Qt调用opencv实现yolov3对视频进行目标检测
依赖：支持CUDA的opencv4.3.0，demo.cfg，demo_final.weights，demo.names
demo.cfg，demo.names是darknet训练用的配置文件，demo_final.weights是训练后的权重文件。
使用GPU或CPU的代码：
//使用GPU检测
#ifdef Enable_GPU
net.setPreferableTarget(DNN_TARGET_CUDA);
net.setPreferableBackend(DNN_BACKEND_CUDA);
//使用CPU检测
#else
net.setPreferableTarget(DNN_TARGET_CPU);
net.setPreferableBackend(DNN_BACKEND_OPENCV);
#endif一般opencv关闭窗口后会马上重新出现，就像有守护进程一样。这样就不能点击关闭按钮退出程序。为了点击关闭按钮就能退出程序，需要在while循环内加这段代码：
// opencv点击关闭按钮则关闭窗口
if (getWindowProperty(kWinName, WND_PROP_VISIBLE) < 1)
{
break;
}完整代码：
#include
#include
#include
#include
#include
#include
#include
#include
#include
#include
#include
using namespace cv;
using namespace dnn;
using namespace std;
string pro_dir = "./input/";
#define Enable_GPU
// Initialize the parameters
//float confThreshold = 0.5; // Confidence threshold
float confThreshold = 0.4; // 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 classes;
// Remove the bounding boxes with low confidence using non-maxima suppression
void postprocess(Mat& frame, const vector& out);
// Draw the predicted bounding box
void drawPred(int classId, float conf, int left, int top, int right, int bottom, Mat& frame);
// Get the names of the output layers
vector getOutputsNames(const Net& net);
void detect_image(string image_path, string modelWeights, string modelConfiguration, string classesFile);
void detect_video(string video_path, string modelWeights, string modelConfiguration, string classesFile);
void dnnM()
{
QString dir = QString("%1/%2").arg(QApplication::applicationDirPath()).arg("input");
String modelConfiguration = QString("%1/%2").arg(dir).arg("demo.cfg").toLocal8Bit();
String modelWeights = QString("%1/%2").arg(dir).arg("demo_final.weights").toLocal8Bit();
String classesFile = QString("%1/%2").arg(dir).arg("demo.names").toLocal8Bit();
String image_path = QString("%1/%2").arg(dir).arg("ammeter.png").toStdString();
String video_path = QString("%1/%2").arg(dir).arg(QStringLiteral("demo.mp4")).toLocal8Bit();
detect_video(video_path, modelWeights, modelConfiguration, classesFile);
}
void detect_image(string image_path, string modelWeights, string modelConfiguration, string classesFile) {
// Load names of classes
ifstream ifs(classesFile.c_str());
string line;
while (getline(ifs, line)) classes.push_back(line);
// Load the network
Net net = readNetFromDarknet(modelConfiguration, modelWeights);
net.setPreferableBackend(DNN_BACKEND_OPENCV);
net.setPreferableTarget(DNN_TARGET_OPENCL);
// Open a video file or an image file or a camera stream.
string str, outputFile;
cv::Mat frame = cv::imread(image_path);
// Create a window
static const string kWinName = "Deep learning object detection in OpenCV";
namedWindow(kWinName, WINDOW_NORMAL);
// Stop the program if reached end of video
// Create a 4D blob from a frame.
Mat blob;
blobFromImage(frame, blob, 1 / 255.0, Size(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 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 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
imshow(kWinName, frame);
cv::waitKey();
}
void detect_video(string video_path, string modelWeights, string modelConfiguration, string classesFile) {
string outputFile = "./out.avi";;
// Load names of classes
ifstream ifs(classesFile.c_str());
string line;
while (getline(ifs, line)) classes.push_back(line);
// Load the network
Net net = readNetFromDarknet(modelConfiguration, modelWeights);
//使用GPU检测
#ifdef Enable_GPU
net.setPreferableTarget(DNN_TARGET_CUDA);
net.setPreferableBackend(DNN_BACKEND_CUDA);
//使用CPU检测
#else
net.setPreferableTarget(DNN_TARGET_CPU);
net.setPreferableBackend(DNN_BACKEND_OPENCV);
#endif
// Open a video file or an image file or a camera stream.
