ROS使用FLIR品牌的相机

本文介绍如何将FLIR品牌的相机应用于ROS系统，将相机实时的数据以rostopic实时广播：

系统平台：Ubuntu16.04 LTS   ROS Kinetic

依赖：OpenCV（本例中所用版本为2.9）， cv_bridge

硬件：FLIR Grasshopper GS3-PGE-23S6C-C

1.首先安装相机在Ubuntu16.04下的驱动：

本例直接安装于买设备时自带的安装盘，于官网也可以下载到相关驱动：

网址：www.flir.com/mv-techsupport/downloads

找到相同设备的下载列表，选择对应的系统版本即可

安装完成后，控制台输入flycap有软件弹出代表安装成功

点击Auto Force IP，可以为设备自动分配IP，可以通过flycap看到设备

点击确定，可以进入照相模式

2.ROS代码编写：

参考地址：https://blog.csdn.net/qq_27050183/article/details/51141998

https://blog.csdn.net/lixujie666/article/details/83303836

项目中用QT编写代码，用QMake的pro文件配置依赖目录：

 TEMPLATE = app

 CONFIG += console c++

 QMAKE_CXXFLAGS += -std=c++

 CONFIG -= app_bundle

 CONFIG -= qt

 SOURCES += \

         main.cpp

 INCLUDEPATH += /usr/include/flycapture

 INCLUDEPATH += /usr/local/include/opencv2

 LIBS += /usr/local/lib/libopencv_*.so \

 LIBS += /usr/lib/libflycapture*.so

 LIBS += /usr/lib/libcv_bridge.so

 #---------------------ROS DEPENDS---------------------#

 INCLUDEPATH += /opt/ros/kinetic/include

 DEPENDPATH += /opt/ros/kinetic/include

 LIBS += -L/opt/ros/kinetic/lib \

         -L/usr/local/lib \

         -lroscpp \

         -lrospack \

         -lpthread \

         -lrosconsole \

         -lrosconsole_log4cxx \

         -lrosconsole_backend_interface \

         -lxmlrpcpp \

         -lroscpp_serialization \

         -lrostime \

         -lcpp_common \

         -lroslib \

         -ltf \

 LIBS += -L /usr/lib \

         -l Ice

 LIBS += /opt/ros/kinetic/lib/libroslib.so \

         /usr/local/lib/liblog4cplus.so

 #---------------------ROS DEPENDS---------------------#

编写C++通信代码：

 #include <iostream>

 #include <stdlib.h>

 #include <sensor_msgs/Image.h>

 #include <cv_bridge/cv_bridge.h>

 #include <ros/ros.h>

 using namespace std;

 #include "FLIR.h"

 using namespace FLIR;

 unsigned int width=;

 unsigned int height=;

 int main(int argc, char ** argv)

 {

     if(argc==){

         width=atoi(argv[]);

         height=atoi(argv[]);

     }

     GigECamera cam;

     CameraInfo camInfo[];

     unsigned int numCameras;

     FlyCapture2::Error error;

     error = BusManager::DiscoverGigECameras(camInfo, &numCameras);

     if (error != PGRERROR_OK)

     {

         PrintError(error);

         return false;

     }

     IPAddress ipAddress=camInfo[camIndex-].ipAddress;

     sprintf(ipStr,"%d.%d.%d.%d",ipAddress.octets[],ipAddress.octets[],ipAddress.octets[],ipAddress.octets[]);

     BusManager busMgr;

     error=busMgr.GetCameraFromIPAddress(ipAddress,&guid);

     if (error != PGRERROR_OK)

     {

         PrintError(error);

         return false;

     }

     error = cam.Connect(&guid);

     if (error != PGRERROR_OK)

     {

         PrintError(error);

         return false;

     }

     GigEImageSettingsInfo imageSettingsInfo;

     error = cam.GetGigEImageSettingsInfo(&imageSettingsInfo);

     if (error != PGRERROR_OK)

     {

         PrintError(error);

         return false;

     }

     GigEImageSettings imageSettings;

     imageSettings.offsetX = (unsigned int)(imageSettingsInfo.maxWidth-width)/;

     imageSettings.offsetY = (unsigned int)(imageSettingsInfo.maxHeight-height)/;

     imageSettings.height = height;

     imageSettings.width = width;

     imageSettings.pixelFormat = PIXEL_FORMAT_RGB;

     error = cam.SetGigEImageSettings(&imageSettings);

     if (error != PGRERROR_OK)

     {

         PrintError(error);

         return false;

     }

     error = cam.StartCapture();

     if (error != PGRERROR_OK)

     {

         PrintError(error);

         return false;

     }

     cv::Mat img;

     cam.RetrieveBGR(img);

     ros::init(argc, argv, "camera_node");

     ros::NodeHandle nh;

     ros::Publisher pub_camera = nh.advertise<sensor_msgs::Image>("flir_camera", );

     ros::Rate r = ros::Rate();

     cv_bridge::CvImage cvi;

     cvi.header.frame_id = "image";

     cvi.encoding = "bgr8";

     while (ros::ok()){

         cam.RetrieveBGR(img);

         sensor_msgs::Image image;

         cvi.image = img;

         cvi.toImageMsg(image);

         pub_camera.publish(image);

         r.sleep();

     }

     return ;

 }

编译，运行

3.完成效果：

发布消息类型为sensor_msgs::Image的话题，话题名为"flir_camera"，可以用rqt_image_view查看效果：