Texture Mapping In Structured Light Terrain 3D Measurement

With the concept of "Metaverse",virtual reality has become a hot trend in current developments.As an important constituent of the environment in virtual reality,threedimensional terrain can provide more accurate environmental data for scene simulation in many fields if the detailed information of real terrain can be expressed.At present,the commonly used three-dimensional laser scanners mostly collect the depth information of the real world by rotational scanning,but the complex detail texture information of the surface cannot be obtained at the same time,but the texture information can be obtained at the same time in the way of structured light by active projecting the light source,and this paper investigates the texture mapping in the structured light based terrain three-dimensional measurement.Using the principle of structured light measurement,a three-dimensional terrain texture mapping system based on multi-point structured light was designed in this paper,and proposes an overall technical scheme of image separation and reconstruction mapping based on dot matrix structured light.By analyzing the characteristics of structured light measurement images and combining the limitations of the traditional robust principal component analysis algorithm for front and rear view separation,which is prone to image distortion and low image quality,the improved robust principal component analysis front and rear view separation algorithm is studied.Based on the idea of Homomorphic filtering,the foreground light spot information in the image and the background texture image are firstly converted from multiplicative to additive relationship,and on the basis of the traditional robust principal component analysis model.This is based on the traditional robust principal component analysis model,in which the noise caused by stray light and equipment in the image is separated as one of the classes,so as to obtain a noise-free foreground image and a background image.The separated foreground spot image uses morphological processing to remove burrs,isolated points,etc.on the spot features,and uses the centroid coordinate extraction algorithm to locate the spot centroid coordinates,and the obtained two-dimensional spot centroid coordinates are brought into the point cloud solution based on the mapping relationship.Calculate the model,obtain the spatial point cloud of the three-dimensional terrain,and then apply the triangular mesh growth method to realize the Delaunay triangular mesh division,and obtain the triangular meshed model of the point cloud data.Finally,the parametric texture mapping method based on harmonic mapping is introduced to obtain the texture coordinates corresponding to the three-dimensional model,so as to realize the one-to-one correspondence between the three-dimensional mesh model and the texture image,and map the texture image on the surface of the three-dimensional mesh model to generate detailed threedimensional terrain model with information and texture image information.By building a structured light measurement system platform,texture mapping experiments were carried out on two types of three-dimensional pavement topography.As the experimental results shown,the quality of the structured light foreground and background images separated by the algorithm proposed in this paper is better,and a realistic three-dimensional virtual terrain model can be obtained after texture mapping,which can restore the details of the real terrain surface to a large extent.Compared with the limitation of traditional three-dimensional measuring instruments that can only collect depth information,the surface texture and depth information of the terrain can be obtained at the same time through the structured light measuring system by only using monocular measuring equipment,which has the advantages of simple equipment,low implementation cost and collecting comprehensive information at the same time.