A case of skin tone detection-using the RGB color model

Code:

/*input: RGB image output: Grayscale image with the same size as the input image, if a pixel in the RGB image is detected as skin color, the corresponding pixel in the grayscale map is 255, otherwise 0*/voidSkinrgb (Mat &rgb,mat &DST) {Assert (Rgb.channels ()==3 ); Const intR =2; Const intG =1; Const intB =0; DST=Mat::zeros (Rgb.rows, Rgb.cols, cv_8u);  for(inti =0; i < rgb.rows; i++) {vec3b* ptr = rgb.ptr<vec3b>(i); Uchar* Ptr_mask = dst.ptr<uchar>(i);  for(intj =0; J < Rgb.cols; J + +) {vec3b pixel=Ptr[j]; if((Pixel[r) > the&& pixel[g]> +&& Pixel[b] > -&&Pixel[r]-Pixel[b]> the&& Pixel[r]-pixel[g]> the) ||(Pixel[r]> $&& pixel[g]> About&& pixel[b]> the&&ABS (PIXEL[R)-Pixel[b]) <= the&& Pixel[r]>pixel[b] && pixel[g]>Pixel[b])) {Ptr_mask[j]=255; }        }    }}

Effect:

A case of skin tone detection-using the RGB color model