一维码二维码识别（opencv c++)

最新推荐文章于 2025-04-17 11:57:38 发布

Start hungry~

最新推荐文章于 2025-04-17 11:57:38 发布

阅读量3.9k

点赞数 1

分类专栏： opencv 文章标签： opencv c++ 边缘检测图像识别

原文链接：https://blog.youkuaiyun.com/dcrmg/article/details/52194896?ops_request_misc=%257B%2522request%255Fid%2522%253A%2522161190784316780255272641%2522%252C%2522scm%2522%253A%252220140713.130102334.pc%255Fall.%2522%257D&request_id=161190784316780255272641&biz_id=0&utm

版权

opencv 专栏收录该内容

3 篇文章

订阅专栏

一维码识别

参考资料：（侵权删）
1、https://blog.youkuaiyun.com/qq_40908493/article/details/79829226?ops_request_misc=%25257B%252522request%25255Fid%252522%25253A%252522161188879716780274186320%252522%25252C%252522scm%252522%25253A%25252220140713.130102334.pc%25255Fall.%252522%25257D&request_id=161188879716780274186320&biz_id=0&utm_medium=distribute.pc_search_result.none-task-blog-2_allfirst_rank_v2~rank_v29-1-79829226.pc_search_result_cache&utm_term=vs2013+opencv+2.4.9%25E5%25AE%2589%25E8%25A3%2585zbar
2、https://blog.youkuaiyun.com/dcrmg/article/details/52194896?ops_request_misc=%257B%2522request%255Fid%2522%253A%2522161190784316780255272641%2522%252C%2522scm%2522%253A%252220140713.130102334.pc%255Fall.%2522%257D&request_id=161190784316780255272641&biz_id=0&utm_medium=distribute.pc_search_result.none-task-blog-2_allfirst_rank_v2~rank_v29-2-52194896.pc_search_result_cache&utm_term=opencv%2Bzbar二维码识别（一维码校正）&spm=1018.2226.3001.4449

zbar.h可以自动扫描识别一维码二维码，不需要校正也可识别。在参考资料的基础上修改了一些代码段及添加了注释，基本原理相同。

1、安装zbar.h后简单测试（如何安装配置参考资料1）

#include"opencv2/highgui/highgui.hpp"
#include "zbar.h" 
#include <iostream> 

using namespace std;
using namespace zbar; //添加zbar名称空间
using namespace cv;

double Point2PointDist(Point2f& a, Point2f& b);
double Line_Angle(Point2f& a, Point2f& b);
int main(int argc, char*argv[])
{
	ImageScanner scanner;	//构造扫描器ImageScanner对象，并使用set_config进行初始化
	scanner.set_config(ZBAR_NONE, ZBAR_CFG_ENABLE, 1);

	//加载灰度图并显示
	Mat imageGray = imread("6.png",0);	//小于0或不写则加载原图，等于0则加载灰度图，大于0则加载RGB图像
	if (!imageGray.data)
	{
		cout << "请确认图片是否存在" << endl;
		system("pause");
		return 0;
	}
	//imshow("原图", image);
	//Mat imageGray;
	//cvtColor(image, imageGray, CV_RGB2GRAY);	//转换为灰度图
	imshow("灰度图", imageGray);

	//扫描并显示信息
	int width = imageGray.cols;
	int height = imageGray.rows;
	uchar *raw = (uchar *)imageGray.data;	
	Image imageZbar(width, height, "Y800", raw, width * height);//y800实际为灰度图代号
	scanner.scan(imageZbar); //扫描条码
	Image::SymbolIterator symbol = imageZbar.symbol_begin();
	if (imageZbar.symbol_begin() == imageZbar.symbol_end())
	{
		cout << "查询条码失败，请检查图片！" << endl;
	}
	for (; symbol != imageZbar.symbol_end(); ++symbol)
	{
		cout << "类型：" << endl << symbol->get_type_name() << endl << endl;
		cout << "条码：" << endl << symbol->get_data() << endl << endl;
	}
	waitKey();
	imageZbar.set_data(NULL, 0);
	return 0;
}

