opencv + yolov3

#include <opencv2/opencv.hpp>
#include <vector>

using namespace std;
using namespace cv;
using namespace cv::dnn;

// Initialize the parameters
float confThreshold = 0.2; // Confidence threshold
float nmsThreshold = 0.4;  // Non-maximum suppression threshold
int inpWidth = 416;        // Width of network's input image
int inpHeight = 416;       // Height of network's input image
vector<string> classes;

                           // Draw the predicted bounding box
void drawPred(int classId, float conf, int left, int top, int right, int bottom, Mat& frame)
{
    //Draw a rectangle displaying the bounding box
    rectangle(frame, Point(left, top), Point(right, bottom), Scalar(0, 0, 255));

    //Get the label for the class name and its confidence
    string label = format("%.2f", conf);
    if (!classes.empty())
    {
        CV_Assert(classId < (int)classes.size());
        label = classes[classId] + ":" + label;
    }

    //Display the label at the top of the bounding box
    int baseLine;
    Size labelSize = getTextSize(label, FONT_HERSHEY_SIMPLEX, 0.5, 1, &baseLine);
    top = max(top, labelSize.height);
    putText(frame, label, Point(left, top), FONT_HERSHEY_SIMPLEX, 0.5, Scalar(255, 255, 255));

}

                           // Get the names of the output layers
vector<String> getOutputsNames(const Net& net)
{
    static vector<String> names;
    if (names.empty())
    {
        //Get the indices of the output layers, i.e. the layers with unconnected outputs
        vector<int> outLayers = net.getUnconnectedOutLayers();

        //get the names of all the layers in the network
        vector<String> layersNames = net.getLayerNames();

        // Get the names of the output layers in names
        names.resize(outLayers.size());
        for (size_t i = 0; i < outLayers.size(); ++i)
            names[i] = layersNames[outLayers[i] - 1];
    }
    return names;
}

// Remove the bounding boxes with low confidence using non-maxima suppression
void postprocess(Mat& frame, const vector<Mat>& outs)
{
    vector<int> classIds;
    vector<float> confidences;
    vector<Rect> boxes;

    for (size_t i = 0; i < outs.size(); ++i)
    {
        // Scan through all the bounding boxes output from the network and keep only the
        // ones with high confidence scores. Assign the box's class label as the class
        // with the highest score for the box.
        float* data = (float*)outs[i].data;
        for (int j = 0; j < outs[i].rows; ++j, data += outs[i].cols)
        {
            Mat scores = outs[i].row(j).colRange(5, outs[i].cols);
            Point classIdPoint;
            double confidence;
            // Get the value and location of the maximum score
            minMaxLoc(scores, 0, &confidence, 0, &classIdPoint);
            if (confidence > confThreshold)
            {
                int centerX = (int)(data[0] * frame.cols);
                int centerY = (int)(data[1] * frame.rows);
                int width = (int)(data[2] * frame.cols);
                int height = (int)(data[3] * frame.rows);
                int left = centerX - width / 2;
                int top = centerY - height / 2;

                classIds.push_back(classIdPoint.x);
                confidences.push_back((float)confidence);
                boxes.push_back(Rect(left, top, width, height));
            }
        }
    }

    // Perform non maximum suppression to eliminate redundant overlapping boxes with
    // lower confidences
    vector<int> indices;
    NMSBoxes(boxes, confidences, confThreshold, nmsThreshold, indices);
    for (size_t i = 0; i < indices.size(); ++i)
    {
        int idx = indices[i];
        Rect box = boxes[idx];
        drawPred(classIds[idx], confidences[idx], box.x, box.y,box.x + box.width, box.y + box.height, frame);
    }
}

void main()
{

    // Load names of classes
    string classesFile = "E:\\data\\label2\\data\\obj.names";
    ifstream ifs(classesFile.c_str());
    string line;
    
    while (getline(ifs, line)) classes.push_back(line);

    // Give the configuration and weight files for the model
    String modelConfiguration = "E:\\data\\label\\yolov3-tiny-1-4.cfg";
    String modelWeights = "E:\\data\\label2\\backup\\yolov3-tiny-1-4_817000.weights";

    // Load the network
    Net net = readNetFromDarknet(modelConfiguration, modelWeights);
    net.setPreferableBackend(DNN_BACKEND_OPENCV);
    net.setPreferableTarget(DNN_TARGET_CPU);

    cv::VideoCapture cap(0);
    cv::Mat frame,blob;
    while (waitKey(1) < 0)
    {
        // get frame from the video
        cap >> frame;
        //std::cout << "frame.size=" << frame.size() << "\n";
        // Create a 4D blob from a frame.
        cv::Mat sml;
        cv::resize(frame, sml, cv::Size(0, 0), 3.0/2, 3/2.0);
        blobFromImage(sml, blob, 1 / 255.0, cvSize(inpWidth, inpHeight), Scalar(0, 0, 0), true, false);

        //Sets the input to the network
        net.setInput(blob);

        // Runs the forward pass to get output of the output layers
        vector<Mat> outs;
        net.forward(outs, getOutputsNames(net));

        // Remove the bounding boxes with low confidence
        postprocess(frame, outs);

        // Put efficiency information. The function getPerfProfile returns the
        // overall time for inference(t) and the timings for each of the layers(in layersTimes)
        vector<double> layersTimes;
        double freq = getTickFrequency() / 1000;
        double t = net.getPerfProfile(layersTimes) / freq;
        string label = format("Inference time for a frame : %.2f ms", t);
        putText(frame, label, Point(0, 15), FONT_HERSHEY_SIMPLEX, 0.5, Scalar(0, 0, 255));

        // Write the frame with the detection boxes
        Mat detectedFrame;
        frame.convertTo(detectedFrame, CV_8U);
        
        cv::imshow("detectedFrame", detectedFrame);
        //cv::waitKey(10);

    }

}