Update speed_filter2d_in_rgb_image.cpp

Author: BBuf

Algorithm optimization/README.md

@@ -1,4 +1,4 @@
-# Algorithm optimization 优化一些常见的OpenCV算法
+# Algorithm optimization 优化一些常见的OpenCV算法
 
 - speed_filter2d_in_gray_image.cpp 使用OpenBLAS和OpenMP优化对灰度图的卷积算法(filter2D)
-- 
+- speed_filter2d_in_rgb_image.cpp 使用OpenBLAS和OpenMP优化对RGB图的卷积算法(filter2D)

Algorithm optimization/speed_filter2d_in_rgb_image.cpp

@@ -0,0 +1,81 @@
+// OpenCVTest.cpp : 定义控制台应用程序的入口点。
+//
+
+#include "stdafx.h"
+
+#include <iostream>
+#include <stdio.h>
+#include <cblas.h>
+#include <opencv2/opencv.hpp>
+
+using namespace std;
+using namespace cv;
+
+//使用OpenMp和OpenBlas加速filter2D(RGB度图）,且卷积核的长宽相等，这里实现的是长宽均为3且通道数为3的卷积核(相当于Tensor)
+//A加pad的运算
+void get_Pad(int pad_Height, int pad_Width, int row, int col, int channel, float *A_pad, const float *A) {
+	int pad_x = (pad_Height - row) >> 1;
+	int pad_y = (pad_Width - col) >> 1;
+	printf("pad_x: %d pad_y: %d\n", pad_x, pad_y);
+	const int pad_one_channel = pad_Height * pad_Width;
+	const int org_one_channel = row * col;
+	for (int c = 0; c < channel; c++) {
+		for (int i = 0; i < pad_Height; i++) {
+			for (int j = 0; j < pad_Width; j++) {
+				int index = c * pad_one_channel + i * pad_Height + j;
+				if (i <= pad_x || i + pad_x > pad_Height) {
+					A_pad[index] = 0;
+				}
+				else {
+					if (j <= pad_y || j + pad_y > pad_Width) {
+						A_pad[index] = 0;
+					}
+					else {
+						A_pad[index] = A[c * org_one_channel + (i - 1) * row + j - 1];
+					}
+				}
+			}
+		}
+	}
+}
+
+
+int main() {
+	//RGB图
+	Mat src = cv::imread("F:\\1.jpg");
+	int row = src.rows;
+	int col = src.cols;
+	int channel = src.channels();
+	float *A = new float[row * col * channel];
+	int cnt = 0;
+	for (int k = 0; k < channel; k++) {
+		for (int i = 0; i < row; i++) {
+			for (int j = 0; j < col; j++) {
+				A[cnt++] = 1.0 * src.at<Vec3b>(i, j)[k];
+			}
+		}
+	}
+	const int KernelHeight = 3;
+	const int KernelWidth = 3;
+	float *B = new float[KernelHeight * KernelWidth * channel];
+	cnt = 0;
+	//这里随便赋值测试
+	for (int k = 0; k < channel; k++) {
+		for (int i = 0; i < KernelHeight; i++) {
+			for (int j = 0; j < KernelWidth; j++) {
+				B[cnt++] = k + 1;
+			}
+		}
+	}
+	//卷积核参数初始化为
+	const int pad = (KernelHeight - 1) / 2; //需要pad的长度
+	const int stride = 1; //卷积核滑动的步长
+	const int OutHeight = (row - KernelHeight + 2 * pad) / stride + 1;
+	const int OutWidth = (col - KernelWidth + 2 * pad) / stride + 1;
+	//计算3维pad_A
+	const int pad_Height = row + 2 * pad;
+	const int pad_Width = col + 2 * pad;
+	float *A_pad = new float[pad_Height * pad_Width * channel];
+	get_Pad(pad_Height, pad_Width, row, col, channel, A_pad, A);
+	//
+}