一种改进的边缘细化方法

摘要

针对Sobel算子检测出的图像边缘较粗且检测效果受噪声影响大的问题，提出了一种结合自适应平滑滤波并改进原细化算法的方法来抑制噪声并细化边缘。使用新的自适应平滑滤波梯度模板对原始图像滤波，在平滑噪声的过程中锐化图像边缘；在Sobel算子检测出边缘后使用改进的细化算法剔除伪边缘点。试验比较表明：在迭代次数相同的情况下，新的自适应平滑滤波模板比原梯度模板总体效果更好；改进的边缘细化算法在保证与原算法运算时间相当的条件下，保留了更多的边缘细节信息，细化结果更加准确。%An improved edge thinning method with adaptive smooth filter is proposed for de-noising and thinning rough edge detected by Sobel operator.A new gradient operator is used for adaptive smooth filter,which can overcome noise influence and sharpen image edge effectively.Then,the improved edge thinning method is used after edge detection with Sobel operator.Experimental results show that the new gradient operator is better than the old one with the same iteration number,and the improved edge thinning method can preserve more edge details with a considerable calcula-tion time meanwhile.