opencv计算信息熵（c++代码）

【核心代码】

// first.cpp : 定义控制台应用程序的入口点。
//

#include "stdafx.h"
#include <opencv2/opencv.hpp>
using namespace cv;
double Entropy(Mat img);

int _tmain(int argc, _TCHAR* argv[])
{	
	String name = "chair_0022_Area01_001.png";
	Mat src_color = imread("picture/1/"   name);//读取原彩色图  
	/*int c = src_color.cols ;
	int r = src_color.rows ;
	int tt = src_color.channels();*/
	Mat src_gray;//彩色图像转化成灰度图  
	cvtColor(src_color, src_gray, COLOR_BGR2GRAY);
	imwrite("picture/Gray/Gray_"   name, src_gray);//保存图像文件

	//Mat img = imread("gray.png");
	double x = Entropy(src_color);
	std::cout << x << std::endl;
	std::cout << src_color.cols << std::endl;
	std::cout << src_color.rows << std::endl;
	std::cout << src_color.channels() << std::endl;

	double x1 = Entropy(src_gray);
	std::cout << x1 << std::endl;
	std::cout << src_gray.cols << std::endl;
	std::cout << src_gray.rows << std::endl;
	std::cout << src_gray.channels() << std::endl;

	system("pause");
	return 0;
}


double Entropy(Mat img)
{
	//将输入的矩阵为图像
	double temp[256];
	/*清零*/
	for (int i = 0; i < 256; i  )
	{
		temp[i] = 0.0;
	}
	/*计算每个像素的累积值*/
	for (int m = 0; m < img.rows; m  )
	{
		const uchar* t = img.ptr<uchar>(m);
		for (int n = 0; n < img.cols; n  )
		{
			int i = t[n];
			temp[i] = temp[i]   1;
		}
	}
	/*计算每个像素的概率*/
	for (int i = 0; i < 256; i  )
	{
		temp[i] = temp[i] / (img.rows*img.cols);
	}
	double result = 0;
	/*根据定义计算图像熵*/
	for (int i = 0; i < 256; i  )
	{
		if (temp[i] == 0.0)
			result = result;
		else
			result = result - temp[i] * (log(temp[i]) / log(2.0));
	}
	return result;
}


void calc_2D_entropy(cv::Mat &input, cv::Mat &output){
	int height = input.rows;
	int width = input.cols;

	cv::Mat out = cv::Mat::zeros(height, width, CV_32FC1);

	//template size
	int w = 3;

	for (int i = w; i < height - w; i  )
	{
		float *data = out.ptr<float>(i);
		for (int j = w; j < width - w; j  )
		{
			//cv::Mat Hist = cv::Mat::zeros(1, 256, CV_32F);
			float Hist[256] = { 0 };
			for (int p = i - w; p < i   w   1; p  )
			{
				uchar *t = input.ptr<uchar>(p);
				for (int q = j - w; q < j   w   1; q  )
				{
					int tmp = t[q];
					//cout << "tmp:" << tmp << endl;
					Hist[tmp] = Hist[tmp]   1;
				}

			}

			float sumHist = 0;
			for (int ii = 0; ii < 256; ii  )
			{

				sumHist  = Hist[ii];
			}

			//get the probality
			for (int ii = 0; ii < 256; ii  )
			{
				Hist[ii] = Hist[ii] / sumHist;
				//if (Hist[ii] != 0)
				//  cout << ii << ":" << Hist[ii] << endl;
			}

			//calculate the entropy
			for (int k = 0; k < 256; k  )
			{
				float v = Hist[k];
				float z = data[j];
				//cout << "z:" << z << endl;
				if (v != 0)
				{
					double H = v * (log(v) / (float)log(2.0));
					//H = H * 80.5 - 1;
					data[j] = data[j] - H;
					//data[j] = data[j]   v * log(1 / v);
					//cout << j << ":" << data[j] << endl;
				}
			}
		}

	}

	normalize(out, output);
	output = output * 255;

}