文章目录
- 相关链接
- 前言
- 阈值
- 阈值使用
- 代码
- Python
- C++
- Csharp
- csharp代码问题
- 总结
相关链接
C++&Python&Csharp in OpenCV 专栏
【2022B站最好的OpenCV课程推荐】OpenCV从入门到实战 全套课程（附带课程课件资料+课件笔记）
OpenCV一个窗口同时显示多张图片 (C++)
前言
我感觉我之前的比较实在是太过于偏代码了，缺少了图形学的知识的记录。之后的笔记我会把图形学的知识也记录上去。
阈值
阈值简单来说就是增强图片的对比。用过Ps的人就知道阈值是什么意思，在电脑绘画中用的还是比较多的
阈值使用
代码
Python
# %%
# 导入包
import cv2
import matplotlib.pyplot as plt
import numpy as np
image = {
"image":'',
"gray":'',
}
image["image"] = cv2.imread("D:/workSpace/OpenCV/HellOpenCV/Resources/image/cat.png")
# 获取颜色通道，因为cv2是BGR的顺序，所以这里是B 灰度
image["gray"]= cv2.split(image['image'])[0]
thresh = {
'1' : '','2': '','3': '','4': '','5': ''
}
# 调用不同的阈值处理算子
ret , thresh['1'] = cv2.threshold(image['gray'],127,255,cv2.THRESH_BINARY)
ret , thresh['2'] = cv2.threshold(image['gray'],127,255,cv2.THRESH_BINARY_INV)
ret , thresh['3'] = cv2.threshold(image['gray'],127,255,cv2.THRESH_TRUNC)
ret , thresh['4'] = cv2.threshold(image['gray'],127,255,cv2.THRESH_TOZERO)
ret , thresh['5'] = cv2.threshold(image['gray'],127,255,cv2.THRESH_TOZERO_INV)
# 将数据总和成数组
titles = ['iamge','binary','binary inv','trunc','tozero','tozero inv']
images = [image['gray'],thresh['1'],thresh['2'],thresh['3'],thresh['4'],thresh['5']]
# 循环赋值到plt中
for i in range(6):
plt.subplot(2,3,i+1),plt.imshow(images[i],'gray')
plt.title(titles[i])
plt.xticks([]),plt.yticks([])
plt.show()
C++
这里为了一次性展示多个窗口，我就CSDN上面找了个方法
OpenCV一个窗口同时显示多张图片 (C++)
#include <opencv2/opencv.hpp>
#include <opencv2/core/core.hpp>
#include <opencv2/highgui/highgui.hpp>
#include <opencv2/imgproc.hpp>
#include<iostream>
#include <vector>
using namespace std;
using namespace cv;
vector<Mat> imageVector;
void multipleImage(vector<Mat> imgVector, Mat& dst, int imgCols)
{
const int MAX_PIXEL = 300;
int imgNum = imgVector.size();
//选择图片最大的一边 将最大的边按比例变为300像素
Size imgOriSize = imgVector[0].size();
int imgMaxPixel = max(imgOriSize.height, imgOriSize.width);
//获取最大像素变为MAX_PIXEL的比例因子
double prop = imgMaxPixel < MAX_PIXEL ? (double)imgMaxPixel / MAX_PIXEL : MAX_PIXEL / (double)imgMaxPixel;
Size imgStdSize(imgOriSize.width * prop, imgOriSize.height * prop); //窗口显示的标准图像的Size
Mat imgStd; //标准图片
Point2i location(0, 0); //坐标点(从0,0开始)
//构建窗口大小 通道与imageVector[0]的通道一样
Mat imgWindow(imgStdSize.height * ((imgNum - 1) / imgCols + 1), imgStdSize.width * imgCols, imgVector[0].type());
for (int i = 0; i < imgNum; i++)
{
location.x = (i % imgCols) * imgStdSize.width;
location.y = (i / imgCols) * imgStdSize.height;
resize(imgVector[i], imgStd, imgStdSize, prop, prop, INTER_LINEAR); //设置为标准大小
imgStd.copyTo(imgWindow(Rect(location, imgStdSize)));
}
dst = imgWindow;
}
int main()
{
auto image = imread("D:/workSpace/OpenCV/HellOpenCV/Resources/image/cat.png");
Mat showImage;
Mat bgr[3];
split(image, bgr);
Mat gray = bgr[0];
Mat thresh[5];
threshold(gray, thresh[0], 127, 255, THRESH_BINARY);
threshold(gray, thresh[1], 127, 255, THRESH_BINARY_INV);
threshold(gray, thresh[2], 127, 255, THRESH_TRUNC);
threshold(gray, thresh[3], 127, 255, THRESH_TOZERO);
