实现功能：
C++语言实现纯高斯模糊处理灰度图像，不受图片格式限制
算法实现：
/// <summary>
/// 程序功能：c语言实现纯高斯模糊处理灰度图像
/// 系统win7，VS2010开发环境,编程语言C++,OpenCV2.4.7最新整理时间 whd 2016.9.9。
/// </summary>
/// <param name=" pixels">源图像数据在内存的起始地址。</param>
/// <param name="width">源和目标图像的宽度。</param>
/// <param name="height">源和目标图像的高度。</param>
/// <param name=" channels">通道数，灰度图像cn=1,彩色图像cn=3</param>
/// <param name="sigma">sigma的平方是高斯函数的方差</param>
/// <remarks> 1: 能处理8位灰度和24位图像。需要分开进行，后面会合成一个程序</remarks>
// 以下为参考函数实现的整个过程
//（1）建立工程，复制粘贴博客代码。
// (2) 添加malloc()和free()函数的头文件
// (3) exp()函数的头文件
// (4) 修改Gasussblur中形参int sigma为float sigma，更加符合实际情况
// (5) 配置OpenCV
// (6) 调用函数
#include "stdafx.h"
#include<stdlib.h> //malloc(),free()函数需要的头文件
#include<math.h>
#include<windows.h> //包含时钟头文件
#include <opencv2/opencv.hpp>
using namespace std;
using namespace cv;
inline int* buildGaussKern(int winSize, int sigma)
{
int wincenter, x;
float sum = 0.0f;
wincenter = winSize / 2;
float *kern = (float*)malloc(winSize*sizeof(float));
int *ikern = (int*)malloc(winSize*sizeof(int));
float SQRT_2PI = 2.506628274631f;
float sigmaMul2PI = 1.0f / (sigma * SQRT_2PI);
float divSigmaPow2 = 1.0f / (2.0f * sigma * sigma);
for (x = 0; x < wincenter + 1; x++)
{
kern[wincenter - x] = kern[wincenter + x] = exp(-(x * x)* divSigmaPow2) * sigmaMul2PI;
sum += kern[wincenter - x] + ((x != 0) ? kern[wincenter + x] : 0.0);
}
sum = 1.0f / sum;
for (x = 0; x < winSize; x++)
{
kern[x] *= sum;
ikern[x] = kern[x] * 256.0f;
}
free(kern);
return ikern;
}
void GaussBlur(unsigned char* pixels, unsigned int width, unsigned int height, unsigned int channels, float sigma)
{
width = 3 * width;
if ((width % 4) != 0) width += (4 - (width % 4));
unsigned int winsize = (1 + (((int)ceil(3 * sigma)) * 2));
int *gaussKern = buildGaussKern(winsize, sigma);
winsize *= 3;
unsigned int halfsize = winsize / 2;
unsigned char *tmpBuffer = (unsigned char*)malloc(width * height* sizeof(unsigned char));
for (unsigned int h = 0; h < height; h++)
{
unsigned int rowWidth = h * width;
for (unsigned int w = 0; w < width; w += channels)
{
unsigned int rowR = 0;
unsigned int rowG = 0;
unsigned int rowB = 0;
int * gaussKernPtr = gaussKern;
int whalfsize = w + width - halfsize;
unsigned int curPos = rowWidth + w;
for (unsigned int k = 1; k < winsize; k += channels)
{
unsigned int pos = rowWidth + ((k + whalfsize) % width);
int fkern = *gaussKernPtr++;
rowR += (pixels[pos] * fkern);
rowG += (pixels[pos + 1] * fkern);
rowB += (pixels[pos + 2] * fkern);
}
tmpBuffer[curPos] = ((unsigned char)(rowR >> 8));
tmpBuffer[curPos + 1] = ((unsigned char)(rowG >> 8));
tmpBuffer[curPos + 2] = ((unsigned char)(rowB >> 8));
}
}
winsize /= 3;
halfsize = winsize / 2;
for (unsigned int w = 0; w < width; w++)
{
for (unsigned int h = 0; h < height; h++)
{
unsigned int col_all = 0;
int hhalfsize = h + height - halfsize;
for (unsigned int k = 0; k < winsize; k++)
{
col_all += tmpBuffer[((k + hhalfsize) % height)* width + w] * gaussKern[k];
}
pixels[h * width + w] = (unsigned char)(col_all >> 8);
}
}
free(tmpBuffer);
free(gaussKern);
}
void GaussBlur1D(unsigned char* pixels,unsigned char* pixelsout, unsigned int width, unsigned int height, float sigma) //删掉unsigned int channels,因为是单通道没有用
{
width = 1 * width; //3修改为1，因为三个通道变为了1个通道，存储每行数据的宽度变为了原来的1/3.
