Then, the detection and edge thinning of the initial frame target edge feature is performed based on the initial frame target local region. Lastly, the Hough transform is used to extract the contour line of the initial frame target local edge detection and thinning image, and the best linear parameters in the four directions of the target contour are selected as the final target contour line and then the maximum gradient method is used to obtain the optimal contour feature extraction. Combining the correlation with the target scale dynamic change between adjacent frame images during the target approaching phase, the contour feature of the target in the second frame image is determined based on the prior information of the contour feature extracted from the initial frame. The current frame image position area is located according to the position of the previous frame image and then the continuous tracking of the contour region feature of the sequence image is realized by local processing. The algorithm proposed does not need to traverse the whole image and can reduce the target area to be processed greatly. It can overcome the drawbacks of the poor contour extraction due to the interference of the target image edge and low processing speed effectively and has a advantage of fast speed, high accuracy and high stability.