一种改进的特征匹配算法在碎片云测量建模中的应用

为满足超高速撞击典型Whipple防护构型的损伤评估需求,利用图像处理技术对碎片云序列阴影图像进行深入研究。使用超高速序列激光阴影成像仪得到三组不同实验条件下碎片云发展过程的高清阴影图像,分别对每组最具代表性的2帧进行图像处理分析;根据碎片云图像特点以及碎片运动特性,提出了一种改进的碎片二次特征匹配算法,该方法包含碎片粗定位、特征定义及初匹配和精确匹配三步策略;通过运用改进的匹配算法,对选取的相邻两帧图片完成碎片高效匹配,并提取匹配碎片的运动参数,进而分析碎片的速度分布和飞行角度分布,获取二次碎片云相关运动特性;得到三组实验各自的轨迹模拟图。根据得到的轨迹分析结果分别对三组实验的后板损伤进行估计,并通过与防护构型的实际损伤结果进行比较,验证了该方法的有效性。

An improved feature matching algorithm and its application in measurement for modelling of debris clouds

In order to satisfy the assessment requirement of the damage effect of hypervelocity impacts(HVIs)of space debris on typical double-plate Whipple shield configuration,an insight investigation on the distributive characters of debris clouds in successive shadowgraphs was carried out by using the image processing technique.The high-resolution shadow images of the debris cloud development process under three different experimental conditions were acquired by the hypervelocity sequence laser shadowgraph image,and the most representative 2 frames of each group were processed and analyzed respectively.According to the characteristics of debris cloud image and the motion of debris,an improved double feature matching algorithm was proposed.In terms of such an improved algorithm,an effective matching of the debris in the selected two frames was obtained,and the corresponding characteristics were extracted.Furthermore,on this basis,the velocity distribution and leap angle distribution of the debris were analyzed,and the related motion characteristics of the secondary debris cloud were obtained.Finally,the damage situation of the rear wall was estimated by a constructed model,and the validity of the proposed algorithm was verified through the comparison with the actual damage results of the front and rear walls.