2018-03-03 21:43:21 +08:00
|
|
|
#include <fstream>
|
|
|
|
#include <sstream>
|
|
|
|
|
|
|
|
#include <opencv2/dnn.hpp>
|
2018-03-04 00:29:37 +08:00
|
|
|
#include <opencv2/imgproc.hpp>
|
|
|
|
#include <opencv2/highgui.hpp>
|
2018-03-03 21:43:21 +08:00
|
|
|
|
|
|
|
const char* keys =
|
|
|
|
"{ help h | | Print help message. }"
|
|
|
|
"{ input i | | Path to input image or video file. Skip this argument to capture frames from a camera.}"
|
|
|
|
"{ model m | | Path to a binary file of model contains trained weights. "
|
|
|
|
"It could be a file with extensions .caffemodel (Caffe), "
|
|
|
|
".pb (TensorFlow), .t7 or .net (Torch), .weights (Darknet) }"
|
|
|
|
"{ config c | | Path to a text file of model contains network configuration. "
|
|
|
|
"It could be a file with extensions .prototxt (Caffe), .pbtxt (TensorFlow), .cfg (Darknet) }"
|
|
|
|
"{ framework f | | Optional name of an origin framework of the model. Detect it automatically if it does not set. }"
|
|
|
|
"{ classes | | Optional path to a text file with names of classes. }"
|
|
|
|
"{ mean | | Preprocess input image by subtracting mean values. Mean values should be in BGR order and delimited by spaces. }"
|
2018-03-07 00:29:23 +08:00
|
|
|
"{ scale | 1 | Preprocess input image by multiplying on a scale factor. }"
|
|
|
|
"{ width | | Preprocess input image by resizing to a specific width. }"
|
|
|
|
"{ height | | Preprocess input image by resizing to a specific height. }"
|
|
|
|
"{ rgb | | Indicate that model works with RGB input images instead BGR ones. }"
|
|
|
|
"{ backend | 0 | Choose one of computation backends: "
|
|
|
|
"0: default C++ backend, "
|
|
|
|
"1: Halide language (http://halide-lang.org/), "
|
|
|
|
"2: Intel's Deep Learning Inference Engine (https://software.seek.intel.com/deep-learning-deployment)}"
|
|
|
|
"{ target | 0 | Choose one of target computation devices: "
|
|
|
|
"0: CPU target (by default),"
|
|
|
|
"1: OpenCL }";
|
2018-03-03 21:43:21 +08:00
|
|
|
|
|
|
|
using namespace cv;
|
|
|
|
using namespace dnn;
|
|
|
|
|
|
|
|
std::vector<std::string> classes;
|
|
|
|
|
|
|
|
int main(int argc, char** argv)
|
|
|
|
{
|
|
|
|
CommandLineParser parser(argc, argv, keys);
|
|
|
|
parser.about("Use this script to run classification deep learning networks using OpenCV.");
|
|
|
|
if (argc == 1 || parser.has("help"))
|
|
|
|
{
|
|
|
|
parser.printMessage();
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
|
|
|
float scale = parser.get<float>("scale");
|
2018-03-07 00:29:23 +08:00
|
|
|
Scalar mean = parser.get<Scalar>("mean");
|
2018-03-03 21:43:21 +08:00
|
|
|
bool swapRB = parser.get<bool>("rgb");
|
2018-03-07 00:29:23 +08:00
|
|
|
CV_Assert(parser.has("width"), parser.has("height"));
|
2018-03-03 21:43:21 +08:00
|
|
|
int inpWidth = parser.get<int>("width");
|
|
|
|
int inpHeight = parser.get<int>("height");
|
2018-03-04 00:29:37 +08:00
|
|
|
String model = parser.get<String>("model");
|
|
|
|
String config = parser.get<String>("config");
|
|
|
|
String framework = parser.get<String>("framework");
|
|
|
|
int backendId = parser.get<int>("backend");
|
|
|
|
int targetId = parser.get<int>("target");
|
2018-03-03 21:43:21 +08:00
|
|
|
|
|
|
|
// Open file with classes names.