结果图示：
在这里插入图片描述

2、（实际上条码有一定角度也能扫描出来）将一维码校正后再次扫描显示信息

#include "opencv2/core/core.hpp"
#include"opencv2/highgui/highgui.hpp"
#include "opencv2/imgproc/imgproc.hpp"
#include "zbar.h" 
#include <iostream> 

using namespace std;
using namespace zbar; //添加zbar名称空间
using namespace cv;

double Point2PointDist(Point2f& a, Point2f& b);	//用于计算频谱图直线距离
double Line_Angle(Point2f& a, Point2f& b);//计算直线角度

int main(int argc, char*argv[])
{
	ImageScanner scanner;	//构造扫描器ImageScanner对象，并使用set_config进行初始化
	scanner.set_config(ZBAR_NONE, ZBAR_CFG_ENABLE, 1);

	//1、加载灰度图并显示
	Mat imageGray = imread("3.png", 0);	//小于0或不写则加载原图，等于0则加载灰度图，大于0则加载RGB图像
	if (!imageGray.data)
	{
		cout << "请确认图片是否存在" << endl;
		system("pause");
		return 0;
	}
	imshow("【1】灰度图", imageGray);

	//2、计算图像x、y方向的梯度，并求和，突出边缘信息
	Mat grad_x, grad_y,dst;
	Mat abs_grad_x, abs_grad_y;

	Sobel(imageGray, grad_x, CV_16S, 1, 0, 3, 1);
	convertScaleAbs(grad_x, abs_grad_x);
	//imshow("【2】x方向Sobel", abs_grad_x);

	Sobel(imageGray, grad_y, CV_16S, 0, 1, 3, 1);
	convertScaleAbs(grad_y, abs_grad_y);
	//imshow("【3】y方向Sobel", abs_grad_y);

	addWeighted(abs_grad_x, 1, abs_grad_y, 1, 0, dst);
	imshow("【4】合并Sobel梯度", dst);

在这里插入图片描述

	//3、离散傅里叶变换，画出频谱信息
	//扩充边界
	int m = getOptimalDFTSize(dst.rows);
	int n = getOptimalDFTSize(dst.cols);
	Mat padded;							//填充边界，当图片尺寸为2,3,5的整数倍时，离散傅里叶变换速度最快
	copyMakeBorder(dst, padded, 0,m - dst.rows, 0, n - dst.cols,BORDER_CONSTANT,Scalar::all(0));
	//为傅里叶变化的结果（实部和虚部）分配存储空间
	Mat planes[] = { Mat_<float>(padded), Mat::zeros(padded.size(), CV_32F) };
	Mat complexI;
	merge(planes, 2, complexI);

	dft(complexI,complexI);//进行就地离散傅里叶变换
	
	//将复数转换为幅值
	split(complexI, planes);
	magnitude(planes[0], planes[1], planes[0]);
	Mat magnitudeImage = planes[0];
	
	//进行对数尺度缩放
	magnitudeImage += Scalar::all(1);
	log(magnitudeImage, magnitudeImage);

	//剪切和重分布幅度图像限
	int cx = magnitudeImage.cols / 2;
	int cy = magnitudeImage.rows / 2;
	Mat q0(magnitudeImage, Rect(0, 0, cx, cy));   // ROI区域的左上
	Mat q1(magnitudeImage, Rect(cx, 0, cx, cy));  // ROI区域的右上
	Mat q2(magnitudeImage, Rect(0, cy, cx, cy));  // ROI区域的左下
	Mat q3(magnitudeImage, Rect(cx, cy, cx, cy)); // ROI区域的右下
	//交换象限（左上与右下进行交换）
	Mat tmp;
	q0.copyTo(tmp);
	q3.copyTo(q0);
	tmp.copyTo(q3);
	//交换象限（右上与左下进行交换）
	q1.copyTo(tmp);
	q2.copyTo(q1);
	tmp.copyTo(q2);

	normalize(magnitudeImage, magnitudeImage, 0, 1, CV_MINMAX);
	imshow("【5】频谱幅值", magnitudeImage);

