小觅双目相机测试

官方SDK  https://github.com/slightech/MYNT-EYE-SDK

官方淘宝 

WINDOS 下配置http://blog.csdn.net/u013749068/article/details/79463736

ORB-SLAM2 https://github.com/slightech/MYNT-EYE-ORB-SLAM2-Sample

每个版本SDK下载：  https://pan.baidu.com/s/1i5REqVz#list/path=%2F&parentPath=%2F

查看 

ls /dev/video*

1 TX1安装记录

1 安装基本依赖

2 下载SDK  

3 安装教程  https://slightech.github.io/MYNT-EYE-SDK/getting_started_tegra.html

从 Install SDK 开始 ，第一个kuda自带

cd <sdk directory>
$ ./install.sh
$ source ~/.bashrc
查看是否安装好
$ echo $MYNTEYE_SDK_ROOT
编译测试
$ cd samples/
$ mkdir build
$ cd build/
$ cmake -DCMAKE_BUILD_TYPE=Release ..
$ make

4 运行示例

环境 配置普通tx1版的 opencv3.2
在路径ubuntu@tegra-ubuntu:~/Code/XiMi/SDK/mynteye/samples$

./samples/bin/camera [name]
[name] 为空 自动选择摄像头

4.1 测试左右视图

./samples/bin/camera 
  • 自动选择摄像头
  • 使用HUB，因为USB问题，可能读不出來，换一个
  • 查看USB  ls /dev/video*

4.2 测试深度图

./samples/bin/camera2 

2ORB-SLAM2  

我们创建了一个stereo_mynteye在ORB-SLAM2中命名的样本，以展示如何使用我们的MYNT EYE相机。

在运行stereo_mynteyeORB-SLAM2之前，请按照以下步骤操作：

  1. 在这里下载MYNT EYE SDK 并按照教程进行安装。

  2. 按照正常程序安装ORB-SLAM2。

  3. $cd stereo_mynteyeORB-SLAM2
    $chmod +x build.sh
    $./build.sh

  4. 校准MYNT EYE相机并更改参数cam_stereo.yaml。

  5. 运行stereo_mynteye，

    $ ./Examples/Stereo/mynt-eye-orb-slam2-sample ./Vocabulary/ORBvoc.txt ./Examples/Stereo/cam_stereo.yaml 1

    ./Vocabulary/ORBvoc.txt是词汇文件./Examples/Stereo/cam_stereo.yaml的路径，是相机配置文件的路径，1代表video1。

  6. 你可以通过这里的视频了解运行效果。

在文件夹
ubuntu@tegra-ubuntu:~/Code/XiMi/MYNT-EYE-ORB-SLAM2-Sample-mynteye/Examples/Stereo$


./stereo_mynteye ./ORBvoc.txt ./cam_stereo.yaml 0

3ROS 

1必须安装ROS 版的SDK

下载地址https://pan.baidu.com/s/1i5REqVz#list/path=%2FMYNT-EYE-SDK%2F1.x%2F1.8&parentPath=%2F

cd <sdk directory>
$ ./install.sh
$ source ~/.bashrc
检查是否安装好
echo $MYNTEYE_SDK_ROOT

2Build SDK Samples

$ cd samples/
$ mkdir build
$ cd build/
$ cmake -DCMAKE_BUILD_TYPE=Release ..
$ make

 3  运行实例

./samples/bin/camera

需要一个相机标定文件,在博文最后自己标定

SN03882E340009070E.conf

4在ROS中打开  发布左右图 深度图 IMU

4.1 下载ROS测试包

地址  https://github.com/slightech/MYNT-EYE-ROS-Wrapper

新建一个ROS包，下载

git clone https://github.com/slightech/MYNT-EYE-ROS-Wrapper.git

4.2 编译

catkin_make
source ./devel/setup.bash

4.3 根据相机USB口修改打开0还是1

修改launchu 文件 0 usb
<param name="device_name" type="int" value="0" />

4.4 开始运行实例

# Run MYNT EYE camera node, and open Rviz to display
$ roslaunch mynteye_ros_wrapper mynt_camera_display.launch

# Run MYNT EYE camera node
$ roslaunch mynteye_ros_wrapper mynt_camera.launch
# Or,
$ rosrun mynteye_ros_wrapper mynteye_wrapper_node

ROS 标定参数

%YAML:1.0
M1: !!opencv-matrix
   rows: 3
   cols: 3
   dt: d
   data: [ 7.0670498518636600e+002, 0., 3.7297901043077900e+002, 0.,
       7.0666416610920203e+002, 2.1240374979570200e+002, 0., 0., 1. ]
D1: !!opencv-matrix
   rows: 1
   cols: 14
   dt: d
   data: [ -4.6388704908098000e-001, 2.3425087249454701e-001, 0., 0., 0.,
       0., 0., 0., 0., 0., 0., 0., 0., 0. ]
M2: !!opencv-matrix
   rows: 3
   cols: 3
   dt: d
   data: [ 7.0503556956133104e+002, 0., 4.0450936863506502e+002, 0.,
       7.0491472237897597e+002, 2.2503054543868200e+002, 0., 0., 1. ]
D2: !!opencv-matrix
   rows: 1
   cols: 14
   dt: d
   data: [ -4.6174421404117999e-001, 2.1536606632286900e-001, 0., 0., 0.,
       0., 0., 0., 0., 0., 0., 0., 0., 0. ]
R: !!opencv-matrix
   rows: 3
   cols: 3
   dt: d
   data: [ 9.9997459395392996e-001, -3.3977699515345002e-003,
       -6.2663072083021200e-003, -3.3890405418889199e-003,
       9.9999327274544803e-001, 1.4031635874333700e-003,
       -6.2710326803329004e-003, -1.3818911694177701e-003,
       9.9997938205040904e-001 ]
T: !!opencv-matrix
   rows: 3
   cols: 1
   dt: d
   data: [ -1.1963887403661900e+002, -1.5522537096063099e-001,
       -9.4210539328372297e-003 ]

