天天看點
傳回

win10安裝openni2以及擷取kinect v1圖像

  • 初始環境：win10(64位)+vs2015+opencv2.4.13
  • 硬體：kinect v1
  • 最終環境：win10(64位)+vs2015+opencv2.4.13+kinect SDK 1.8+OpenNI2.2+NITE2.2

1.配置openni2環境（kinect SDK 1.8+OpenNI2.2+NITE2.2）

參考網址：

-(1)http://blog.csdn.net/chenxin_130/article/details/8580636

-(2)https://social.msdn.microsoft.com/Forums/en-US/bae80fb0-fce5-468e-a292-fe46381af3e5/how-to-install-openni-22-nite-22-kinect-sdk-18-windows-7-3264-bit?forum=kinectsdk

-(3)http://www.cnblogs.com/tornadomeet/archive/2012/11/16/2772681.html

（1）一開始不要把kinect插在電腦上！先安裝Kinect for Windows SDK 1.8（Kinect for Windows SDK 1.8 官方下載下傳網址），再連接配接kinect，可以從裝置管理器裡面看或者通過運作sample檢驗是否安裝成功。（如果一開始就先把kinect連上了，windows會自動裝上不正确的驅動，後面在裝置管理器裡會發現不能識别kinect，隻能識别為“未知裝置”，這時候先把這個未知裝置的驅動删了，然後把kinect拔了，裝了SDK以後再插上kinect）

（2）安裝好驅動以後，先安裝openni2.2（64位openni2.2 CSDN下載下傳網址，32位openni2.2 CSDN下載下傳網址），再安裝nite2.2（64位nite2.2 CSDN下載下傳網址，32位nite2.2 CSDN下載下傳網址）。安裝包從網址(2)下載下傳，或者從CSDN上搜。網址（2）中建議64位電腦需要同時安裝openni2.2和nite2.2的64位和32位版本，但我嘗試過兩種版本都安裝，安裝後會導緻打不開kinect，是以隻裝64位或者隻裝32位的openni和nite就好了。

（3）每次運作exe程式，将OpenNI2\Redist\檔案夾下所有檔案粘貼到exe同目錄下。

（4）如果有kinect初始化失敗的情況，可以嘗試用管理者身份打開記事本，打開”C:\Program Files (x86)\OpenNI2\Redist\OpenNI2\Drivers\PS1080.ini”，将”;UsbInterface=2”修改為”UsbInterface=0”（記得去掉UsbInterface前的”;”）。（測試openni2 sample 如NiViewer時，需修改sample同目錄下的”OpenNI2\Drivers\PS1080.ini”檔案）

（4）配置VS開發環境，見參考網址(1)。

2.擷取kinect v1圖像

（1）擷取彩色圖和深度圖程式

參考網址：http://blog.csdn.net/chenxin_130/article/details/8619909

（2）擷取紅外圖像

代碼如下：

#include <stdlib.h> 
#include <iostream> 
#include <string> 
#include "OpenNI.h" 
#include "opencv2/core/core.hpp" 
#include "opencv2/highgui/highgui.hpp" 
#include "opencv2/imgproc/imgproc.hpp" 
using namespace std;
using namespace cv;
using namespace openni;


void CheckOpenNIError(Status result, string status)
{
if (result != STATUS_OK)
cerr << status << " Error: " << OpenNI::getExtendedError() << endl;
}

int main(int argc, char** argv)
{
Status result = STATUS_OK;

//OpenNI2 image 
VideoFrameRef oniIrImg;

//OpenCV image 
cv::Mat cvIrImg;

cv::namedWindow("Ir");
char key = ;

//【1】 
// initialize OpenNI2 
result = OpenNI::initialize();
CheckOpenNIError(result, "initialize context");

// open device 
Device device;
result = device.open(openni::ANY_DEVICE);

//【2】 
// create Ir stream 
VideoStream oniIrStream;
result = oniIrStream.create(device, openni::SENSOR_IR);

