Three-dimensional reconstruction system and method for electronic microscopic scene

Abstract

The invention discloses a three-dimensional reconstruction system and method for an electronic microscopic scene. The method comprises the steps: collecting an image sequence of corresponding frames at a certain step through an electronic microscope, carrying out the segmentation processing of the images, carrying out the compiling of the images into a dictionary, solving a sparse coefficient, carrying out the fusion, and carrying out the recovery of the images to generate a clear image. In the specific embodiments, the focusing motion adjustment is completed through the combination of a stepping motor and a screw. The approximate object distance difference of all frames is calculated through a control signal and a positioning height, thereby achieving the three-dimensional reconstruction,and solving a problem of insufficient three-dimensional information caused by a small field depth in a microscopic scene.

Description

technical field [0001] The invention belongs to the field of microscopic vision technology, and more specifically, relates to a three-dimensional reconstruction system and method in an electron microscopic scene. Background technique [0002] With the development of computer vision technology, 3D modeling technology has provided strong technical support in the fields of industrial inspection, quality control, archaeology and biological research, and its wide application has also made great progress in computer vision. Such as binocular stereo vision, texture method, contour method, motion method and statistical learning method. [0003] At present, due to the short depth of field of optical microscopes, rich three-dimensional information cannot be seen when observing objects, especially when the surface of the object is of different heights, no clear image can be obtained at any focal length. Therefore, the 3D reconstruction method in the microscopic scene requires multi-fo...

AI Technical Summary

Problems solved by technology

Traditional 3D reconstruction, such as binocular 3D reconstruction, reconstructs the 3D information of the scene based on the geometric relationship between the two 3D images captured by the camera, but the defocused image of the microscopic scene cannot obtain a complete texture information; laser scanning 3D reconstruction can only obtain 3D data, but cannot obtain the texture information of the object itself, that is, the restoration information is incomplete, and it is not suitable for common microscopic scenarios such as biological research, material analysis, and quality inspection; feature-based The statistical learning method is based on a large target database. Through the learning method, each target in the database is extracted, and then a probability function is established for the features of the reconstructed target. Finally, the reconstructed target and the similar target in the database are combined. The degree of similarity is expressed as the size of the probability, and the depth of the target with the highest probability is taken as the depth of the reconstruction target, and combined with texture mapping or interpolation methods for 3D reconstruction, it is possible to perform 3D reconstruction with only one picture, but the biggest difficulty lies in Obtain a complete database of large objects, also less suitable for microscopic scenarios

Images