20180628例程一解释 coherence_enhancing_diff.hdev

----------------------------------------------------------------------------------------------------------
* This example program shows how coherence_enhancing_diff can be used
* to enhance a fingerprint image by joining disconnected parts of the
* fingerprint lines.
read_image (Image, 'fingerprint')
get_image_size (Image, Width, Height)
dev_close_window ()
dev_open_window (0, 0, Width * 2, Height * 2, 'black', WindowHandle)
dev_display (Image)
stop ()
coherence_enhancing_diff (Image, ImageCED, 0.5, 4, 0.5, 50)
dev_display (ImageCED)
----------------------------------------------------------------------------------------------------------
重点算子：

coherence_enhancing_diff (Operator)
执行图像的相干增强扩散。
请注意，如果将具有缩小域的图像用作输入，则筛选器运算符可能会返回意外的结果。
coherence_enhancing_diff(Image : ImageCED : Sigma, Rho, Theta, Iterations : )

Sigma：平滑倒数运算符（>=0）(建议值0,0.1,0.5,1)
在这个例程里，小于0.5会在左下角出有轻微模糊情况，如下图

Rho：平滑扩散系数（>=0）(建议值0,1,3,5,10,30)

Theta：时间步骤（>0 <=0.5）(建议值0.1,0.2,0.3,0.4,0.5)

Iterations：迭代次数（>=1）(建议值1,5,10,20,50,100,500)