　　SGM-双目深度图 ROS

#include <iostream>
#include <numeric>
#include <sys/time.h>
#include <vector>
#include <stdlib.h>
#include <typeinfo>
#include <opencv2/opencv.hpp>

//ROS 
#include <ros/ros.h>
#include <image_transport/image_transport.h>
#include <opencv2/highgui/highgui.hpp>
#include <cv_bridge/cv_bridge.h>

//CUDA sgm
#include <numeric>
#include <stdlib.h>
#include <ctime>
#include <sys/types.h>
#include <stdint.h>
#include <linux/limits.h>
#include <dirent.h>
#include <iostream>
#include <fstream>
#include "disparity_method.h"

//MY-EYE
#include "camera.h"
#include "utility.h"



char command;
cv::Mat left,right;

void imageCallbackLeft(const sensor_msgs::ImageConstPtr &msg)
{
   try
   {
       left = cv_bridge::toCvShare(msg,"bgr8")->image;
       cv::imshow("left",left );
       command=cv::waitKey(1);
       if(command == ' ')
       {
	 cv::imwrite("left.jpg",left);
         cv::imwrite("right.jpg",right);
       }
   }
   catch(cv_bridge::Exception &e)
   {
    
   }
}

void imageCallbackRight(const sensor_msgs::ImageConstPtr &msg)
{
   try
   {
       right =  cv_bridge::toCvShare(msg,"bgr8")->image;
       cv::imshow("right",right);
       //cv::waitKey(1);
   }
   catch(cv_bridge::Exception &e)
   {
    
   }
}


int main(int argc, char *argv[]) {
	
        ros::init(argc, argv, "sgm_ros");
        ros::NodeHandle nh("~");
        image_transport::ImageTransport it(nh);
        image_transport::Subscriber sub_left = it.subscribe("/left/image_rect_color", 1,imageCallbackLeft);
        image_transport::Subscriber sub_right = it.subscribe("/right/image_rect_color", 1,imageCallbackRight);

	//open camera
        std::string name;
	if (argc >= 2) {
	    name = argv[1];
	} else {
	    bool found = false;
	    name = FindDeviceName(&found);
	}
        cout << "Open Camera: " << name << endl;

        CalibrationParameters *calib_params = nullptr;
        if (argc >= 3) {
            stringstream ss;
            ss << argv[2];
            calib_params = new CalibrationParameters;
            calib_params->Load(ss.str());
        }
        InitParameters init_params(name, calib_params);

        Camera cam;
       //cam.SetMode(Mode::MODE_CPU);
        cam.Open(init_params);

        if (calib_params)
           delete calib_params;

        if (!cam.IsOpened()) {
            std::cerr << "Error: Open camera failed" << std::endl;
            return 1;
         }
         cout << "33[1;32mPress ESC/Q on Windows to terminate33[0m
";        


        
	uint8_t p1 = 7;
	uint8_t p2 = 84;

	init_disparity_method(p1, p2);
   
        ros::Rate loop_rate(50);
        
        while(ros::ok())
        {
                ros::spinOnce();
                loop_rate.sleep();
 
		cv::Mat h_im0 = left.clone();
		if(!h_im0.data) {
			std::cerr << "Couldn't read the file " << std::endl;
			continue;
		}
		cv::Mat h_im1 = right.clone();
		if(!h_im1.data) {
			std::cerr << "Couldn't read the file " << std::endl;
			continue;
		}

		// Convert images to grayscale
		if (h_im0.channels()>1) {
			cv::cvtColor(h_im0, h_im0, CV_RGB2GRAY);
		}

		if (h_im1.channels()>1) {
			cv::cvtColor(h_im1, h_im1, CV_RGB2GRAY);
		}

		if(h_im0.rows != h_im1.rows || h_im0.cols != h_im1.cols) {
			std::cerr << "Both images must have the same dimensions" << std::endl;
			return EXIT_FAILURE;
		}
		if(h_im0.rows % 4 != 0 || h_im0.cols % 4 != 0) {
		        std::cerr << "Due to implementation limitations image width and height must be a divisible by 4" << std::endl;
		        return EXIT_FAILURE;
		}


		// Compute
		float elapsed_time_ms;
		cv::Mat disparity_im = compute_disparity(h_im0, h_im1, &elapsed_time_ms);
                std::cout << "matching time:  " << elapsed_time_ms << std::endl;
  
		const int type = disparity_im.type();
		const uchar depth = type & CV_MAT_DEPTH_MASK;
		if(depth == CV_8U) {
		        cv::imshow("disp",disparity_im);
                       //cv::imwrite("disp.jpg", disparity_im);
		} else {
			cv::Mat disparity16(disparity_im.rows, disparity_im.cols, CV_16UC1);
			for(int i = 0; i < disparity_im.rows; i++) {
				for(int j = 0; j < disparity_im.cols; j++) {
					const float d = disparity_im.at<float>(i, j)*256.0f;
					disparity16.at<uint16_t>(i, j) = (uint16_t) d;
				}
			}
			cv::imshow("disp",disparity_im);
                       //cv::imwrite("disp.jpg", disparity16);
		}

       }

         

	return 0;
}

　　

【推广】 免费学中医，健康全家人
原文地址：https://www.cnblogs.com/kekeoutlook/p/8619896.html