摘要
传统B-snake模型以整体迭代为基础,已经逼近到目标的轮廓曲线仍要参加迭代计算,影响轮廓曲线的收敛速度。提出了一种改进的B-snake模型——分段迭代B-snake模型,它以分段迭代为基础,在迭代过程中只计算尚未逼近的轮廓曲线,直至所有的轮廓曲线均逼近目标轮廓。实验证明,与传统B-snake模型相比,分段迭代B-snake模型可以快速收敛到目标轮廓。
Traditional B-snake is on the basis of the integral iteration. All of the data are calculated on each iteration steps, including the approximated. It has a large influence on the speed of approximating to the target curve. An improved Bsnake model, the segmented B-snake, is presented in this paper, which is base upon the segmented iteration. In this model, only the unapproximated segments will take part in the later iteration computing until all segments are approximated to the target contour. Experiments indicate that the improved B-snake model has a faster speed of convergence to the target contour than the traditional.
出处
《四川理工学院学报（自然科学版）》
CAS
2009年第5期96-100,共5页
Journal of Sichuan University of Science & Engineering(Natural Science Edition)
关键词
分段迭代
B-SNAKE
轮廓提取
整体迭代
the local iteration
B-snake
contour extraction
the integral iteration