VideoCapture cap;
VideoWriter video;
Mat frame, blob;
try {
// Open the video file
ifstream ifile(video_path);
if (!ifile) throw("error");
cap.open(video_path);
//video.open(outputFile, VideoWriter::fourcc('M', 'J', 'P', 'G'), 25, Size(cap.get(CAP_PROP_FRAME_WIDTH), cap.get(CAP_PROP_FRAME_HEIGHT)));
bool b = video.open(outputFile, VideoWriter::fourcc('M', 'P', '4', '2'), 25, Size(cap.get(CAP_PROP_FRAME_WIDTH), cap.get(CAP_PROP_FRAME_HEIGHT)));
}
catch (...) {
cout << "Could not open the input image/video stream" << endl;
return;
}
static const string kWinName = "demo";
namedWindow(kWinName, WINDOW_NORMAL);
while (waitKey(1) < 0)
{
// opencv点击关闭按钮则关闭窗口
if (getWindowProperty(kWinName, WND_PROP_VISIBLE) < 1)
{
break;
}
// get frame from the video
cap >> frame;
// Stop the program if reached end of video
if (frame.empty()) {
cout << "Done processing !!!" << endl;
cout << "Output file is stored as " << outputFile << endl;
waitKey(3000);
break;
}
// Create a 4D blob from a frame.
blobFromImage(frame, blob, 1 / 255.0, Size(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 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 layersTimes;
double freq = getTickFrequency() / 1000;
double t = net.getPerfProfile(layersTimes) / freq;
double FPS = 1 / (t / 1000);
string label = format("FPS: %.2f", FPS);
//rectangle(frame, Size(0, 0), Size(100, 20), Scalar(0, 0, 0), 0);
//cv::line(frame, Size(0, 0), Size(100, 0), Scalar(0, 0, 0), 0);
putText(frame, label, Point(0, 15), FONT_HERSHEY_SIMPLEX, 0.5, Scalar(255, 255, 255), 1);
// Write the frame with the detection boxes
Mat detectedFrame;
frame.convertTo(detectedFrame, CV_8U);
video.write(detectedFrame);
//imshow(kWinName, frame);
imshow(kWinName, detectedFrame);
}
cap.release();
video.release();
cv::destroyWindow(kWinName);
}
// Remove the bounding boxes with low confidence using non-maxima suppression
void postprocess(Mat& frame, const vector& outs)
{
vector classIds;
vector confidences;
vector 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 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);
}
}
// Draw the predicted bounding box
void drawPred(int classId, float conf, int left, int top, int right, int bottom, Mat& frame)
{
string type;
//Get the label for the class name and its confidence
string label = format("%.2f", conf);
if (!classes.empty())
{
CV_Assert(classId < (int)classes.size());
type = classes[classId];
label = type + ":" + label;
}
Scalar scalar;
switch (classId)
{
case 0:
scalar = Scalar(255, 104, 0);
break;
case 1:
scalar = Scalar(255, 39, 255);
break;
case 2:
scalar = Scalar(255, 31, 107);
break;
case 3:
scalar = Scalar(255, 239, 0);
break;
case 4:
scalar = Scalar(0, 122, 254);
break;
case 5:
scalar = Scalar(131, 216, 0);
break;
case 6:
scalar = Scalar(0, 226, 189);
break;
case 7:
scalar = Scalar(250, 25, 19);
break;
case 8:
scalar = Scalar(111, 111, 22);
break;
default:
scalar = Scalar(255, 104, 22);
break;
}
//Draw a rectangle displaying the bounding box
rectangle(frame, Point(left, top), Point(right, bottom), scalar, 2);
//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);
baseLine -= 1;
rectangle(frame, Point(left, top - 2 * labelSize.height - baseLine), Point(left + round(labelSize.width), top - baseLine + 2), scalar, FILLED);
putText(frame, label, Point(left, top - 0.5 * labelSize.height - baseLine - 1), FONT_HERSHEY_SIMPLEX, 0.5, Scalar(0, 0, 0), 1);
}
// Get the names of the output layers
vector getOutputsNames(const Net& net)
{
static vector names;
if (names.empty())
{
//Get the indices of the output layers, i.e. the layers with unconnected outputs
vector outLayers = net.getUnconnectedOutLayers();
//get the names of all the layers in the network
vector 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;
}
int main(int argc, char* argv[])
{
//QApplication a(argc, argv);
dnnM();
//return a.exec();
return 1;
}
实测：
1650s显卡下跑这个程序，FPS在33左右，比CPU 5帧好多了。但是很奇怪的是我用darknet检测的时候FPS只有7，实在是不知道哪里有问题。
2070s：FPS 65 。