if ((width % 4) != 0) width += (4 - (width % 4));
unsigned int winsize = (1 + (((int)ceil(3 * sigma)) * 2)); //窗的大小
int *gaussKern = buildGaussKern(winsize, sigma); //构建高斯核，计算高斯系数
winsize *= 1; //3改为1，高斯窗的宽度变为原来的1/3
unsigned int halfsize = winsize / 2; //窗的边到中心的距离
unsigned char *tmpBuffer = (unsigned char*)malloc(width * height* sizeof(unsigned char)); //开辟新的内存存储处理高斯模糊后的数据
for (unsigned int h = 0; h < height; h++) //外层循环，图像的高度
{
unsigned int rowWidth = h * width; //当前行的宽度为图像的高度乘以每行图像的数据所占的宽度。因为是按行存储的数组。
for (unsigned int w = 0; w < width; w++) //w+=channels，可以修改为w++，因为是单通道数据，而不是三通道数据
{
unsigned int rowR = 0; //存储r分量的数据
int * gaussKernPtr = gaussKern;//将高斯系数赋值给gaussKernPtr
int whalfsize = w + width - halfsize;
unsigned int curPos = rowWidth + w; //当前位置
for (unsigned int k = 1; k < winsize;k++) // k += channels修改为k++
{
unsigned int pos = rowWidth + ((k + whalfsize) % width);
int fkern = *gaussKernPtr++;
rowR += (pixels[pos] * fkern); //当前像素值乘以高斯系数，rowR这了泛指单通道的当前像素点高斯处理后的数
}
tmpBuffer[curPos] = ((unsigned char)(rowR >> 8)); //除以256
}
}
halfsize = winsize / 2;
for (unsigned int w = 0; w < width; w++)
{
for (unsigned int h = 0; h < height; h++)
{
unsigned int col_all = 0;
int hhalfsize = h + height - halfsize;
for (unsigned int k = 0; k < winsize; k++)
{
col_all += tmpBuffer[((k + hhalfsize) % height)* width + w] * gaussKern[k];
}
pixelsout[h * width + w] = (unsigned char)(col_all >> 8);
}
}
free(tmpBuffer);
free(gaussKern);
}
int _tmain(int argc, _TCHAR* argv[])
{
const char* imagename = "C:\\Users\\Administrator.IES7LSEJAZ1GGRL\\Desktop\\PureGaussian-master\\GaussianBlur\\GaussianBlur\\InputName.bmp";
//从文件中读入图像
Mat img = imread(imagename);
Mat dst = imread(imagename);
Mat gray_img;
Mat gray_dst;
cvtColor(img, gray_img, CV_BGR2GRAY);
cvtColor(dst, gray_dst, CV_BGR2GRAY);
//如果读入图像失败
if(img.empty())
{
fprintf(stderr, "Can not load image %s\n", imagename);
return -1;
}
LARGE_INTEGER m_nFreq;
LARGE_INTEGER m_nBeginTime;
LARGE_INTEGER nEndTime;
QueryPerformanceFrequency(&m_nFreq); // 获取时钟周期
QueryPerformanceCounter(&m_nBeginTime); // 获取时钟计数
GaussBlur1D(gray_img.data,gray_dst.data,gray_img.cols,gray_img.rows,2);
QueryPerformanceCounter(&nEndTime);
cout << (nEndTime.QuadPart-m_nBeginTime.QuadPart)*100/m_nFreq.QuadPart << endl;
//显示图像
imshow("原图像",gray_img);
imshow("模糊图像", gray_dst);
//此函数等待按键，按键盘任意键就返回
waitKey();
return 0;
}
算法实现效果：sigma=2.0
