对图像进行模板卷积运算C++

编制3X3模板对图像进行模板卷积运算程序。
要求：手动输入源图像、平均系数、卷积系数；输出卷积处理后图像和源图像。
提示：可用cvt（image,bgr2gray）函数将彩色图像转换为灰度图像后再进行处理。

原理：先将彩色图像转为灰度图像，然后依次对每个像素进行模板卷积运算。对边缘像素我采用的是不进行模板卷积运算

代码实现 ：

类声明部分

#pragma once
#include<iostream>
#include<string>

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

using namespace std;
using namespace  cv;
class TemplateConvolution
{
public:
	TemplateConvolution();
	~TemplateConvolution();
	void readImage(string filename);
	void changeToGray();
	void convolutionImage(int T[3][3]);
private:
	Mat src;
	Mat out;
	Mat gray_image;
};

类实现部分：

#include "TemplateConvolution.h"

TemplateConvolution::TemplateConvolution()
{
}

TemplateConvolution::~TemplateConvolution()
{
}

void TemplateConvolution::readImage(string filename)
{
	src = imread(filename);
}

void TemplateConvolution::changeToGray()
{
     cvtColor(src, gray_image, COLOR_BGR2GRAY); 
	 namedWindow("Gray_Image");
	 imshow("Gray_Image", gray_image);
	 waitKey(0);
}

void TemplateConvolution::convolutionImage(int T[3][3])
{
	int result = 0;
	for (int i = 0; i < 3; i++) {
		for (int j = 0; j < 3; j++) {
			result += T[i][j];
		}
	}
	if (result == 0) {
		out = gray_image.clone();
		int rows = gray_image.rows;
		int cols = gray_image.cols;
		for (int i = 0; i < rows; i++) {
			for (int j = 0; j < cols; j++) {
				if (i == 0 || i == rows - 1 || j == 0 || j == cols - 1) {
					out.at<uchar>(i, j) = gray_image.at<uchar>(i, j);
				}
				else {
					double temp = 0;
					temp = gray_image.at<uchar>(i, j) * T[1][1] + gray_image.at<uchar>(i, j - 1) * T[0][1] + gray_image.at<uchar>(i, j + 1) * T[2][1] +
						gray_image.at<uchar>(i - 1, j) * T[1][0] + gray_image.at<uchar>(i + 1, j) * T[1][2] + gray_image.at<uchar>(i - 1, j - 1) * T[0][0] +
						gray_image.at<uchar>(i - 1, j + 1) * T[2][0] + gray_image.at<uchar>(i + 1, j - 1) * T[0][2] + gray_image.at<uchar>(i + 1, j + 1) * T[2][2];
					if (int(temp + 0.5) > 255) {
						out.at<uchar>(i, j) = 255;
					}
					else out.at<uchar>(i, j) = int(temp + 0.5);
				}
			}
		}
	}
	else {
		double k = 1 / result;
		out = gray_image.clone();
		int rows = gray_image.rows;
		int cols = gray_image.cols;
		for (int i = 0; i < rows; i++) {
			for (int j = 0; j < cols; j++) {
				if (i == 0 || i == rows - 1 || j == 0 || j == cols - 1) {
					out.at<uchar>(i, j) = gray_image.at<uchar>(i, j);
				}
				else {
					double temp = 0;
					temp = gray_image.at<uchar>(i, j) * T[1][1] + gray_image.at<uchar>(i, j - 1) * T[0][1] + gray_image.at<uchar>(i, j + 1) * T[2][1] +
						gray_image.at<uchar>(i - 1, j) * T[1][0] + gray_image.at<uchar>(i + 1, j) * T[1][2] + gray_image.at<uchar>(i - 1, j - 1) * T[0][0] +
						gray_image.at<uchar>(i - 1, j + 1) * T[2][0] + gray_image.at<uchar>(i + 1, j - 1) * T[0][2] + gray_image.at<uchar>(i + 1, j + 1) * T[2][2];
					temp = temp * k;
					if (int(temp + 0.5) > 255) {
						out.at<uchar>(i, j) = 255;
					}
					else out.at<uchar>(i, j) = int(temp + 0.5);
				}
			}
		}
	}
	namedWindow("changeImage");
	imshow("changeImage", out);
	waitKey(0);
}

main函数：

#include"TemplateConvolution.h"
int main() {
	TemplateConvolution tc;
	tc.readImage("C:\\Users\\86183\\Desktop\\图片1.png");
	tc.changeToGray();
	int Sobel[3][3] = { -1,-2,-1,0,0,0,1,2,1 };
	tc.convolutionImage(Sobel);
   return 0;
}

结果展示：