An Improved Optical Flow Field Model Method Based on Eigenvectors

Abstract

The invention discloses an improved optical flow field model algorithm based on a characteristic vector. The improved optical flow field model algorithm comprises the steps of respectively constructing Gaussian pyramid image layers of a reference image and a floating image, respectively extracting character vectors of the reference image layer and the floating image layer; replacing a brightness constancy assumption in a traditional optical flow field model by characteristic vector constancy, and constructing an energy function based on characteristic vector constancy; minimizing an energy function in each image layer, and solving a motion displacement field between the reference image and the floating image by means of optical flow iteration; and correcting the floating image by means ofthe obtained motion displacement field, and obtaining a registering image. The improved optical flow field model algorithm performs improvement for aiming at incapability of effectively registering large-displacement deformation and easy smoothing in optical flow estimation in a traditional optical flow field model, thereby improving registering precision of a non-rigid image. The improved opticalflow field model algorithm can perform automatic registering on the non-rigid image with relatively large displacement deformation, and can be widely applied to the field of medical image processing,image fusion, mode identification, etc.

Description

technical field [0001] The invention belongs to the field of computer vision, and more specifically relates to an improved optical flow field model method based on feature vectors. Background technique [0002] Image registration is to align two or more images of the same target at different acquisition times, different sensors, and different acquisition conditions on physical coordinates, so as to realize information sharing and complementarity. Gain more comprehensive information and understanding. Due to different imaging conditions, multiple images of the same object have differences in resolution, imaging mode, grayscale attributes, etc. Therefore, the registration of these images is a typical problem and technical difficulty in the field of image processing research. [0003] Image registration has a wide range of applications in the fields of aerospace, medical image processing, remote sensing images, pattern recognition, etc., and has important research value and s...

AI Technical Summary

Problems solved by technology

[0005] The purpose of the present invention is to overcome the deficiencies in the prior art and provide an improved optical flow field model method based on eigenvectors, aiming at the inability of the traditional optical flow field model to effectively register large displacement deformation and optical flow estimation The over-smoothing problem that is prone to occur in the image has been improved to estimate the large displacement motion in the image to improve the registration accuracy of the non-rigid image. It can automatically register the non-rigid image with large displacement and deformation, and can be widely used in Medical image processing, image fusion, pattern recognition and other fields

Images