Minimizing the computing complexity of iterative sequential thinning algorithm
Yun Hai Xia · 2003
The sequential thinning method proposed, based on the grassfire model, overcomes the two essential difficulties of thinning. First, using contour tracing, thinning can be fulfilled in less than two scans of image. Due to the notion of associated point defined, fire propagation and quenching are simulated in discrete space. This simulation makes it possible to find quench points without neighborhood testing, and the computing complexity is almost minimized. Second, protrusion analysis makes the method capable of eliminating nonsignificant branches of skeleton. Moreover, like median-axis transformation, the obtained skeletons are labelled by quench function. Thus the method is suitable for practical applications.>