A Fast Image Registration Method Based on Visually Salient Regions

Abstract

The present invention discloses a quick image registration method based on a visual remarkable area, which belongs to the field of image processing. The quick image registration method is used for matching an image A to be matched with a reference image B and respectively acquiring remarkable areas of A and B. The mass centers of three remarkable areas of A are vertexes of a triangle, and the triangle is used as a first characteristic triangle. Similarly a second characteristic triangle is acquired. The first characteristic triangle and the second characteristic triangle which are similar form a similar triangle pair, wherein the similarity of (a,b) is maximal. An affine transformation model for transforming from A to B is established based on (a,b), wherein the related initial matching parameters comprise horizontal translation amount, vertical translation amount, rotation angle and scale parameter. From the initial matching parameter, Powell searching for setting step length is performed for acquiring searching values. Each searching value is used for performing multiple registration tests on the A and B, and the searching value which corresponds with an optical registration test is the optimal matching parameter as a result. The optimal matching parameter is used for performing affine transformation on A for acquiring a final registration result.

Description

technical field [0001] The invention belongs to the field of image processing, and in particular relates to a method for quickly registering images using visually significant region-based assistance. Background technique [0002] Multi-source image registration refers to transforming two or more images of the same scene from different sensors, different viewing angles, and different times into the same coordinate system based on the optimally determined image transformation parameters based on some similarity measures, and The process of getting the best match at the pixel level. From this definition, registration includes the transformation relationship between the spatial position and the pixel value of the source images. Because image registration enables the interpretation of a specific scene to be supported by a variety of data, it is widely used in medical image processing, remote sensing image analysis, target recognition, target change detection and other fields. ...

AI Technical Summary

Problems solved by technology

Although this method is fast, it only uses binary images for calculation and uses too little image information, which is easy to cause mismatching; ②The fast image registration method based on shape matching uses the target shape in the two images to match, Use the matching result as the initial value of the optimized search algorithm to improve search efficiency

Although this method combines the characteristics of feature-based and gray-level-based registration methods, since the features extracted in the method are point features of independent target contours and there are many typical value settings in the extraction process, the algorithm is not suitable for multi-shape It is difficult to show strong adaptability when applied in similar more general natural scenes

[0006] To sum up, the grayscale-based image matching method has high registration accuracy and robustness but poor real-time performance; the feature-based image registration method has high real-time and heterogeneous adaptability but Registration accuracy and robustness are somewhat lacking, so the two types of methods have their own different characteristics and good complementarity

Most of the current methods strictly belong to a certain category of methods. While presenting the advantages of their respective algorithms, they also expose their own defects. They cannot achieve automatic registration with high precision, strong real-time performance, and wide adaptability.

Images