//【3】 
// set Ir video mode 
VideoMode modeIr;
modeIr.setResolution(, );
modeIr.setFps();
modeIr.setPixelFormat(PIXEL_FORMAT_GRAY16);
oniIrStream.setVideoMode(modeIr);
// start Ir stream 
result = oniIrStream.start();

while (key != )
{
if (oniIrStream.readFrame(&oniIrImg) == STATUS_OK)
{
cv::Mat cvRawImg16U(oniIrImg.getHeight(), oniIrImg.getWidth(), CV_16UC1, (void*)oniIrImg.getData());
cvRawImg16U.convertTo(cvIrImg, CV_8U);

cv::imshow("Ir", cvIrImg);
}
key = cv::waitKey();
}

//cv destroy 
cv::destroyWindow("Ir");

//OpenNI2 destroy 
oniIrStream.destroy();
device.close();
OpenNI::shutdown();

return ;
}

3.運作程式

注意：假設project位置為”..\Visual Studio 2015\Projects\GetKinectIRImageProject\”，程式在Debug x86模式下運作，需要将”C:\Program Files (x86)\OpenNI2\Redist”下所有檔案粘貼到”..\Visual Studio 2015\Projects\GetKinectIRImageProject\Debug\”檔案夾下。

win10安裝openni2以及擷取kinect v1圖像

4.讀取彩色（RGB）圖和紅外（IR）圖并儲存 

/*************************
OpenNI2 Deep, Color and Fusion Image
Author: Xin Chen, 2013.2
Blog: http://blog.csdn.net/chenxin_130
*************************/
/*************************
Add IR image getting and Image saving
Author: LuHaiyan, 2017.2
Blog: http://blog.csdn.net/u011988573
*************************/

//否則會出現錯誤：This function or variable may be unsafe.Consider using localtime_s instead
#pragma warning(disable:4996)
#include<time.h> //擷取時間
#include <direct.h> //建立檔案夾
#include <stdlib.h>  
#include <iostream>  
#include <string>  
#include "OpenNI.h"  
#include "opencv2/core/core.hpp"  
#include "opencv2/highgui/highgui.hpp"  
#include "opencv2/imgproc/imgproc.hpp"  
using namespace std;
using namespace cv;
using namespace openni;

/*
部落格（http://www.cnblogs.com/tornadomeet/archive/2012/11/16/2772681.html）中提到，
Kinect在使用時，微軟官方推薦的距離為1220mm（4’）~3810mm(12.5’)，
網友heresy在他的博文Kinect + OpenNI 的深度值中統計過，
kinect在距離為1.0m時其精度大概是3mm,而當距離是3.0m時,其精度大概是3cm，
是以當kinect在官方推薦的距離範圍内使用時，
如果是1.2m時，其精度應該在3mm附近，
如果是3.6m時其精度就大于3cm了，
是以距離越遠，其深度值精度越低。
另外，通過OpenNI擷取到的深度資訊（即z坐标）的機關是mm，
這一點在程式程式設計中要注意，且一般的深度值用12bit(其實是可以用13bit表示的)表示，即最大值為4095，也就是代表4.095m，
是以平時我們采集到的深度資料如果需要擴充到灰階圖，可以乘以一個因子255/4095
(為了加快該結果，一般設定為256/4096，即轉換後的灰階值每變化1，代表kinect采集到的深度值變化了16mm，
如果當人的距離為1米左右時，本來精度是3mm，現在經過歸一化後精度确更加下降了，
如果其深度值為0表示該位置處偵測不到像素點的深度。
*/
#define DEPTH_SCALE_FACTOR 256./4096.

string getTime()
{
    time_t rawtime;
    struct tm * timeinfo;
    char buffer[];

    time(&rawtime);
    timeinfo = localtime(&rawtime);

    strftime(buffer, sizeof(buffer), "%Y_%m_%d_%H_%M_%S", timeinfo);
    string time_now = buffer;
    cout << time_now << endl;
    return time_now;
}

void CheckOpenNIError(Status result, string status)
{
    if (result != STATUS_OK)
        cerr << status << " Error: " << OpenNI::getExtendedError() << endl;
}

int main(int argc, char** argv)
{
    Status result = STATUS_OK;

