C++基于HSV和C1C2C3彩色空间的阴影检测源码+超详细注释+说明(课程大作业).zip

// ShadowDetection.cpp : 此文件包含 "main" 函数。程序执行将在此处开始并结束。 // #include <iostream> #include<opencv2/opencv.hpp> using namespace cv; using namespace std; //形态学闭运算 先膨胀后腐蚀 Mat Closure(const Mat& img){ Mat dst; //设置卷积核大小 int element_size = 1; int s = element_size * 2 + 1; //闭运算 Mat structureElement = getStructuringElement(MORPH_RECT, Size(s, s), Point(-1, -1)); dilate(img, dst, structureElement, Point(-1, -1), 1);//先膨胀 erode(dst, dst, structureElement, Point(-1, -1));//再腐蚀 return dst; } //基于HSV彩色空间的阴影检测/////////////////////////////////////////// Mat HSV(const Mat& img) { int width = img.cols; int height = img.rows; Mat dst,img_hsv; img_hsv.create(img.size(), img.type()); dst.create(img.size(), CV_8UC3); //BGR转换到HSV空间 cvtColor(img, img_hsv, COLOR_BGR2HSV);//H色调 S饱和度 V明度 double h, s, v, m; //计算指数m(-1~1)并进行阈值分割 for (int i = 0; i < height; i++) for (int j = 0; j < width; j++) { h = img_hsv.at<Vec3b>(i, j)[0]; s = img_hsv.at<Vec3b>(i, j)[1]; v = img_hsv.at<Vec3b>(i, j)[2]; m = (((s - v) / (h + s + v))); if (m > 0.3)//分割阈值设为0.3 dst.at<Vec3b>(i, j) = { 0,255,255 }; else dst.at<Vec3b>(i, j) = { 100,0,100 }; } //执行闭运算 dst = Closure(dst); imshow("阴影检测-HSV", dst); imwrite("HSV.jpg", dst); return dst; } //基于C1C2C3彩色空间的阴影检测//////////////////////////////////////// Mat C1C2C3(const Mat& img) { int width = img.cols; int height = img.rows; Mat dst,temp; dst.create(img.size(), CV_8UC3); double C3; //计算C3指数，并进行双阈值分割 for (int i = 0; i < height; i++) for (int j = 0; j < width; j++) { double R = img.at<Vec3b>(i, j)[2]; double G = img.at<Vec3b>(i, j)[1]; double B = img.at<Vec3b>(i, j)[0]; C3 = atan(B / MAX(R,G)); if (C3 > 0.4 && B < 34) //阈值设为C3>0.4和Blue<34 dst.at<Vec3b>(i, j) = { 0,255,255 }; else dst.at<Vec3b>(i, j) = { 100,0,100 }; } //执行闭运算 dst = Closure(dst); imshow("阴影检测-C1C2C3", dst); imwrite("C1C2C3.jpg", dst); return dst; } int main() { Mat srcImg = imread("第三部分 阴影检测/Color.bmp", 4);//保持原通道数原深度 HSV(srcImg); C1C2C3(srcImg); waitKey(0); imshow("test", srcImg); }