Image positioning matching method based on radial annular histogram

Abstract

The invention discloses an image positioning matching method based on a radial annular histogram. The method comprises the following steps: selecting the maximum inscribed circle in a template image as an effective area, the circle center O thereof being (iO, jO), wherein every time when the template image is translated on a target image, a circular area, which is same with the effective area, is obtained on the target image; obtaining radial gradient code of any one point in the effective area and all circular areas, summarizing a histogram of each radial gradient code and obtaining the radial annular histogram of the template image and the target image; and carrying out calculation according to the radial annular histogram to obtain similarity measurement amount S, and through the comparison of the similarity measurement amount S and the threshold value, realizing positioning matching of the target image. The image positioning matching method realizes positioning matching of target objects which are not obvious in characteristics by utilizing rotation invariance and gradient characteristic information of the radial annular histogram, has good robustness in illumination variation and can be suitable for multiple-target matching requirements.

Description

technical field [0001] The invention belongs to the technical field of image processing, and more specifically relates to an image positioning and matching algorithm based on a radial circular histogram. Background technique [0002] IC packaging products are developing toward integration and miniaturization, which requires IC electronic packaging equipment to have high-speed and high-precision positioning capabilities, so as to quickly and accurately pick up and place chips. In the above process, the image matching positioning algorithm plays a role. key role. [0003] There are two common image matching and positioning algorithms: grayscale matching and feature matching. Gray level matching uses gray level information to calculate directly. The algorithm is simple and the accuracy is high, but it is sensitive to nonlinear lighting changes. The algorithm of feature matching usually extracts edge features, which greatly reduces the amount of calculation, has the characteri...

AI Technical Summary

Problems solved by technology

[0004] However, in current IC packaging applications, the template image may undergo translation, rotation, and other transformations relative to the target image, and nonlinear lighting changes and image blur often appear at the same time. If the grayscale matching algorithm is used, it cannot adapt to nonlinear lighting changes; if the feature matching algorithm is used , due to the blurred target image, the features are not obvious, and the fine edges cannot be normally extracted for feature matching

Therefore, it is necessary to find a matching algorithm that has better robustness to illumination changes and can better deal with the problem of indistinct features caused by blur

Images