Wide-baseline color image template matching method

Abstract

The invention belongs to the technical field of image processing and provides a wide-baseline color image template matching method. Firstly, according to image multi-channel characteristics, a vector VSAD is adopted for measuring uniformity of sample pixels of two images, on the basis of which, a Lab-SAD method is given to increase effective information of to-be-matched areas in target images; secondly, a hypothesis that the less colors or color combinations of selected templates in to-be-matched images, the higher the matching correctness probability is provided, a score graph calculation method is given, and the higher pixel gray values in a score graph, the higher scores, the higher accuracy in matching by template selection based on the position; finally, a method for sorting template areas according to the score graph is given, and high-score areas are taken as final template selection areas. According to experimental verification, the wide-baseline color image template matching method has high accuracy and effectiveness under wide-baseline conditions.

Description

technical field [0001] The invention belongs to the technical field of image processing, and in particular relates to a template matching method for a wide-baseline color image. Background technique [0002] Template matching algorithms usually consider all possible transformations, including rotation, scale, and affine transformation; Alexe et al. provide an efficient calculation method to process high-dimensional vectors in two image matching windows, which extracts the overlapping part of the two windows Boundary, and use it to limit and match multiple windows; Tsai et al. proposed a template matching method, which uses wave decomposition and ring projection method and focuses on rotation transformation; Kim et al. proposed a grayscale template matching method, the It has better resistance to rotation and scale transformation; Yao et al. proposed a method for searching color textures, which also considered rotation and scale transformation; under the condition of wide bas...

AI Technical Summary

Problems solved by technology

[0003] Under wide baseline conditions, as the photographic baseline increases, the effective information in the target image of the area to be matched gradually decreases. In this case, effectively using various information in the image and selecting the appropriate position to be matched becomes the key to improving the matching accuracy. An effective means; since the Fast-Match method is a new method based on grayscale image matching, it is necessary to convert the color space first when matching color images. This mapping process will lead to the loss of color information, resulting in the matching of color images Mis-matching occurs in areas with large spatial differences; the template matching method has the following problems that affect the matching accuracy when matching wide baseline images: 1) With the increase of the photographic baseline, the effective information of the target image in the area to be matched gradually decreases; 2) The selection of the matching area has a great influence on the accuracy of the matching results

Images