Method ParseOldYOLO

src/Detector/OCVDNNDetector.cpp:438–507

\brief OCVDNNDetector::ParseOldYOLO \param crop \param detections \param tmpRegions

/// \param tmpRegions
///
void OCVDNNDetector::ParseOldYOLO(const cv::Rect& crop, const std::vector<cv::Mat>& detections, regions_t& tmpRegions)
{
    if (m_outLayerTypes[0] == "DetectionOutput")
    {
        // Network produces output blob with a shape 1x1xNx7 where N is a number of detections and an every detection is a vector of values
        // [batchId, classId, confidence, left, top, right, bottom]
        CV_Assert(detections.size() > 0);
        for (size_t k = 0; k < detections.size(); ++k)
        {
            const float* data = reinterpret_cast<float*>(detections[k].data);
            for (size_t i = 0; i < detections[k].total(); i += 7)
            {
                float confidence = data[i + 2];
                if (confidence > m_confidenceThreshold)
                {
                    int left = (int)data[i + 3];
                    int top = (int)data[i + 4];
                    int right = (int)data[i + 5];
                    int bottom = (int)data[i + 6];
                    int width = right - left + 1;
                    int height = bottom - top + 1;
                    if (width <= 2 || height <= 2)
                    {
                        left = cvRound(data[i + 3] * crop.width);
                        top = cvRound(data[i + 4] * crop.height);
                        right = cvRound(data[i + 5] * crop.width);
                        bottom = cvRound(data[i + 6] * crop.height);
                        width = right - left + 1;
                        height = bottom - top + 1;
                    }
                    size_t objectClass = (int)(data[i + 1]) - 1;
                    if (m_classesWhiteList.empty() || m_classesWhiteList.find(T2T(objectClass)) != std::end(m_classesWhiteList))
                        tmpRegions.emplace_back(cv::Rect(left + crop.x, top + crop.y, width, height), T2T(objectClass), confidence);
                }
            }
        }
    }
    else if (m_outLayerTypes[0] == "Region")
    {
        for (size_t i = 0; i < detections.size(); ++i)
        {
            // Network produces output blob with a shape NxC where N is a number of detected objects and C is a number of classes + 4 where the first 4
            // numbers are [center_x, center_y, width, height]
            const float* data = reinterpret_cast<float*>(detections[i].data);
            for (int j = 0; j < detections[i].rows; ++j, data += detections[i].cols)
            {
                cv::Mat scores = detections[i].row(j).colRange(5, detections[i].cols);
                cv::Point classIdPoint;
                double confidence = 0;
                cv::minMaxLoc(scores, 0, &confidence, 0, &classIdPoint);
                if (confidence > m_confidenceThreshold)
                {
                    int centerX = cvRound(data[0] * crop.width);
                    int centerY = cvRound(data[1] * crop.height);
                    int width = cvRound(data[2] * crop.width);
                    int height = cvRound(data[3] * crop.height);
                    int left = centerX - width / 2;
                    int top = centerY - height / 2;