Light field depth estimation method based on splitting iteration algorithm

Abstract

The invention discloses a light field depth estimation method based on a splitting iteration algorithm, and belongs to the field of computer vision and light fields. The method comprises the following steps: A, correcting an input light field image to obtain a corrected light field image; B, obtaining a sub-aperture image I (x + [delta]x) of the other view angle of the corrected light field image I (x) after the displacement [delta]x occurs based on the Fourier phase shift theorem; C, performing stereo matching on all the sub-aperture images, and constructing an energy function to obtain an initial depth map; D, performing confidence calculation on the initial depth map, finding out inaccurate pixel points in the initial depth map, and optimizing the initial depth map; and E, obtaining an optimized depth map. According to the method, the problem that the baseline between the sub-aperture images is too narrow is effectively solved, the matching quality is improved, and the accuracy degree of the depth map is ensured as much as possible in the calculation process of the initial depth map.

Description

technical field [0001] The invention relates to the technical fields of computer vision and light field, and relates to a depth estimation method, in particular to a light field depth estimation method based on a split iteration algorithm. Background technique [0002] Since 1981, light field technology has gradually attracted the attention of top scientific research institutions in the world, and has carried out many continuous researches. With the rapid development of computer technology and the gradual improvement of microlens manufacturing technology, the combination of light field and computer vision becomes possible. In 1992, Adelson proposed the theory of the seven-dimensional plenoptic function, using the seven-dimensional function to represent the light reflected by the surface of the object, and describing any available information of the observer in any space and any time point. In 1996, Levoy further proposed the four-dimensional light field theory, which parame...

AI Technical Summary

Problems solved by technology

However, since the energy function cannot comprehensively constrain the matching, the depth map obtained by this method still has the problem of inaccurate depth information of some pixels, especially in the edge part

Therefore, accurate depth estimation from 4D light-field images remains a great challenge.

Images