如何在ubuntu（或其他Linux環境下）將realsense RGB幀轉換爲cv :: Mat？

Question

有沒有官方SDK這樣做，有人可以幫忙嗎？ 如何在ubuntu（或其他Linux環境下）將realsense RGB幀轉換爲cv :: Mat？

Answer 1

似乎是你可以創建3 cv :: Mat來表示圖像RGB +深度+紅外的每個組件。因此，它應該像

CV ::墊frameColor = CV ::墊::零（resolutionColor.height， resolutionColor.width，CV_8UC3）; cv :: Mat frameDepth = cv :: Mat :: zeros（resolutionDepth.height，resolutionDepth.width， CV_32FC1）; cv :: Mat frameIR = cv :: Mat :: zeros（resolutionIR.height， resolutionIR.width，CV_8UC1）;

檢查這個論壇 - > https://software.intel.com/en-us/forums/realsense/topic/538066和 Convert a PXCImage into an OpenCV Mat

Answer 2

一旦你的樣本圖像

PXCImage::ImageData c_image; 
cv::Mat p_image; 
PXCCapture::Sample *sample = sm->QuerySample(); 

      if (sample->color != NULL) 
      { 

       sample->color->AcquireAccess(PXCImage::ACCESS_READ_WRITE, PXCImage::PIXEL_FORMAT_RGB32, &c_image); 
       PXCImage::ImageInfo inputInfo = sample->color->QueryInfo(); 
       cv::Mat img = cv::Mat(inputInfo.height, inputInfo.width, CV_8UC4, c_image.planes[0], c_image.pitches[0]); 
       sample->color->ReleaseAccess(&c_image); 


      } 

Answer 3

如果您使用的是新librealsense SDK爲Ubuntu：那麼它是一樣簡單創建一個cv :: Mat的新實例，以獲取相機的高度和寬度以及所獲取圖像的尺寸。

cv::Mat vis = cv::Mat(1080,1920,CV_8UC3);

之後，唯一的步驟左側是分配由流提供給矩陣的vis.data屬性數據

vis.data = (uchar*)reinterpret_cast<const uint8_t *>(dev->get_frame_data(rs::stream::color));` 

的（UCHAR *）鑄造是投uint8_t格式轉換爲cv :: Mat的原生格式。

因爲Realsense流提供顏色BGR和OpenCV使用RGB來表示圖像，最後一步是將矩陣變換，以正確的格式：

cv::cvtColor(vis, vis, CV_BGR2RGB); 

Answer 4

/*** 
Returns the next frame if next frame is recorded 
Returns the previous frame if next frame is not recorded 
***/ 
void Converter::ConvertPXCImageToOpenCVMat(Intel::RealSense::Image *inImg, Intel::RealSense::ImageData data, cv::Mat *outImg) { 
    auto cvDataType = 0; 
    auto cvDataWidth = 0; 

    auto imgInfo = inImg->QueryInfo(); 

    switch (data.format) { 
     /* STREAM_TYPE_COLOR */ 
    case Intel::RealSense::Image::PIXEL_FORMAT_YUY2: /* YUY2 image */ 
    case Intel::RealSense::Image::PIXEL_FORMAT_NV12: /* NV12 image */ 
     throw; // Not implemented 
    case Intel::RealSense::Image::PIXEL_FORMAT_RGB32: /* BGRA layout on a little-endian machine */ 
     cvDataType = CV_8UC4; 
     cvDataWidth = 4; 
     break; 
    case Intel::RealSense::Image::PIXEL_FORMAT_RGB24: /* BGR layout on a little-endian machine */ 
     cvDataType = CV_8UC3; 
     cvDataWidth = 3; 
     break; 
    case Intel::RealSense::Image::PIXEL_FORMAT_Y8: /* 8-Bit Gray Image, or IR 8-bit */ 
     cvDataType = CV_8U; 
     cvDataWidth = 1; 
     break; 
     /* STREAM_TYPE_DEPTH */ 
    case Intel::RealSense::Image::PIXEL_FORMAT_DEPTH: /* 16-bit unsigned integer with precision mm. */ 
    case Intel::RealSense::Image::PIXEL_FORMAT_DEPTH_RAW: /* 16-bit unsigned integer with device specific precision (call device->QueryDepthUnit()) */ 
     cvDataType = CV_16U; 
     cvDataWidth = 2; 
     break; 
    case Intel::RealSense::Image::PIXEL_FORMAT_DEPTH_F32: /* 32-bit float-point with precision mm. */ 
     cvDataType = CV_32F; 
     cvDataWidth = 4; 
     break; 
     /* STREAM_TYPE_IR */ 
    case Intel::RealSense::Image::PIXEL_FORMAT_Y16: /* 16-Bit Gray Image */ 
     cvDataType = CV_16U; 
     cvDataWidth = 2; 
     break; 
    case Intel::RealSense::Image::PIXEL_FORMAT_Y8_IR_RELATIVE: /* Relative IR Image */ 
     cvDataType = CV_8U; 
     cvDataWidth = 1; 
     break; 
    default: 
     break; 
    } 

    // suppose that no other planes 
    if (data.planes[1] != nullptr) throw; // not implemented 
              // suppose that no sub pixel padding needed 
    if (data.pitches[0] % cvDataWidth != 0) throw; // not implemented 

    outImg->create(imgInfo.height, data.pitches[0]/cvDataWidth, cvDataType); 

    //memcpy(outImg->data, data.planes[0], imgInfo.height*imgInfo.width*cvDataWidth * sizeof(pxcBYTE)); 
    memcpy(outImg->data, data.planes[0], imgInfo.height*imgInfo.width*cvDataWidth * sizeof(uint8_t)); 
}