代码之家  ›  专栏  ›  技术社区  ›  Mee

opencv标准视差图不工作

  •  0
  • Mee  · 技术社区  · 7 年前

    我似乎无法使用标准的opencv函数(stereoBM)获得任何类型的深度图像。

    我试过:

    Mat disp, disp8;
    
    StereoBM *sbm = StereoBM::create(16, 2);
    sbm->setDisp12MaxDiff(1);
    sbm->setSpeckleRange(8);
    sbm->setSpeckleWindowSize(0);
    sbm->setUniquenessRatio(0);
    sbm->setTextureThreshold(507);
    sbm->setMinDisparity(-39);
    sbm->setPreFilterCap(61);
    sbm->setPreFilterSize(5);
    sbm->compute(imgLeft, imgRight, disp);
    normalize(disp, disp8, 0, 255, CV_MINMAX, CV_8U);
    
    cv::imshow("disp", disp8);
    

    它可以编译,但会产生大量错误。不确定我是否正确使用了抽象类?

    谢谢

    1 回复  |  直到 7 年前
        1
  •  0
  •   Arritmic    7 年前

    这是我的功能。我希望这对你有帮助。

    cv::Mat Arritmic::DepthMap(cv::Mat &imageL, cv::Mat &imageR)
    {
    
        /// DISPARITY MAP AND DEPTH MAP
        cv::Mat left_for_matcher, right_for_matcher;
        cv::Mat left_disp,right_disp;
        cv::Mat filtered_disp;
        cv::Mat conf_map =  cv::Mat(imageL.rows, imageL.cols, CV_8U);
        conf_map =  cv::Scalar(255);
        cv::Rect ROI;
        int max_disp = 16; // n*16
        int wsize = 15;
    
    
    
    
       // Perform matching and create the filter instance
       /* I am using StereoBM for faster processing. If speed is not critical, 
       though, StereoSGBM would provide better quality.
       The filter instance is created by providing the StereoMatcher instance
       that we intend to use. Another matcher instance is returned by the
       createRightMatcher function. These two matcher instances are then used
       to compute disparity maps both for the left and right views, that are
       required by the filter. */
    
        cv::Ptr<cv::ximgproc::DisparityWLSFilter> wls_filter;
    
        cv::Ptr<cv::StereoBM >left_matcher = cv::StereoBM::create(max_disp,wsize);
        wls_filter = cv::ximgproc::createDisparityWLSFilter(left_matcher);
        cv::Ptr<cv::StereoMatcher> right_matcher = cv::ximgproc::createRightMatcher(left_matcher);
        cv::cvtColor(imageL,  left_for_matcher,  cv::COLOR_BGR2GRAY);
        cv::cvtColor(imageR, right_for_matcher, cv::COLOR_BGR2GRAY);
    
    
        left_matcher-> compute(left_for_matcher, right_for_matcher,left_disp);
        right_matcher->compute(right_for_matcher,left_for_matcher, right_disp);
    
        // Perform filtering
        /* Disparity maps computed by the respective matcher instances, as
        well as the source left view are passed to the filter. Note that we
        are using the original non-downscaled view to guide the filtering 
        process. 
        The disparity map is automatically upscaled in an edge-aware fashion
        to match the original view resolution. The result is stored in
        filtered_disp. */
    
    
        double lambda = 6000.0;   // hardcode
        double sigma = 2.0;       // hardcode
    
        //! [filtering]
        wls_filter->setLambda(lambda);
        wls_filter->setSigmaColor(sigma);
    
        wls_filter->filter(left_disp, imageL, filtered_disp, right_disp);
    
        //! [filtering]
        conf_map = wls_filter->getConfidenceMap();
    
        // Get the ROI that was used in the last filter call:
        ROI = wls_filter->getROI();
    
          cv::Mat raw_disp_vis;
        cv::ximgproc::getDisparityVis(left_disp,raw_disp_vis, 21.0);
        cv::Mat filtered_disp_vis;
        cv::ximgproc::getDisparityVis(filtered_disp,filtered_disp_vis, 15.0);
    
    
        return raw_disp_vis;  // rerturning de depth map image.
    }