本文主要是介绍在ROS中使用opencv-识别白线,希望对大家解决编程问题提供一定的参考价值,需要的开发者们随着小编来一起学习吧!
当时写的一个识别白线的程序,还不算完整,后面要自己用程序算出两天线之间中点的坐标,并反馈坐标信息回来,跟底层通讯,做一个闭环。
#include<ros/ros.h> //ros标准库头文件
#include<iostream> //C++标准输入输出库
#include<cv_bridge/cv_bridge.h>
#include<sensor_msgs/image_encodings.h>
#include<image_transport/image_transport.h>
#include<opencv2/core/core.hpp>
#include<opencv2/highgui/highgui.hpp>
#include<opencv2/imgproc/imgproc.hpp>
#include <cv.h>
#include <math.h>
using namespace std;
using namespace cv;static const std::string OPENCV_WINDOW1 = "Image window"; //定义输入窗口名称
static const std::string OPENCV_WINDOW2 = "Gray window"; //定义输出窗口名称
static const std::string OPENCV_WINDOW3 = "Canny window"; //定义输出窗口名称
static const std::string OPENCV_WINDOW4 = "Hough window"; //定义输出窗口名称
//定义一个转换的类
class RGB_GRAY
{
private:ros::NodeHandle nh_; //定义ROS句柄image_transport::ImageTransport it_; //定义一个image_transport实例image_transport::Subscriber image_sub_; //定义ROS图象接收器image_transport::Publisher image_pub_; //定义ROS图象发布器
public:RGB_GRAY():it_(nh_) //构造函数{image_sub_ = it_.subscribe("/cv_camera/image_raw", 1, &RGB_GRAY::convert_callback, this); //定义图象接受器,订阅话题是“camera/rgb/image_raw”image_pub_ = it_.advertise("/image_converter/output_video", 1); //定义图象发布器//初始化输入输出窗口cv::namedWindow(OPENCV_WINDOW1);cv::namedWindow(OPENCV_WINDOW2);cv::namedWindow(OPENCV_WINDOW3);cv::namedWindow(OPENCV_WINDOW4);}~RGB_GRAY() //析构函数{cv::destroyWindow(OPENCV_WINDOW1);cv::destroyWindow(OPENCV_WINDOW2);cv::destroyWindow(OPENCV_WINDOW3);cv::destroyWindow(OPENCV_WINDOW4);}/*这是一个ROS和OpenCV的格式转换回调函数,将图象格式从sensor_msgs/Image ---> cv::Mat*/void convert_callback(const sensor_msgs::ImageConstPtr& msg){cv_bridge::CvImagePtr cv_ptr1; // 声明一个CvImage指针的实例cv_bridge::CvImagePtr cv_ptr2; // 声明一个CvImage指针的实例cv_bridge::CvImagePtr cv_ptr3; // 声明一个CvImage指针的实例cv_bridge::CvImagePtr cv_ptr4; // 声明一个CvImage指针的实例try{cv_ptr1 = cv_bridge::toCvCopy(msg, sensor_msgs::image_encodings::BGR8); //将ROS消息中的图象信息提取,生成新cv类型的图象,复制给CvImage指针cv_ptr2 = cv_bridge::toCvCopy(msg, sensor_msgs::image_encodings::BGR8); //将ROS消息中的图象信息提取,生成新cv类型的图象,复制给CvImage指针cv_ptr3 = cv_bridge::toCvCopy(msg, sensor_msgs::image_encodings::BGR8); //将ROS消息中的图象信息提取,生成新cv类型的图象,复制给CvImage指针cv_ptr4 = cv_bridge::toCvCopy(msg, sensor_msgs::image_encodings::BGR8); //将ROS消息中的图象信息提取,生成新cv类型的图象,复制给CvImage指针}catch(cv_bridge::Exception& e) //异常处理{ROS_ERROR("cv_bridge exception: %s", e.what());return;}image_process1(cv_ptr1->image); //得到了cv::Mat类型的图象,在CvImage指针的image中,将结果传送给处理函数image_process2(cv_ptr2->image); //得到了cv::Mat类型的图象,在CvImage指针的image中,将结果传送给处理函数image_process3(cv_ptr3->image); //得到了cv::Mat类型的图象,在CvImage指针的image中,将结果传送给处理函数image_process3(cv_ptr3->image); //得到了cv::Mat类型的图象,在CvImage指针的image中,将结果传送给处理函数}/*这是图象处理的主要函数,一般会把图像处理的主要程序写在这个函数中。