说明：  舌体轮廓的图像分割是计算机辅助舌诊系统进行图像分析的前提条件。提出了一种基于分水岭交换和主动轮廓模型相结合的自动轮廓分割算法。采用了分水岭变换的结果作为主动轮廓模型算法的初始曲线。在分水岭变换中，通过引入标记函数和强制最小值技术解决了传统分水岭变换可能导致的过分割问题，同时，在保证分割质量的基础上，通过降采样的方法来提高分割速度
(Tongue contours for image segmentation is aided tongue diagnosis system, a prerequisite for image analysis. A watershed-based exchange and active contour model by integrating the contour segmentation. Watershed transform using active contour model results as the initial curve. In the watershed transform, through the introduction of mandatory minimum marking function and technology to solve the traditional watershed transformation may lead to over-segmentation, while guaranteeing the quality of the basis of segmentation, by reducing the sampling method to improve segmentation speed)