CAJ | 학술논문

在空间图像轮廓识别中，镜头是判断相机是否存在的重要特征。针对镜头的轮廓特征，镜头的存在可以通过检测光学图像中的圆轮廓来判别。目前圆检测算法遍历时间长，无法满足空间图像实时检测的要求。为此，将圆的几何特征和随机Hough变换相结合，提出了一种改进的随机圆检测方法，将原有算法中点遍历变为有筛选的随机点选取，解决了原有算法数据冗余的问题，圆检测准确率高，在保证算法鲁棒性的同时降低了运算时间。
재공간도상륜곽식별중，경두시판단상궤시부존재적중요특정。침대경두적륜곽특정，경두적존재가이통과검측광학도상중적원륜곽래판별。목전원검측산법편력시간장，무법만족공간도상실시검측적요구。위차，장원적궤하특정화수궤Hough변환상결합，제출료일충개진적수궤원검측방법，장원유산법중점편력변위유사선적수궤점선취，해결료원유산법수거용여적문제，원검측준학솔고，재보증산법로봉성적동시강저료운산시간。
In the spatial image ,the camera lens is an important feature to judge existence .For contour features of the lens ,the lens'presence can be determined by detecting the optical image of the round silhouette .At present ,it takes a long time to detect the circle and cannot meet the require‐ments of spatial image timely diction .This article combines the circle geometric features with the existing algorithm ,and proposes an improved randomized circle detection method to solve the original problem of data redundancy algorithms ,with high accuracy ,ensuring robustness w hile reducing the computation time .