how to convert 3D point clouds to 2D images matlab

25 proposed a simplification method based on k-means clustering. Song and Feng 24 reduced the points globally according to the specified simplification ratio. One is based on the global ideology for simplification, and the other is based on the partition strategy, extracting the feature points before simplification. These methods can mainly be summarized into two categories. In contrast, the scattered-point-based methods can consume the point cloud directly. The main disadvantages of these mesh-based methods are that they are seriously complex and building polygonal structural meshes require a great quantity of extra information and memory space. This method is apt to discard feature points in the area where the curvature changes. 23 combined voxel grid with the bounding box, and confirmed the center of each small grid by calculating the distance of k neighborhood and the normal for simplification. Kalvin and Taylor 22 proposed a bounded approximation method, which placed the vertices of the polyhedral mesh in an error region, and this method improved the practical feasibility. The implement is simple and easy, however, different point clouds need different division scales, and the simplification accuracy cannot be guaranteed.

The algorithm constructed a bounding box that surrounded all the points of the 3D model, and the bounding box was divided into several small cubes evenly, and the closest point from the center of the small cube replaced all other points. 21 proposed a simplification algorithm based on the bounding box.

20 simplified the connected triangular mesh through wavelet representation. Hamann 19 developed an algorithm to iteratively delete triangles according to the triangulation, and it had a good effect on the model surface. The mesh-based methods convert the point cloud to the mesh model with polygons, and then reduce the points based on specific rules for simplification. One is mesh-based and the other is scattered-point-based. The previous research on point cloud simplification methods can be generally classified into two kinds. High precision of the 3D scanning technology can provide abundant details of the point clouds, but it will simultaneously generate a mass of redundant data, which seriously reduces the efficiency of data processing, transmission and display, thereby affecting subsequent applications. With the improvement of 3D data acquisition capabilities 1, 2, 3, point cloud models are gaining increasing attention and being widely utilized in many fields, such as scene reconstruction 4, 5, 6, 7, 8, 9, museum display 10, 11, 12, 13, and virtual projection 14, 15, 16, 17, 18.

How to convert 3D point clouds to 2D images matlab