Corner detection method based on non-causal fractional gradient operator

Abstract

The invention discloses a corner detection method based on a non-causal fractional gradient operator, which belongs to the technical field of image processing. According to the invention, non-causal fractional gradient operation of a gray-scale image to be detected is realized through the combination of cause-effect and anti-causal fractional integral and cause-effect and anti-causal fractional differential. The method comprises the steps that an image is read to generate a gray scale matrix f (x, y); non-causal fractional gradients Dx and Dy of f (x, y) in x and y directions are calculated; the product DxDy of the gradient direction is calculated; Gaussian kernel is used to filter DxDy; the corner point intensity is calculated; and non-maximum suppression is carried out to acquire an accurate image corner. According to the invention, the novel algorithm based on non-causal fractional gradient is used to carry out gradient and corner energy operation; the corner detection precision is improved; the method is suitable for the computer vision fields of image registration and matching, image fusion, target recognition and the like.

Description

technical field [0001] The invention belongs to the technical field of image processing, and relates to an image corner detection method, in particular to a corner detection method of a non-causal fractional gradient operator. Background technique [0002] Image corner detection is an important issue in the field of image processing. It is mainly a method used to obtain image features in computer vision systems. It is widely used in motion detection, image matching, video tracking, 3D modeling and target recognition. Used to extract the corner points of the image. The existing corner detection algorithms can be mainly classified into three categories: corner detection based on grayscale images, corner detection based on binary images, and corner detection based on contour curves. The corner detection method of high-degree image, its algorithm is stable and uniform, it retains the important characteristic information of the object in the image and reduces the amount of infor...

AI Technical Summary

Problems solved by technology

[0006] The purpose of the present invention is to overcome the sensitivity to noise when using the Harris corner detection method based on the traditional integer order differential for corner detection, and a large number of false corners will also appear at the same time, and the image usually needs to be smoothed after the corner detection , so that it is easy to cause the weakening or loss of image corner information, which affects the insufficiency of the image corner detection effect. A non-causal fractional gradient operator corner detection method is provided.

Images