An edge-detection algorithm based on improvement average filter for noise image

Aiming at the traditional edge detection algorithm can not effectively detect the edge in noise image, a new combined edge detection algorithm is proposed in this paper. There are mainly two steps in algorithm, the first step is image filtering, an new improved average filter algorithm based on Rank transform is used in this step, the separate noise points can be found from the original image by Rank transform; the second step is mainly on edge detection, the segment threshold is automatic used in the image after the filter process, a new kind of improved iterative algorithm is used in this selection which can adjust to obtain the optimal threshold. Finally, the compared experiment results shows that the performance of the algorithm in this paper is better than traditional algorithm; the detection accuracy is very high and maximize the image edge information and maintain the details of image.