基于递归滤波的边缘检测算法

针对包含不同程度噪声数字图像的边缘检测问题，提出了两种建立在房顶型模糊边缘模型基础上的平滑和边缘检测滤波器。首先用这些滤波算子以三阶递归的形式实现图像的平滑及梯度计算，再进行非极大值抑制及双阈值的边缘检测连接。实验结果表明，该方法得到的梯度图像均比Canny和Deriche滤波算子清晰，所得边缘图像也更加完整，检测时间也少于其他方法。

Edge detection algorithm based on recursive filtering

For the edges detection problem of digital images containing different degrees of noise, two smoothing and edge detection filters, based on a model of crest-line blurred edge contours, were proposed. First, these filters were used to achieve image smoothing and gradient calculation by the third-order recursive form, and then non-maximum suppression and dual-threshold edge detection were performed. The experimental results show that the proposed method can obtain more clear gradient images, accurate detected edge and less computational time than the ones with Canny and Deriche operators.