A three-dimensional measurement system and method for obtaining real 3D texture point cloud data

Abstract

The invention discloses a three-dimensional measurement system and a method for obtaining real 3D texture point cloud data. The method includes the following steps: S1: acquiring an image I1 when the laser is on and an image I11 when the laser is off; S2: locking the images obtained by I1 and I11 Perform image processing to obtain the binarized contour line l1(u,v); S3: Obtain the grayscale curve h(u,v) corresponding to l1(u,v) without laser and restore it to obtain an object in a three-dimensional space A real image of a cross section; S4: Repeat steps S1-S3 to obtain a series of contour lines with grayscale information; S5: Obtain point cloud data with grayscale information, and use the three-dimensional surface contour reconstruction technology to reconstruct the point cloud data Construct a 3D digital model with surface texture information. The invention combines texture mapping with three-dimensional model reconstruction to reconstruct a three-dimensional model with real surface texture, and maps single-pixel contour lines obtained by locking imaging to grayscale images obtained without laser illumination, thereby improving texture mapping accuracy.

Description

technical field [0001] The invention belongs to the technical field of image processing, in particular to a three-dimensional measurement system and a method for obtaining real 3D texture point cloud data. Background technique [0002] With the rapid development of machine vision, 3D measurement and computer graphics, 3D measurement of objects has become a research hotspot because of its wide application value. The optical three-dimensional measurement technology based on laser scanning can easily obtain and display the real geometric contour structure of the object surface, but the displayed three-dimensional image is generally a mesh and point structure, which can only reflect the geometric contour of the object and cannot reflect the real texture of the object surface Variety. In many applications, it is hoped that the displayed 3D image can not only reflect the geometric structure of the object, but also reflect the real texture and even color of the object surface. Fo...

AI Technical Summary

Problems solved by technology

Generating a realistic 3D model involves several challenging issues: acquisition of depth images and texture images, system calibration, matching of depth images, fusion of depth images, mapping of texture images and depth images, 3D model reconstruction, etc.

However, there are two disadvantages: (1) The texture mapping is realized by not fixing the camera position, the camera parameters need to be estimated, and the accuracy is restricted, and error estimation and compensation must be performed to improve the mapping accuracy; (2) 3D mesh model It is stored separately from the 2D texture image, and the 3D model with texture is generated through texture mapping, which is a simulation of the texture of the 3D model, not a true restoration of the surface texture of the model

Images