Structured light three-dimensional measurement method based on orthogonal Gray code and line shift combination

Abstract

A structured light three-dimensional measurement method based on orthogonal Gray code and line shift combination is applied to rapid three-dimensional reconstruction. Gray code fringe edges are used for encoding and decoding to eliminate Gray code single bit decoding errors; a line shift pattern is further projected after the projection of a Gray code pattern, the Gray code fringe edge positioning and the line shift fringe center positioning are combined, the density of image sampling points is improved; through symmetric line shift method from the fringe center to two sides, the edge positioning errors caused by edge spread and global illumination are avoided, the decoding errors of line shift fringe cycle positioning are therefore eliminated, the accuracy of the global decoding is improved, and the edge positioning time is reduced; and the two-way orthogonal projection is used for achieving globally unique encoding of left and right fields of view, matching features are used for completing the one-to-one correspondence of the same name sampling points, and the quantization error in pixel center decoding is eliminated. The invention improves the speed of three-dimensional measurement while ensuring the accuracy and robustness of system measurement.

Description

technical field [0001] The invention relates to a machine vision measurement method, in particular to a structured light three-dimensional measurement method based on the combination of orthogonal gray code and line shift. Background technique [0002] The acquisition of three-dimensional images has important research and application value for computer vision and robotics, and has been widely used in specific fields such as face recognition, human-computer interaction, automatic measurement, visual navigation, virtual reality, aerospace and medical diagnosis. At present, there are various scene 3D information acquisition technologies such as binocular and multi-eye stereo vision, structured light 3D vision, optical zoom 3D measurement, motion-based shape recovery, shadow-based shape recovery, and texture-based shape recovery. As a method of 3D reconstruction and industrial measurement with high accuracy, fast measurement speed and low cost, structured light 3D vision technol...

AI Technical Summary

Problems solved by technology

This encoding method can achieve high spatial resolution, but the color bandwidth of the projector, the change of the color or depth of the measurement surface, the error of the camera and the sensitivity to noise will restrict the measurement accuracy and application of the system.

[0009] Different encoding methods need to project different numbers of encoded images in different modes. The simpler the encoding mode, the more the number of projection patterns and the longer the processing time, but the decoding is simple and the positioning accuracy is high; the more complex the mode, the fewer the number of projections and the processing speed The faster, but the decoding is more difficult, susceptible to noise interference, and the positioning accuracy is not high

Images