Fractional order edge detection method for noise pollution image

Abstract

The invention discloses a fractional order edge detection method for a noise pollution image. The method is characterized by comprising the following steps: a noise pollution gray level image is denoised; the denoised image enters a two-dimensional fractional differential filter for edge extraction; the de-noising processing method comprises the following steps: carrying out convolution on a graylevel image and a fractional order integral filter or a Gaussian filter to obtain a filtered image; the edge extraction method comprises the following steps: carrying out convolution on a denoised image and two-dimensional fractional differential filter masks in the horizontal direction and the vertical direction, and solving a gradient image. The improved two-dimensional fractional order integralfilter and the improved two-dimensional fractional order differential filter are adopted, and when the improved two-dimensional fractional order integral filter and the improved two-dimensional fractional order differential filter are used for noise image edge detection at the same time, the characteristics of high robustness, edge positioning accuracy and edge enhancement on noise are achieved.

Description

technical field [0001] The invention relates to the field of image edge detection, in particular to a fractional order edge detection method for noise polluted images. Background technique [0002] Most existing edge detection methods use integer order difference, mainly including Robert operator, Prewitt operator, Sobel operator, Canny operator, Laplace operator and so on. Integer-order difference has the advantage of orientation, but since noise and object edges are high-frequency signals, none of the existing algorithms can effectively compromise noise resistance and edge detection accuracy when performing edge detection on noisy images. [0003] Canny discloses an edge detection operator optimized between anti-noise interference and precise positioning in the article "A computational approach to edge detection". The operator includes four steps: [0004] A. Filter and smooth the image [0005] B. Calculate the gradient magnitude in the horizontal direction and vertical...

AI Technical Summary

Problems solved by technology

[0012]Composite derivative operator first proposes to use the first-order fractional differential and integral to image edge processing at the same time, the image after edge detection, the target internal and background areas are affected by the noise It can resist the interference of some noise, but the image after non-maximum suppression and threshold processing, the target edge is discontinuous, and the edge is sensitive to noise, etc.

[0013] Fractional calculus can enhance high-frequency signals while maintaining the characteristics of low-frequency signals. CRONE operator and compound derivative operator have made some breakthroughs in edge detection, but still does not solve the conflict between detection accuracy and noise immunity well

Images