https://blog.csdn.net/qq_37124237/article/details/82183177
void cv::Sobel( InputArray _src, OutputArray _dst, int ddepth, int dx, int dy, int ksize, double scale, double delta, int borderType )
| src | 输入图像 |
| dst | 输出图像 the same size and the same number of channels as src . |
| ddepth | 输出图像深度，输入图像为CV_8U，sobel的结果一般为CV_16S |
| dx | order of the derivative x.：x导数的阶数？？？（只能0、1、2） |
| dy | order of the derivative y.：y导数的阶数？？？ |
| ksize | size of the extended Sobel kernel; it must be 1, 3, 5, or 7. ：差分算子的核大小 |
| scale | optional scale factor for the computed derivative values; by default, no scaling is applied ：导数结果的比例系数 |
| delta | optional delta value that is added to the results prior to storing them in dst.：偏移值 |
| borderType | pixel extrapolation method, see cv::BorderTypes |
说明：
1、Sobel方法：离散性差分算子
原文：https://www.cnblogs.com/wllwqdeai/p/11761344.html