这里的例子只是一个彩色图象到灰度图象的转化*/void image_process1(cv::Mat img1)//这里是灰度处理{cv::Mat img_out1;cv::cvtColor(img1, img_out1, CV_RGB2GRAY); //转换成灰度图象cv::imshow(OPENCV_WINDOW1, img1);cv::imshow(OPENCV_WINDOW2, img_out1);cv::waitKey(5);}void image_process2(cv::Mat img2)//这里是边缘检测{cv::Mat dstframe;cv::Mat edge;cv::Mat grayVideo;dstframe.create(img2.size(),img2.type());cv::cvtColor(img2,grayVideo,CV_BGR2GRAY);cv::blur(grayVideo,edge,cvSize(15,15));cv::Canny(edge, edge, 0, 30,3);cv::imshow(OPENCV_WINDOW3, edge);cv::waitKey(5);}void image_process3(cv::Mat img3){cv::Mat dst2;cv::Mat cdst2;cv::Canny(img3, dst2, 50, 200, 3);cv::cvtColor(dst2, cdst2, CV_GRAY2BGR);//灰度化vector<Vec2f> lines;HoughLines(dst2, lines, 1, CV_PI/180, 100, 0, 0 );for( size_t i = 0; i < lines.size(); i++ )//将求得的线条画出来{float rho = lines[i][0], theta = lines[i][1];Point pt1, pt2;double a = cos(theta), b = sin(theta);double x0 = a*rho, y0 = b*rho;pt1.x = cvRound(x0 + 1000*(-b));pt1.y = cvRound(y0 + 1000*(a));pt2.x = cvRound(x0 - 1000*(-b));pt2.y = cvRound(y0 - 1000*(a));line( cdst2, pt1, pt2, Scalar(0,0,255), 2, CV_AA);cv::imshow(OPENCV_WINDOW4, cdst2);cout<<"x="<<(pt1.x+pt2.x)/2<<endl; cout<<"y="<<(pt1.y+pt2.y)/2<<endl;}cv::waitKey(5);}};//主函数int main(int argc, char** argv){ros::init(argc, argv, "RGB");RGB_GRAY obj;ros::spin();}
当时写的一个识别白线的程序,还不算完整,后面要自己用程序算出两天线之间中点的坐标,并反馈坐标信息回来,跟底层通讯,做一个闭环。
#include<ros/ros.h> //ros标准库头文件
#include<iostream> //C++标准输入输出库
#include<cv_bridge/cv_bridge.h>
#include<sensor_msgs/image_encodings.h>
#include<image_transport/image_transport.h>
#include<opencv2/core/core.hpp>
#include<opencv2/highgui/highgui.hpp>
#include<opencv2/imgproc/imgproc.hpp>
#include <cv.h>
#include <math.h>
using namespace std;
using namespace cv;static const std::string OPENCV_WINDOW1 = "Image window"; //定义输入窗口名称
static const std::string OPENCV_WINDOW2 = "Gray window"; //定义输出窗口名称
static const std::string OPENCV_WINDOW3 = "Canny window"; //定义输出窗口名称
static const std::string OPENCV_WINDOW4 = "Hough window"; //定义输出窗口名称
//定义一个转换的类
class RGB_GRAY
{
private:ros::NodeHandle nh_; //定义ROS句柄image_transport::ImageTransport it_; //定义一个image_transport实例image_transport::Subscriber image_sub_; //定义ROS图象接收器image_transport::Publisher image_pub_; //定义ROS图象发布器
