CAJ | 학술논문

针对空间绳系机器人在轨捕获任务中，对非合作目标卫星帆板支架线条提取时出现的直线段重复交叉检测、运行速度慢等问题，提出一种区域生长Hough变换算法。该算法利用区域生长算法对种子节点周围的邻域进行拓展，通过对生长区域内的点进行Hough变换检测，逐步缩小新的拓展点的搜索范围，从而降低计算量。同时，利用角度范围限定和动态聚类的方法解决了Hough变换中检测直线过多和直线交点检测的问题。实验表明，该算法计算速度较快，计算精度较高，能够解决Hough变换中同一直线多次检测的问题，可以满足空间绳系机器人在轨抓捕任务的要求。
침대공간승계궤기인재궤포획임무중，대비합작목표위성범판지가선조제취시출현적직선단중복교차검측、운행속도만등문제，제출일충구역생장Hough변환산법。해산법이용구역생장산법대충자절점주위적린역진행탁전，통과대생장구역내적점진행Hough변환검측，축보축소신적탁전점적수색범위，종이강저계산량。동시，이용각도범위한정화동태취류적방법해결료Hough변환중검측직선과다화직선교점검측적문제。실험표명，해산법계산속도교쾌，계산정도교고，능구해결Hough변환중동일직선다차검측적문제，가이만족공간승계궤기인재궤조포임무적요구。
To accomplish the real-time detection of satellite solar panel support when the Tethered Space Robot ( TSR) captures non-cooperative target satellite, a novel Hough transform based on the thought of Region Growth Algorithm is proposed. The Region Growth Algorithm is employed to enlarge the boundary of the set pixels;after the detection of the every single pixel in the growing region using Hough transform, the searching boundary of the new set pixels is gradually reduced so that the amount of calculation is decreased. In addition, the limit of angle range and the dynamic clustering method solve the problem of excessive linear detection as well as the intersection point detection of Hough transform. The experiment results and their analysis show preliminarily that the proposed method can detect accurately with less computational time, and can also solve the problem of Hough transform that one line may be detected for several times. Each result can satisfy the requirement of TSR on-orbit grasp task.