Skeleton revision algorithm using maximal circles
Tsuyoshi Nakamura, Hiroshi ENOWAKI, Hideaki Itoh, Lifeng He · 2002
Thinning is a fundamental concept of computer processing of line images, and it plays an important role in pattern recognition. Most existing thinning algorithms have the problem to produce distorted skeletons. Some methods to correct the distorted skeleton have been proposed, but non of them produce complete results. In order to solve the problem, we propose an improved thinning algorithm which is a skeleton revision algorithm using maximal circles (SRAMCs). SRAMCs is a hybrid algorithm and the processing route is divided into three stages: 1) thinning using Hilditch algorithm; 2) removal of spurious pixels using maximal circle; and 3) skeleton reconstruction. The results of experiments indicates that SRAMCs can be useful for detection of lines and curves in digitized line images.