public:RGB_GRAY():it_(nh_) //构造函数{image_sub_ = it_.subscribe("/cv_camera/image_raw", 1, &RGB_GRAY::convert_callback, this); //定义图象接受器,订阅话题是“camera/rgb/image_raw”image_pub_ = it_.advertise("/image_converter/output_video", 1); //定义图象发布器//初始化输入输出窗口cv::namedWindow(OPENCV_WINDOW1);cv::namedWindow(OPENCV_WINDOW2);cv::namedWindow(OPENCV_WINDOW3);cv::namedWindow(OPENCV_WINDOW4);}~RGB_GRAY() //析构函数{cv::destroyWindow(OPENCV_WINDOW1);cv::destroyWindow(OPENCV_WINDOW2);cv::destroyWindow(OPENCV_WINDOW3);cv::destroyWindow(OPENCV_WINDOW4);}/*这是一个ROS和OpenCV的格式转换回调函数,将图象格式从sensor_msgs/Image ---> cv::Mat*/void convert_callback(const sensor_msgs::ImageConstPtr& msg){cv_bridge::CvImagePtr cv_ptr1; // 声明一个CvImage指针的实例cv_bridge::CvImagePtr cv_ptr2; // 声明一个CvImage指针的实例cv_bridge::CvImagePtr cv_ptr3; // 声明一个CvImage指针的实例cv_bridge::CvImagePtr cv_ptr4; // 声明一个CvImage指针的实例try{cv_ptr1 = cv_bridge::toCvCopy(msg, sensor_msgs::image_encodings::BGR8); //将ROS消息中的图象信息提取,生成新cv类型的图象,复制给CvImage指针cv_ptr2 = cv_bridge::toCvCopy(msg, sensor_msgs::image_encodings::BGR8); //将ROS消息中的图象信息提取,生成新cv类型的图象,复制给CvImage指针cv_ptr3 = cv_bridge::toCvCopy(msg, sensor_msgs::image_encodings::BGR8); //将ROS消息中的图象信息提取,生成新cv类型的图象,复制给CvImage指针cv_ptr4 = cv_bridge::toCvCopy(msg, sensor_msgs::image_encodings::BGR8); //将ROS消息中的图象信息提取,生成新cv类型的图象,复制给CvImage指针}catch(cv_bridge::Exception& e) //异常处理{ROS_ERROR("cv_bridge exception: %s", e.what());return;}image_process1(cv_ptr1->image); //得到了cv::Mat类型的图象,在CvImage指针的image中,将结果传送给处理函数image_process2(cv_ptr2->image); //得到了cv::Mat类型的图象,在CvImage指针的image中,将结果传送给处理函数image_process3(cv_ptr3->image); //得到了cv::Mat类型的图象,在CvImage指针的image中,将结果传送给处理函数image_process3(cv_ptr3->image); //得到了cv::Mat类型的图象,在CvImage指针的image中,将结果传送给处理函数}/*这是图象处理的主要函数,一般会把图像处理的主要程序写在这个函数中。这里的例子只是一个彩色图象到灰度图象的转化*/void image_process1(cv::Mat img1)//这里是灰度处理{cv::Mat img_out1;cv::cvtColor(img1, img_out1, CV_RGB2GRAY); //转换成灰度图象cv::imshow(OPENCV_WINDOW1, img1);cv::imshow(OPENCV_WINDOW2, img_out1);cv::waitKey(5);}void image_process2(cv::Mat img2)//这里是边缘检测{cv::Mat dstframe;cv::Mat edge;cv::Mat grayVideo;dstframe.create(img2.size(),img2.type());cv::cvtColor(img2,grayVideo,CV_BGR2GRAY);cv::blur(grayVideo,edge,cvSize(15,15));cv::Canny(edge, edge, 0, 30,3);cv::imshow(OPENCV_WINDOW3, edge);cv::waitKey(5);}void image_process3(cv::Mat img3){cv::Mat dst2;cv::Mat cdst2;cv::Canny(img3, dst2, 50, 200, 3);cv::cvtColor(dst2, cdst2, CV_GRAY2BGR);//灰度化vector<Vec2f> lines;HoughLines(dst2, lines, 1, CV_PI/180, 100, 0, 0 );for( size_t i = 0; i < lines.size(); i++ )//将求得的线条画出来{float rho = lines[i][0], theta = lines[i][1];Point pt1, pt2;double a = cos(theta), b = sin(theta);double x0 = a*rho, y0 = b*rho;pt1.x = cvRound(x0 + 1000*(-b));pt1.y = cvRound(y0 + 1000*(a));pt2.x = cvRound(x0 - 1000*(-b));pt2.y = cvRound(y0 - 1000*(a));line( cdst2, pt1, pt2, Scalar(0,0,255), 2, CV_AA);cv::imshow(OPENCV_WINDOW4, cdst2);cout<<"x="<<(pt1.x+pt2.x)/2<<endl; cout<<"y="<<(pt1.y+pt2.y)/2<<endl;}cv::waitKey(5);}