	//4、二值化后找出频率变化最大的直线角度
	//二值化并画出频谱图
	Mat thresholdImg(magnitudeImage.size(), CV_8UC1);
	magnitudeImage.convertTo(thresholdImg, CV_8UC1, 255, 0);//因归一化后数值均小于1，故应扩大回原有数值方便阈值化及后续操作
	threshold(thresholdImg, thresholdImg,160,255,CV_THRESH_BINARY);
	imshow("【6】频谱二值化图",thresholdImg);

在这里插入图片描述

在此与参考资料2代码做了找出倾斜直线角度的修改，考虑到频谱二值化图中仍存在干扰，提高霍夫转换阈值可提取出最长线段,但阈值如果取得较低可能会出现多条线段，资料2中没有考虑到此种情况，将所有线段都画了出来（虽然只有一条），我则采用累计概率霍夫变换（好像更复杂了。。。）求出最长线段

//找出最长直线及其角度
	Mat angleImg(thresholdImg.size(),thresholdImg.type(),Scalar::all(0));
	vector<Vec4i> lines;
	HoughLinesP(thresholdImg, lines, 1, CV_PI / 180, 150, 0, 0);	
	int maxNum = 0;
	double len = 0;
	for (int i = 0; i < lines.size(); i++)
	{
		Vec4f j = lines[i];
		Point2f point1 = Point2f(j[0], j[1]);
		Point2f point2 = Point2f(j[2], j[3]);
		if (len <=(Point2PointDist(point1, point2)))
		{
			len = Point2PointDist(point1, point2);
			maxNum = i;
		}
	}
	Point2f a = Point(lines[maxNum][0], lines[maxNum][1]);
	Point2f b = Point(lines[maxNum][2], lines[maxNum][3]);
	line(angleImg, a,b, Scalar::all(255),1,CV_AA);
	imshow("条码倾斜角度", angleImg);
	double angle=Line_Angle(a, b);
	Point center = Point(imageGray.cols/2,imageGray.rows/2);
	Mat rotMat = getRotationMatrix2D(center,angle,1);
	Mat rotImg(imageGray.size(), imageGray.type());
	warpAffine(imageGray,rotImg,rotMat,rotImg.size());
	imshow("校正图像", rotImg);

	//扫描并显示信息
	int width = rotImg.cols;
	int height = rotImg.rows;
	uchar *raw = (uchar *)rotImg.data;
	Image imageZbar(width, height, "Y800", raw, width * height);//y800实际为灰度图代号
	scanner.scan(imageZbar); //扫描条码
	Image::SymbolIterator symbol = imageZbar.symbol_begin();
	if (imageZbar.symbol_begin() == imageZbar.symbol_end())
	{
		cout << "查询条码失败，请检查图片！" << endl;
	}
	for (; symbol != imageZbar.symbol_end(); ++symbol)
	{
		cout << "类型：" << endl << symbol->get_type_name() << endl << endl;
		cout << "条码：" << endl << symbol->get_data() << endl << endl;
	}
	
	waitKey();
	imageZbar.set_data(NULL, 0);
	return 0;
}

//计算两点之间距离
double Point2PointDist(Point2f& a,Point2f& b)
{
	double res = sqrt(pow(a.x - b.x, 2) + pow(a.y - b.y, 2));
	return res;
}

//计算一条直线与水平线角度
double Line_Angle(Point2f& a,Point2f& b)
{
	float k = 0;
	k= (b.y -a.y) / (b.x-a.x);
	double angle = atan(k) / CV_PI * 180;	//图片x轴在正上方
	return angle;
}

在这里插入图片描述