    //OpenNI2 image  
    //VideoFrameRef oniDepthImg;
    VideoFrameRef oniColorImg;
    VideoFrameRef oniIrImg;

    //OpenCV image  
    //cv::Mat cvDepthImg;
    cv::Mat cvBGRImg;
    //cv::Mat cvFusionImg;
    cv::Mat cvIrImg;

    //cv::namedWindow("depth");
    cv::namedWindow("rgb");
    //cv::namedWindow("fusion");
    cv::namedWindow("Ir");

    //【1】  
    // initialize OpenNI2  
    result = OpenNI::initialize();
    CheckOpenNIError(result, "initialize context");

    // open device    
    Device device;
    result = device.open(openni::ANY_DEVICE);

    //【2】  
     create depth stream   
    //VideoStream oniDepthStream;
    //result = oniDepthStream.create(device, openni::SENSOR_DEPTH);
    //
    【3】  
     set depth video mode  
    //VideoMode modeDepth;
    //modeDepth.setResolution(640, 480);
    //modeDepth.setFps(30);
    //modeDepth.setPixelFormat(PIXEL_FORMAT_DEPTH_1_MM);
    //oniDepthStream.setVideoMode(modeDepth);
     start depth stream  
    //result = oniDepthStream.start();

    // create Ir stream 
    VideoStream oniIrStream;
    result = oniIrStream.create(device, openni::SENSOR_IR);

    //【3】 
    // set Ir video mode 
    VideoMode modeIr;
    modeIr.setResolution(, );
    modeIr.setFps();
    modeIr.setPixelFormat(PIXEL_FORMAT_GRAY16);
    oniIrStream.setVideoMode(modeIr);
    // start Ir stream 
    //result = oniIrStream.start();

    // create color stream  
    VideoStream oniColorStream;
    result = oniColorStream.create(device, openni::SENSOR_COLOR);
    // set color video mode  
    VideoMode modeColor;
    modeColor.setResolution(, );
    modeColor.setFps();
    modeColor.setPixelFormat(PIXEL_FORMAT_RGB888);
    oniColorStream.setVideoMode(modeColor);

    //【4】  
    // set depth and color imge registration mode  
    /*if (device.isImageRegistrationModeSupported(IMAGE_REGISTRATION_DEPTH_TO_COLOR))
    {
    device.setImageRegistrationMode(IMAGE_REGISTRATION_DEPTH_TO_COLOR);
    }*/
    // start color stream  
    result = oniColorStream.start();

    // mkdir
    /*string path_of_all = "F:\\Data_kinect\\test\\";
    string path_of_rgb_d = path_of_all + getTime();
    _mkdir(const_cast<char*>(path_of_rgb_d.c_str()));*/

    char pathOfAll[]= "F:\\Data_kinect\\test\\";
    strcat(pathOfAll, getTime().c_str());
    strcat(pathOfAll, "\\");
    _mkdir(pathOfAll);
    char path[];
    char path_rgb[];
    char path_ir[];
    //string path_of_rgb = path_of_rgb_d + "\\rgb_image";
    //string path_of_depth = path_of_rgb_d + "\\depth_image";
    //string path_of_ir = path_of_rgb_d + "\\ir_image";
    //string path_of_fusion = path_of_rgb_d + "\\fusion_image";
    //_mkdir(const_cast<char*>(path_of_rgb.c_str()));
    //_mkdir(const_cast<char*>(path_of_depth.c_str()));
    //_mkdir(const_cast<char*>(path_of_fusion.c_str()));
    //_mkdir(const_cast<char*>(path_of_ir.c_str()));

    // image index
    int frame_index = ;
    char key = 'a';
    while (key != )
    {
        // read frame  
        if (oniColorStream.readFrame(&oniColorImg) == STATUS_OK)
        {
            // convert data into OpenCV type  
            cv::Mat cvRGBImg(oniColorImg.getHeight(), oniColorImg.getWidth(), CV_8UC3, (void*)oniColorImg.getData());
            cv::cvtColor(cvRGBImg, cvBGRImg, CV_RGB2BGR);
            flip(cvBGRImg, cvBGRImg, );
            cv::imshow("rgb", cvBGRImg);

