论文部分内容阅读
为了更有效地从二值图像中提取更高质量的曲线轮廓,提出一种基于邻域统计编码的轮廓提取算法,它能在快速地提取出待测零件轮廓的同时,滤除毛刺和噪声。在检测精度不变的情况下,具有速度高和鲁棒性较好的特点。
In order to extract more high-quality curve contours from binary images more efficiently, a contour extraction algorithm based on neighborhood statistical coding is proposed, which can remove burrs and noise while quickly extracting the contours of the part to be measured. In the case of the same detection accuracy, with high speed and good robustness characteristics.