opencv大津算法二值化

二值化的前提是灰度化，手动二值化需要不断进行调整，找到合适的阈值，大于阈值为白色。
大津二值化又叫最大类间方差法，可以自动确定二值化的阈值。基本思想是遍历阈值，当某个阈值下分割的两类像素点之间灰度方差最大的时候，就是最优的阈值。（因为类间方差越大，两类的差别越大，不管是前景错分到背景还是背景错分到前景都会使类间方差变小）。数学推导如下：

#include <opencv2/core.hpp>
#include <opencv2/highgui.hpp>
#include <iostream>

cv::Mat BGR2GRAY(cv::Mat img) {
	int width = img.cols;
	int height = img.rows;

	cv::Mat out = cv::Mat::zeros(height, width, CV_8UC1);//8位无符号灰度图像

	for (int y = 0; y < height; y++) {
		for (int x = 0; x < width; x++) {
			//RGB分量取不同的权重
			out.at<uchar>(y, x) = 0.2126 *img.at<cv::Vec3b>(y, x)[2] \
				+ 0.7152 * img.at<cv::Vec3b>(y, x)[1] \
				+ 0.0722 * img.at<cv::Vec3b>(y, x)[0];
		}
	}
	return out;
}

cv::Mat Binarize_Otsu(cv::Mat gray) {
	int width = gray.cols;
	int height = gray.rows;

	double w0 = 0, w1 = 0;//像素的比例
	double m0 = 0, m1 = 0;//像素的均值
	double max_sb = 0, sb = 0;
	int th = 0;
	int val;

	for (int t = 0; t < 255; t++) {
		w0 = 0;
		w1 = 0;
		m0 = 0;
		m1 = 0;
		//获取各类像素的个数 计算均值
		for (int y = 0; y < height; y++) {
			for (int x = 0; x < width; x++) {
				val = (int)(gray.at<uchar>(y, x));

				if (val < t) {
					w0++;
					m0 += val;
				}
				else {
					w1++;
					m1 += val;
				}
			}
		}
		m0 /= w0;
		m1 /= w1;
		w0 /= (height * width);
		w1 /= (height * width);
		sb = w0 * w1 * pow((m0 - m1), 2);//计算类间方差

		if (sb > max_sb) {
			max_sb = sb;
			th = t;
		}
	}

	std::cout << "大津算法阈值:" << th << std::endl;

	cv::Mat out = cv::Mat::zeros(height, width, CV_8UC1);

	for (int y = 0; y < height; y++) {
		for (int x = 0; x < width; x++) {
			// Binarize
			if (gray.at<uchar>(y, x) > th) {
				out.at<uchar>(y, x) = 255;
			}
			else {
				out.at<uchar>(y, x) = 0;
			}

		}
	}

	return out;
}

int main(int argc, const char* argv[]) {
	
	cv::Mat img = cv::imread("C:/Users/zxdn/Desktop/girl.jpg", cv::IMREAD_COLOR);
	cv::Mat gray = BGR2GRAY(img);
	cv::Mat out = Binarize_Otsu(gray);
	cv::imshow("原图", img);
	cv::imshow("大津二值化", out);
	cv::waitKey(0);
	cv::destroyAllWindows();

	return 0;
}