opencv例子里没有提供cvsnakeimage的使用方法，在此整理一个例子，可以形象的看看snake算法的结果，大致做法是：



首先设定域值分割，把基本的轮廓找出来，见图中蓝色轮廓线，再将轮廓点传入cvSnakeImage函数，计算出绿色的snake轮廓线。



其中参数alpha代表点相互靠拢的权值(0-1.0)，beta表示弯曲能量（越小越容易弯曲）(0-1.0)，gamma表示整体能量(0-1.0)。其中参数我自己也不确定具体的范围，最好自己更改不同的范围试试. -opencv example does not provide cvsnakeimage way to use an example in this order, you can see the image of the results of snake algorithm, roughly approach is: first to set the threshold segmentation, the basic outline to find out, the blue outline of Figure line, then passed cvSnakeImage function of contour points to calculate the green snake contour. Parameter alpha which each representative point to move closer to the weight (0-1.0), beta that bending energy (the smaller the easier bending) (0-1.0), gamma that the overall energy (0-1.0). I am not sure which specific parameters of the scope, it is best to try different areas to change their own.