};
//主函数
int main(int argc, char** argv)
{ros::init(argc, argv, "RGB");RGB_GRAY obj;ros::spin();
}
看着好烦,稍微简化了一下,我写代码的风格是代码量越少越好。可能坐标计算这里还需要改进。
#include<ros/ros.h> //ros标准库头文件
#include<iostream> //C++标准输入输出库
#include<cv_bridge/cv_bridge.h>
#include<sensor_msgs/image_encodings.h>
#include<image_transport/image_transport.h>
#include<opencv2/highgui/highgui.hpp>
#include<opencv2/imgproc/imgproc.hpp>
using namespace std;
using namespace cv;static const std::string OPENCV_WINDOW = "Hough window"; //定义输出窗口名称
//定义一个转换的类
class RGB_GRAY
{
private:ros::NodeHandle nh_; //定义ROS句柄image_transport::ImageTransport it_; //定义一个image_transport实例image_transport::Subscriber image_sub_; //定义ROS图象接收器image_transport::Publisher image_pub_; //定义ROS图象发布器
public:RGB_GRAY():it_(nh_) //构造函数{image_sub_ = it_.subscribe("/cv_camera/image_raw", 1, &RGB_GRAY::convert_callback, this); //定义图象接受器,订阅话题是“camera/rgb/image_raw”image_pub_ = it_.advertise("/image_converter/output_video", 1); //定义图象发布器//初始化输入输出窗口cv::namedWindow(OPENCV_WINDOW);}~RGB_GRAY() //析构函数{cv::destroyWindow(OPENCV_WINDOW);}/*这是一个ROS和OpenCV的格式转换回调函数,将图象格式从sensor_msgs/Image ---> cv::Mat */void convert_callback(const sensor_msgs::ImageConstPtr& msg){cv_bridge::CvImagePtr cv_ptr; // 声明一个CvImage指针的实例try{cv_ptr = cv_bridge::toCvCopy(msg, sensor_msgs::image_encodings::BGR8); //将ROS消息中的图象信息提取,生成新cv类型的图象,复制给CvImage指针}catch(cv_bridge::Exception& e) //异常处理{ROS_ERROR("cv_bridge exception: %s", e.what());return;}image_process(cv_ptr->image); //得到了cv::Mat类型的图象,在CvImage指针的image中,将结果传送给处理函数}/*这是图象处理的主要函数,一般会把图像处理的主要程序写在这个函数中。这里的例子只是一个彩色图象到灰度图象的转化*/void image_process(cv::Mat img)//这里是灰度处理{Mat dst;Mat cdst;Canny(img, dst, 50, 200, 3);cvtColor(dst, cdst, CV_GRAY2BGR);//灰度化vector<Vec2f> lines;HoughLines(dst, lines, 1, CV_PI/180, 100, 0, 0 );for( size_t i = 0; i < lines.size(); i++ )//将求得的线条画出来{float rho = lines[i][0], theta = lines[i][1];Point pt1, pt2;double a = cos(theta), b = sin(theta);double x0 = a*rho, y0 = b*rho;pt1.x = cvRound(x0 + 1000*(-b));pt1.y = cvRound(y0 + 1000*(a));pt2.x = cvRound(x0 - 1000*(-b));pt2.y = cvRound(y0 - 1000*(a));line( cdst, pt1, pt2, Scalar(0,0,255), 2, CV_AA);cout<<"x="<<(pt1.x+pt2.x)/2<<endl; cout<<"y="<<(pt1.y+pt2.y)/2<<endl;}imshow(OPENCV_WINDOW, cdst);waitKey(5);}
};
//主函数
int main(int argc, char** argv)
{ros::init(argc, argv, "RGB");RGB_GRAY obj;ros::spin();
}
这篇关于在ROS中使用opencv-识别白线的文章就介绍到这儿,希望我们推荐的文章对编程师们有所帮助!