Microscopic 3D (Three-Dimensional) reconstruction method based on improved iterative shrinkage thresholding algorithm

Abstract

The invention discloses a microscopic 3D (Three-Dimensional) reconstruction method based on an improved iterative shrinkage thresholding algorithm. The method comprises the steps of setting a relativefuzziness initial value; calculating a solution of a thermal diffusion equation; solving an energy equation to judge whether the solution of the energy equation is smaller than an energy threshold value; when the solution of the energy equation is smaller than the energy threshold value, directly obtaining a depth information value; when the solution of the energy equation is not smaller than theenergy threshold value, calculating an acceleration operator to be substituted into an iterative formula; performing gradient estimation again; and substituting an optimal iterative step length solved by use of linear search of a secant line into the iterative formula to solve the depth information value. With the method, the convergence speed that a defocus depth restores a dynamic optimizationproblem can be accelerated and the precision of microscopic 3D topography reconstruction is improved.

Description

technical field [0001] The invention relates to a microscopic 3D reconstruction method, in particular to a microscopic 3D reconstruction method based on an improved iterative shrinkage threshold algorithm. Background technique [0002] In micro-nano computer vision, vision-based micro-nano-scale 3D shape reconstruction is of great significance for a more comprehensive understanding of sample characteristics and evaluation of operating processes. The commonly used microscopic 3D reconstruction methods mainly include volume restoration methods, stereo vision depth restoration methods, focus depth restoration methods, and defocus depth restoration methods. Compared with other 3D reconstruction methods, the defocus depth restoration method has received extensive attention and in-depth research in recent years due to the advantages of fewer pictures required, simple equipment, and convenient operation. Defocus depth restoration was first proposed by Pentland. This method uses th...

AI Technical Summary

Problems solved by technology

[0005] In view of the deficiencies in the prior art, the purpose of the present invention is to provide a microscopic 3D reconstruction method based on an improved iterative shrinkage threshold algorithm to solve the problem of poor convergence efficiency in the dynamic optimization process of defocus depth restoration, resulting in poor accuracy of reconstructed microscopic 3D topography The problem

Images