论文部分内容阅读
基于椭圆和圆的对称性，对传统的霍夫变换检测曲线方法进行改进，提出分三步快速检测椭圆和圆的方法：（１）利用椭圆的几何对称性，快速确定可能的形心，生成相应子图，并滤除图象中的噪声；（２）在各子图中由边界点梯度向量和候选形心坐标，获得椭圆旋转角和长短轴比率；（３）根据椭圆参数方程确定椭圆的长短轴，并验证所检测到的椭圆．实验结果表明，该方法有效地降低了霍夫变换参数空间的大小和运算时间，提高了检测精度，实现了椭圆特征的快速提取．
Based on the symmetry of ellipse and circle, the traditional method of Hough transform detection curve is improved, and the method of rapid detection of ellipse and circle in three steps is proposed. (1) The geometrical symmetry of ellipse is used to quickly determine the possible centroid and generate (2) Obtain the rotation angle of ellipses and the ratio of long axis to short axis in each subgraph by the boundary point gradient vector and candidate centroid coordinates; (3) According to the ellipse parameter equation, determine the ellipse Of the long axis, and verify the detected ellipse. Experimental results show that this method can effectively reduce the size and computation time of the Huff transform parameter space, improve the detection accuracy, and achieve the rapid extraction of elliptical features.