Corner point detection method based on diagonal lines of parallelogram

Abstract

The invention provides a corner detection method based on parallelogram diagonals. The corner detection method is used for solving the problem that an existing corner detection method is low in detection efficiency. The method comprises the following steps: extracting a contour from an original image by using a Canny edge detector, and marking the extracted contour line; finding a T-shaped angular point of a linear contour or a closed contour; smoothing the contour line through an anisotropic Gaussian directional derivative filter to obtain a smoothed curve; calculating the discrete curvature of each pixel point on the smoothed curve by using the ratio of the diagonal lines of the parallelogram as a corner response function, and taking the pixels of which the curvature values are greater than a set threshold value as candidate corner points; and performing non-maximum suppression on the candidate angular points to obtain an angular point set. According to the method, the curvature is rapidly estimated according to the ratio of the diagonal lines of the parallelogram to achieve angular point detection, the excellent angular point detection performance is achieved, the calculation complexity is greatly reduced, the time complexity is low, the calculation efficiency is high, and good robustness to noise is achieved.

Description

technical field [0001] The invention relates to the technical field of image processing, in particular to a corner detection method based on parallelogram diagonals. Background technique [0002] Image corner detection is a key preprocessing step in image analysis and computer vision, and it is often used in image registration, object recognition, scene analysis and other fields. Existing corner detection algorithms can be roughly divided into three categories: model-based methods, grayscale-based methods and contour-based methods. Among them, the contour-based corner detection algorithm has the advantage of small positioning error. [0003] The most important thing for contour-based corner detectors is how to construct effective and efficient corner response functions. The complexity of the construction of the corner response function or the multiple smoothing operations of the curve will restrict the detection efficiency of the corner detection method. Mokhartian et al....

AI Technical Summary

Problems solved by technology

[0005] Aiming at the technical problem of low detection efficiency of existing corner point detection methods, the present invention proposes a corner point detection method based on parallelogram diagonals, using the ratio of quotients as the curvature estimation function, which not only greatly reduces the amount of calculation but also enhances the Robustness to consistent scale transformation, based on the average repetition rate and localization error evaluation criteria, the experiment proves its efficiency and robustness

Images