Image-based 3D Model Construction Method

Abstract

The present invention proposes a method for forming a three-dimensional model based on an image. The method includes: a representation method for a three-dimensional model based on an image: uniformly select a suitable number of vertices on the outer spherical surface surrounding the object, and take the position of each vertex as the coordinate of the camera , taking the position of the center of the sphere as the target point of the camera to obtain the color image and depth image of the 3D object at each camera angle of view, and represent the 3D model through the color image and depth image, as well as the associated camera parameters; image-based 3D Model acquisition method: Determine the level of detail (LOD, level of details) of the 3D model according to the needs of practical applications, triangulate the spherical surface surrounding the 3D object according to the LOD, and use the triangle approximation method to construct a spherical interior with triangles as the basic shape Connect the polyhedron, take each vertex of the polyhedron as the camera position, and use the center of the sphere as the camera target point position to obtain the depth image and color image of the object at this sampling point, and the depth image, color image and camera parameters together constitute the object. 3D model. The present invention has low coupling between the modeling complexity and the geometric appearance complexity of the object, so it is suitable for modeling complex three-dimensional objects.

Description

technical field [0001] The invention relates to the technical field of three-dimensional model modeling, in particular to an image-based three-dimensional model construction method, which is mainly used for the acquisition and representation of the three-dimensional model of complex objects. Background technique [0002] 3D model representation is the basic work in graphics, and the representation method of the model determines the drawing and processing method of the model. In graphics, a variety of 3D model representation methods have been developed, each of which has its own advantages and disadvantages for specific problems. These representation methods can be roughly divided into three categories: geometric representation, point cloud representation, and hybrid representation. Geometric representation is to use basic geometry to represent the surface of an object or an object entity. Point cloud representation is to represent the object by sampling the surface of the o...

AI Technical Summary

Problems solved by technology

[0010] The existing 3D model modeling methods have the following disadvantages: In geometric representation, polygonal representation is the most common method, which requires professionals to model in 3D modeling software, and the modeling method takes a long time. As the complexity of the modeling object increases, modeling will become more and more difficult, and the later realistic rendering will consume a lot of time

Point cloud representation can accurately describe complex models, but the preprocessing process is complicated. Point cloud rendering has high requirements on the pixel filling rate of graphics hardware, and the rendering effect is not realistic enough. Point cloud models often need to use primitives to represent 3D objects. Sampling point interpolation approximates the details of the original model, and is less efficient for large plane representations in objects

The hybrid method combines the advantages of geometric representation and point cloud representation, and can automatically select the representation of the model according to the actual situation, but this method requires preprocessing of the model, which increases time consumption

Images