threshold(gray, thresh[4], 127, 255, THRESH_TOZERO_INV);
vector<Mat> images{gray,thresh[0],thresh[1], thresh[2], thresh[3], thresh[4], };
multipleImage(images, showImage, 3);
imshow("C++", showImage);
waitKey(0);
destroyAllWindows();
return 0;
}
Csharp
还是那句话，C++跑通了，Csharp抄抄就可以了。
using OpenCvSharp;
using System;
using System.Collections.Generic;
using System.Linq;
using System.Numerics;
using System.Security.Cryptography;
using System.Text;
using System.Threading.Tasks;
namespace HelloOpenCV.Utils
{
public static class MyOpenCV_Extensions
{
/// <summary>
/// 3通道遍历
/// </summary>
/// <param name="mat"></param>
/// <returns></returns>
public static int[,,] MatToArray(Mat mat)
{
var res = new int[mat.Rows, mat.Cols, mat.Channels()];
for (var i = 0; i < mat.Rows; i++)
{
for (var j = 0; j < mat.Cols; j++)
{
var temp = mat.At<Vec3b>(i, j);
res[i, j, 0] = temp[0];
res[i, j, 1] = temp[1];
res[i, j, 2] = temp[2];
}
}
return res;
}
public static Mat MultipleImage(List<Mat> lists, int imgCols)
{
const int MAX_PIXEL = 300;
int imgNum = lists.Count;
//选择图片最大的一边 将最大的边按比例变为300像素
Size imgOriSize = lists[0].Size();
int imgMaxPixel = Math.Max(imgOriSize.Height, imgOriSize.Width);
//获取最大像素变为MAX_PIXEL的比例因子
double prop = imgMaxPixel < MAX_PIXEL ? (double)imgMaxPixel / MAX_PIXEL : MAX_PIXEL / (double)imgMaxPixel;
Size imgStdSize= new Size(imgOriSize.Width* prop, imgOriSize.Height* prop); //窗口显示的标准图像的Size
Mat imgStd = new Mat(); //标准图片
Point location = new Point(0, 0); //坐标点(从0,0开始)
//构建窗口大小 通道与imageVector[0]的通道一样
Mat imgWindow = new Mat(imgStdSize.Height* ((imgNum -1) / imgCols + 1), imgStdSize.Width* imgCols, lists[0].Type());
for (int i = 0; i < imgNum; i++)
{
location.X = (i % imgCols) * imgStdSize.Width;
location.Y = (i / imgCols) * imgStdSize.Height;
Cv2.Resize(lists[i], imgStd, imgStdSize, prop, prop, InterpolationFlags.Linear); //设置为标准大小
imgStd.CopyTo(new Mat(imgWindow, new Rect(location, imgStdSize)) );
}
return imgWindow;
}
}
}
运行代码
static void Main(string[] args)
{
Mat image = Cv2.ImRead("D:/workSpace/OpenCV/HellOpenCV/Resources/image/cat.png");
Mat[] bgr = new Mat[3];
Cv2.Split(image, out bgr);
Mat gray = bgr[0];
Mat[] threshs = new Mat[5];
for(var i = 0; i < threshs.Length; i++)
{
threshs[i] = new Mat();
}
Cv2.Threshold(gray, threshs[0], 127, 255, ThresholdTypes.Binary);
Cv2.Threshold(gray, threshs[1], 127, 255, ThresholdTypes.BinaryInv);
Cv2.Threshold(gray, threshs[2], 127, 255, ThresholdTypes.Trunc);
Cv2.Threshold(gray, threshs[3], 127, 255, ThresholdTypes.Tozero);
Cv2.Threshold(gray, threshs[4], 127, 255, ThresholdTypes.TozeroInv);
var res = MyOpenCV_Extensions.MultipleImage(new List<Mat>() {
gray, threshs[0], threshs[1], threshs[2], threshs[3], threshs[4]
}, 3);
Cv2.ImShow("Csharp", res);
Cv2.WaitKey(0);
Cv2.DestroyAllWindows();
}
csharp代码问题
遇到了几个坑
C++是能直接把变量当做构造函数的，C# 对应的代码是构造函数里面带Mat参数的
而且C++的构造函数还不好定位，我进去找不到源码。我不知道C++ 的代码是怎么调试的，毕竟没怎么接触过
总结
跟之前的套路差不多，Python跑通，C++ 翻译Python+上网查资料，Csharp翻译C++。我还是觉得Csharp优雅多了，写起来舒服，C++ 这个语法用起来有点痛苦。而且C++ 报错天天给我报内存变量报错，我哪里能看得懂啊，顶多定位到哪行代码。
