A Simple Way to Realize the Accurate Detection of Cells´ Edge

Cell edge detection is a basic and important issue in medical image processing, which was crucial to subsequent target feature extraction and object recognition. The traditional methods of image edge detection such as Roberts, Sobel, Prewitt, Laplacian, and Canny, all of which are gradient-based algorithm and are sensitive to noise often cannot locate the edge accuracy. In this paper, the disadvantage of the traditional methods are introduced at first, and then a new mathematical morphological and threshold segmentation edge detection algorithm is proposed to detect the edge of cell image with white noise. In the process, threshold segmentation was used to reduce the effect of noise, and the dilate algorithm was used to extract the boundary. At last compare with the traditional methods. The experimental results show that the new approach is more effective than traditional methods to reduce the impact of noise on test results.