Research On 3D Model Segmentation Method Based On Skeleton

The three-dimensional model is a collection of physical appearance surface points obtained by measuring instruments in reverse engineering,and generally includes information such as point coordinate data and color data,and is widely used in fields such as medicine and virtual reality.The 3D model segmentation is based on the relationship between the point coordinates of the 3D model and the points between the patches without any prior knowledge.By calculating the mathematical features of each point,the clustering method will be used to have similar mathematical features.The points come together and are called "sub-components" of the 3D model.This process is often used as an initial step in applications such as 3D retrieval,compression transfer,texture mapping,animation and geometry transformation,and model simplification.Although the existing 3D model segmentation method can obtain better segmentation results,there are difficulties in selecting the region to grow seed points in the segmentation process,how to determine the growth rules of the skeleton data region,and determining the number of "subcomponents" in the final segmentation result of the 3D model.Three difficult questions.This paper has studied the above problems,and the research content is mainly divided into three parts.In the first part,in view of the difficulty of growing seed points in the region,this paper analyzes the characteristics of seed points and the rules of selecting seed points in detail,studies the geometric characteristics of the three-dimensional model skeleton,and proposes the scheme of using the segmentation points of the skeleton as the regional growth seed point.It solves the problem of randomly selecting seed points during the growing process of the region.In the second part,focusing on the problem of how to grow the 3D model,the topological features of the 3D model skeleton data are analyzed.The main idea of the region growing algorithm is studied and the existing region growing algorithm is improved.The experimental results show that the direction vector formed by the seed point and the neighborhood point is used as the region growth direction,which makes the region growth algorithm more suitable for the curve skeleton.In the third part,according to the problem of how to determine the "sub-component" of the 3D model segmentation result,according to the curve skeleton,the 3D model data topology can be effectively preserved.This paper proposes a skeleton-based 3D model segmentation method.Firstly,the candidate key points of the skeleton are extracted and optimized.Then,the skeleton data is segmented by the improved region growing algorithm.Finally,the segmentation results of the original model are obtained by mapping the skeleton data with the original three-dimensional data.The experimental results show that the method can effectively improve the quality of 3D model segmentation results.