基于区域划分拓扑分析的GVF骨架线提取算法研究  

Research on GVF Skeletonization Extraction Algorithm Based on Topological Analysis of Regions Division

作　　者：原凤英[1] 蔡元学 殷大卫 朱昊 明成国 秦月婷[1] 

机构地区：[1]天津科技大学理学院光子学与先进传感技术实验室,天津

出　　处：《光电子》2021年第3期125-131,共7页Optoelectronics

摘　　要：本文的研究目的主要是针对传统骨架线提取算法在提取时容易出现精度低、分叉等问题而提出的一种新型区域划分拓扑分析算法。该算法在提取骨架线时采用区域划分的方法,在每个区域上进行拓扑提取,一旦出现断点噪点的矛盾问题便再次划分,直至得到理想的骨架线。通过对比发现,利用这种新型的区域划分拓扑分析算法得到了比传统算法更为完整清晰的骨架线,为后续研究提供了有力保障。The purpose of this paper is to propose a new region division topology analysis algorithm to solve the problems of low precision and bifurcation in traditional skeleton extraction algorithm. When extracting the skeleton lines, the algorithm uses the method of region division, and extracts the topology in each region. Once the contradiction problem of breakpoint noise occurs, it divides again until the ideal skeleton line is obtained. Through the comparison, it is found that this new algorithm can get more complete and clear skeleton lines than the traditional algorithm, which provides a more powerful guarantee for the follow-up research.

关 键 词：SNAKES模型 GVF算法 拓扑分析 区域划分 

分 类 号：TM7[电气工程—电力系统及自动化]