Microscopic sequence image automatic splicing method, system and device based on affine transformation

Abstract

The invention belongs to the technical field of computer vision and image processing, particularly relates to a microscopic sequence image automatic splicing method, system and device based on affinetransformation, and aims to solve the problems that in the prior art, the influence of image deformation on splicing in the microscopic image shooting process cannot be effectively solved, splicing errors can be accumulated and spread, therefore, the expected splicing effect cannot be achieved. The method comprises the following steps: extracting features of an overlapping region of a microscopicsequence image and carrying out feature matching; endowing each pair of feature points with different weights according to the position information of the features on the image; fitting changes between adjacent images through an affine transformation model, and setting a globally optimized energy function; minimizing the energy function to obtain an affine transformation relationship of each adjacent image; and performing splicing according to the affine transformation relationship to obtain a microscopic spliced image. Through the global splicing method, the influence of error accumulation and image edge distortion on the result in the splicing process is avoided, and the image with higher splicing precision can be obtained.

Description

technical field [0001] The invention belongs to the technical field of computer vision and image processing, and in particular relates to an affine transformation-based automatic splicing method, system and device for microsequence images. Background technique [0002] From the manufacture of the first optical microscope in the early 17th century to the design of the first transmission electron microscope in the early 20th century, human beings have gradually deepened their exploration of the microscopic world. With the rapid development of electronic technology, the resolution of some transmission electron microscopes can reach sub-angstrom level, and people can clearly see the structure of atoms. Microscopes are increasingly used in fields such as medical research and biology, as the images obtained are at higher and higher resolutions. It also brings a problem, the higher the resolution of the image obtained, the smaller the imaging field of view of the microscope. In t...

AI Technical Summary

Problems solved by technology

[0008] In order to solve the above-mentioned problems in the prior art, that is, the prior art cannot effectively solve the influence of the image deformation on the splicing during the microscopic image shooting process, and the splicing errors will accumulate and propagate, so the splicing effect cannot reach the expected problem. The invention provides an affine transformation-based automatic mosaic method for microscopic sequence images, the method comprising:

Images