一种基于IPCM的自适应曲线检测算法

A Self-Adaptive Curve Detection Algorithm Based on IPCM

模糊球壳聚类(FCSS)算法和基于改进型可能性C-均值聚类(IPCM)的球壳聚类(IPCSS)算法都是基于梯度的交错寻优方法,在检测圆或圆弧曲线时容易陷入局部极小值,从而得到错误的检测结果,同时其不能自动识别曲线的条数.针对上述两个缺点,在IPCM的基础上用拟合法计算半径和圆心,很大程度上克服了陷入局部极小值的缺点,同时引入特征间隙的方法,实现了曲线条数的自动识别.大量数值仿真实验和实际数据实验表明,提出的算法对圆或圆弧型曲线具有良好的自适应检测效果.

Fuzzy C-spherical shell cluster（FCSS） algorithm and improved possibilistic C- spherical shell cluster（IPCSS） algorithm based on improved possibilistic C-means cluster（IPCM） are alternating optimization strategies based on gradient. Therefore, the two algorithms are easy to fall into local minimum value when detecting round or circular arc curve, and then lead to error results. In addition, they cannot identify automatically the number of the curves. According to the two weaknesses, this paper proposed a new algorithm on the basis of the IFCM. It puts to use fitting method to calculate radius and center, and overcomes the shortcoming of trapping in local minimum to a great extent. At the same time, a method of characteristics gap is introduced and implements the automatic identification of the curves. A large number of emulational experiments and actual data experiments show that the proposed algorithm for round or circular arc curve has a good adaptive detection effect.