        }

        //if (oniDepthStream.readFrame(&oniDepthImg) == STATUS_OK)
        //{
        //  cv::Mat cvRawImg16U(oniDepthImg.getHeight(), oniDepthImg.getWidth(), CV_16UC1, (void*)oniDepthImg.getData());
        //  //cvRawImg16U.convertTo(cvDepthImg, CV_8U, 255.0 / (oniDepthStream.getMaxPixelValue()));
        //  cvRawImg16U.convertTo(cvDepthImg, CV_8U, DEPTH_SCALE_FACTOR);
        //  //cvRawImg16U.convertTo(cvDepthImg, CV_8U);

        //  //inverse depth color
        //  /*for (int i = 0; i < cvDepthImg.rows; i++)
        //  {
        //      uchar* pData = cvDepthImg.ptr<uchar>(i);
        //      for (int j = 0; j < cvDepthImg.cols ; j++)
        //      {
        //          pData[j] = cv::saturate_cast<uchar>(255-pData[j]);
        //      }
        //  }*/

        //  //【5】  
        //  // convert depth image GRAY to BGR  
        //  //cv::cvtColor(cvDepthImg, cvFusionImg, CV_GRAY2BGR);
        //  cv::imshow("depth", cvDepthImg);
        //}



        //【6】  
        //cv::addWeighted(cvBGRImg, 0.5, cvFusionImg, 1.0, 0, cvFusionImg);
        //cv::imshow("fusion", cvFusionImg);

        key = waitKey();
        if (key == 's') {
            //save image to certain path
            /*stringstream stream;
            stream << frame_index;
            string frame_index_str = stream.str();*/

            /*string rgb_path = path_of_rgb + "\\";
            rgb_path += frame_index_str + ".jpg";*/

            /*string depth_path = path_of_depth + "\\";
            depth_path += frame_index_str + ".jpg";*/

            //string fusion_path = path_of_fusion + "\\";
            //fusion_path += frame_index_str + ".jpg";

            /*string ir_path = path_of_ir + "\\";
            ir_path += frame_index_str + ".jpg";*/

            strcpy(path, pathOfAll);
            sprintf(path_rgb, "%04d_rgb.jpg", frame_index);
            strcat(path, path_rgb);
            cout <<"rgb:"<< path << endl;
            imwrite(path, cvBGRImg);
            //imwrite(depth_path, cvDepthImg);
            //imwrite(fusion_path, cvFusionImg);

            oniColorStream.stop();
            oniIrStream.start();
            if (oniIrStream.readFrame(&oniIrImg) == STATUS_OK)
            {
                cv::Mat cvRawImg16U(oniIrImg.getHeight(), oniIrImg.getWidth(), CV_16UC1, (void*)oniIrImg.getData());
                cvRawImg16U.convertTo(cvIrImg, CV_8U);
                flip(cvIrImg, cvIrImg, );
                cv::imshow("Ir", cvIrImg);
            }
            strcpy(path, pathOfAll);
            sprintf(path_ir, "%04d_ir.jpg", frame_index);
            strcat(path, path_ir);
            cout << "ir:" << path << endl;
            imwrite(path, cvIrImg);
            cout <<"Index "<< frame_index << " rgb and ir image saved." << endl;
            oniIrStream.stop();
            oniColorStream.start();
            frame_index++;
        }


    }

    //cv destroy  
    //cv::destroyWindow("depth");
    cv::destroyWindow("rgb");
    //cv::destroyWindow("fusion");
    cv::destroyWindow("Ir");

    //OpenNI2 destroy  
    //oniDepthStream.destroy();
    oniColorStream.destroy();
    oniIrStream.destroy();
    device.close();
    OpenNI::shutdown();

    return ;
}
Kinect OpenNI2
上一篇: OpenNI2 開發者指南 總覽 OpenNI 類 裝置類（Device Class） 重放控制類(PlaybackControl Class) 視訊流類(VideoStream Class) 視訊幀引用類（VideoFrameRef Class） 記錄器類（Recorder Class）
下一篇: 迷你搜尋引擎

繼續閱讀