Papers by Author: Xu E

Abstract: In order to improve the edge detection efficiency, and decrease the impulse noise impact on the edge detection, a new edge detection algorithm for impulse noise image is proposed. The algorithm combined the median filter idea, and used cross convolution template to calculate image gradients of horizontal and vertical direction. First, detected the point that will attend gradient calculation by threshold, for the noise point, its value will be replaced by median of those points associated with it in detection window. For non-noise point, we directly calculate gradient it initial pixel values. Experiments show that, the new algorithm has strong suppressing noise ability for images corrupted by impulse noise, the extracted image edge is also fine, the outline is clear, the algorithm is better than the traditional Sobel algorithm and the wavelet transform algorithm in noise suppression performance and the detection efficiency, it has strong practicability.

Abstract: Edge detection plays an important role in medical image processing; its accuracy directly affects the diagnosis and treatment of the disease. In view of the shortcomings of the traditional Sobel algorithm, an improved Sobel edge detection algorithm is proposed in this paper. Algorithm increased 135 º and 45 º direction templates, used the local gradient and standard deviation to filter and strengthen the gradient of initial gradient image. Experiments show that the image edge detected by the new algorithm is relatively accurate, complete and clear compared with the traditional Sobel algorithm, which verifies the effectiveness of the new algorithm.