图像处理（一）：灰度化及RGB2HSV

图像灰度化，即将图像从RBG三通道彩色图像转换成灰度图像，现有的方法主要有三种：

1、取最大值：即从RBG三个彩色通道中，去亮度的最大值

2、取平均值：即取RBG三个彩色通道亮度的平均值

3、取加权平均值：RBG三通道占有比例分别为0.11,0.59,0.3

下面是实现代码：

 

#include <opencv2/opencv.hpp>

using namespace cv;

int main()
{
	Mat src = imread("1.jpg");

	//1、 max
	Mat gray = Mat::zeros(src.rows, src.cols, CV_8UC1);
	for (int i = 0; i < src.rows; i++)
		for (int j = 0; j < src.cols; j++)
		{
			gray.at<uchar>(i, j) = fmax(src.at<Vec3b>(i, j)[0], 
			fmax(src.at<Vec3b>(i, j)[1], src.at<Vec3b>(i, j)[2]));
		}

	//2、 average
	Mat gray1 = Mat::zeros(src.rows, src.cols, CV_8UC1);
	for (int i = 0; i < src.rows; i++)
		for (int j = 0; j < src.cols; j++)
		{
			gray1.at<uchar>(i, j) = (src.at<Vec3b>(i, j)[0] +
				src.at<Vec3b>(i, j)[1] +  src.at<Vec3b>(i, j)[2]) / 3;
		}

	// 3、weighted mean
	Mat gray2 = Mat::zeros(src.rows, src.cols, CV_8UC1);
	for (int i = 0; i < src.rows; i++)
		for (int j = 0; j < src.cols; j++)
		{
			gray2.at<uchar>(i, j) = 0.3 * src.at<Vec3b>(i, j)[0] + 
				0.59 * src.at<Vec3b>(i, j)[1] + 0.11 * src.at<Vec3b>(i, j)[2];
		}
	namedWindow("src", CV_WINDOW_AUTOSIZE);
	namedWindow("max", CV_WINDOW_AUTOSIZE);
	namedWindow("ave", CV_WINDOW_AUTOSIZE);
	namedWindow("wei", CV_WINDOW_AUTOSIZE);

	imshow("src", src);
	imshow("max", gray);
	imshow("ave", gray1);
	imshow("wei", gray2);

	waitKey();

	return 0;



}

 

 

 

效果如下：

打完收工！

RGB颜色空间转HSV颜色空间，具体原理可以参考博客：点击打开链接，代码如下：

 

#include <opencv2/opencv.hpp>
#include <vector>
#include <iostream>

using namespace cv;
using namespace std;

int main()
{
	Mat src = imread("1.jpg");
	vector<Mat> channels;

	// 分离出三个通道
	split(src, channels);

	Mat channel_b = channels.at(0);
	Mat channel_g = channels.at(1);
	Mat channel_r = channels.at(2);

	int i, j;
	Mat channel_h = Mat::zeros(src.rows, src.cols, CV_32FC1);
	Mat channel_s = Mat::zeros(src.rows, src.cols, CV_32FC1);
	Mat channel_v = Mat::zeros(src.rows, src.cols, CV_32FC1);

	float min_v = 0.0, max_v = 0.0;
	for (i = 0; i < src.rows; i++)
		for (j = 0; j < src.cols; j++)
		{
			min_v = channel_b.at<uchar>(i, j);
			max_v = channel_b.at<uchar>(i, j);
			//最小值
            if (min_v > channel_g.at<uchar>(i, j))
			{
				min_v = channel_g.at<uchar>(i, j);
			}
			if (min_v > channel_r.at<uchar>(i, j))
			{
				min_v = channel_r.at<uchar>(i, j);
			}
			// 最大值
			if (max_v < channel_g.at<uchar>(i, j))
			{
				max_v = channel_g.at<uchar>(i, j);
			}
			if (max_v < channel_r.at<uchar>(i, j))
			{
				max_v = channel_r.at<uchar>(i, j);
			}

			channel_v.at<float>(i, j) = max_v;
			if (abs(max_v) > 0)
			{
				channel_s.at<float>(i, j) = 1.0 * (max_v - min_v) / max_v;
			}
			else
			{
				channel_s.at<float>(i, j) = 0;
			}

			if (abs(max_v - channel_b.at<uchar>(i, j)) < 0.1)
			{
				channel_h.at<float>(i, j) = 60 * (channel_g.at<uchar>(i, j) - channel_b.at<uchar>(i, j)) / (max_v - min_v);
				if (channel_h.at<float>(i, j) < 0)
				{
					channel_h.at<float>(i, j) += 360;
				}
			}
			else if (abs(max_v - channel_g.at<uchar>(i, j)) < 0.1)
			{
				channel_h.at<float>(i, j) = 120 + 60 * (channel_b.at<uchar>(i, j) - channel_r.at<uchar>(i, j)) / (max_v - min_v);
				if (channel_h.at<float>(i, j) < 0)
				{
					channel_h.at<float>(i, j) += 360;
				}

			}
			else
			{
				channel_h.at<float>(i, j) = 240 + 60 * (channel_r.at<uchar>(i, j) - channel_g.at<uchar>(i, j)) / (max_v - min_v);	
				if (channel_h.at<float>(i, j) < 0)
				{
					channel_h.at<float>(i, j) += 360;
				}
			}
		}


	imshow("H", channel_h);
	imshow("S", channel_s);
	imshow("V", channel_v);

	waitKey();

	return 0;

}

效果如下：

 

 

其中，H分量对应色相（RBG三个通道去最大值），S分量对应饱和度（取值范围0-1），V分量对应明度。

注意一下，下面这个点可能非常有用：

H是色彩

S是深浅， S = 0时，只有灰度

V是明暗，表示色彩的明亮程度，但与光强无直接联系

参考资料：

http://lib.csdn.net/article/opencv/23601