|
|
|
|
if (parser.has("classes"))
|
|
|
|
{
|
|
|
|
std::string file = parser.get<String>("classes");
|
|
|
|
std::ifstream ifs(file.c_str());
|
|
|
|
if (!ifs.is_open())
|
|
|
|
CV_Error(Error::StsError, "File " + file + " not found");
|
|
|
|
std::string line;
|
|
|
|
while (std::getline(ifs, line))
|
|
|
|
{
|
|
|
|
classes.push_back(line);
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
CV_Assert(parser.has("model"));
|
2018-03-04 00:29:37 +08:00
|
|
|
//! [Read and initialize network]
|
|
|
|
Net net = readNet(model, config, framework);
|
|
|
|
net.setPreferableBackend(backendId);
|
|
|
|
net.setPreferableTarget(targetId);
|
|
|
|
//! [Read and initialize network]
|
2018-03-03 21:43:21 +08:00
|
|
|
|
|
|
|
// Create a window
|
|
|
|
static const std::string kWinName = "Deep learning image classification in OpenCV";
|
|
|
|
namedWindow(kWinName, WINDOW_NORMAL);
|
|
|
|
|
2018-03-04 00:29:37 +08:00
|
|
|
//! [Open a video file or an image file or a camera stream]
|
2018-03-03 21:43:21 +08:00
|
|
|
VideoCapture cap;
|
|
|
|
if (parser.has("input"))
|
|
|
|
cap.open(parser.get<String>("input"));
|
|
|
|
else
|
|
|
|
cap.open(0);
|
2018-03-04 00:29:37 +08:00
|
|
|
//! [Open a video file or an image file or a camera stream]
|
2018-03-03 21:43:21 +08:00
|
|
|
|
|
|
|
// Process frames.
|
|
|
|
Mat frame, blob;
|
|
|
|
while (waitKey(1) < 0)
|
|
|
|
{
|
|
|
|
cap >> frame;
|
|
|
|
if (frame.empty())
|
|
|
|
{
|
|
|
|
waitKey();
|
|
|
|
break;
|
|
|
|
}
|
|
|
|
|
2018-03-04 00:29:37 +08:00
|
|
|
//! [Create a 4D blob from a frame]
|
2018-03-03 21:43:21 +08:00
|
|
|
blobFromImage(frame, blob, scale, Size(inpWidth, inpHeight), mean, swapRB, false);
|
2018-03-04 00:29:37 +08:00
|
|
|
//! [Create a 4D blob from a frame]
|
2018-03-03 21:43:21 +08:00
|
|
|
|
2018-03-04 00:29:37 +08:00
|
|
|
//! [Set input blob]
|
2018-03-03 21:43:21 +08:00
|
|
|
net.setInput(blob);
|
2018-03-04 00:29:37 +08:00
|
|
|
//! [Set input blob]
|
|
|
|
//! [Make forward pass]
|
|
|
|
Mat prob = net.forward();
|
|
|
|
//! [Make forward pass]
|
2018-03-03 21:43:21 +08:00
|
|
|
|
2018-03-04 00:29:37 +08:00
|
|
|
//! [Get a class with a highest score]
|
2018-03-03 21:43:21 +08:00
|
|
|
Point classIdPoint;
|
|
|
|
double confidence;
|
2018-03-04 00:29:37 +08:00
|
|
|
minMaxLoc(prob.reshape(1, 1), 0, &confidence, 0, &classIdPoint);
|
2018-03-03 21:43:21 +08:00
|
|
|
int classId = classIdPoint.x;
|
2018-03-04 00:29:37 +08:00
|
|
|
//! [Get a class with a highest score]
|
2018-03-03 21:43:21 +08:00
|
|
|
|
|
|
|
// Put efficiency information.
|
|
|
|
std::vector<double> layersTimes;
|
2018-03-04 00:29:37 +08:00
|
|
|
double freq = getTickFrequency() / 1000;
|
|
|
|
double t = net.getPerfProfile(layersTimes) / freq;
|
|
|
|
std::string label = format("Inference time: %.2f ms", t);
|
2018-03-03 21:43:21 +08:00
|
|
|
putText(frame, label, Point(0, 15), FONT_HERSHEY_SIMPLEX, 0.5, Scalar(0, 255, 0));
|
|
|
|
|
|
|
|
// Print predicted class.
|
|
|
|
label = format("%s: %.4f", (classes.empty() ? format("Class #%d", classId).c_str() :
|
|
|
|
classes[classId].c_str()),
|
|
|
|
confidence);
|
|
|
|
putText(frame, label, Point(0, 40), FONT_HERSHEY_SIMPLEX, 0.5, Scalar(0, 255, 0));
|
|
|
|
|
|
|
|
imshow(kWinName, frame);
|
|
|
|
}
|
|
|
|
return 0;
|
|
|
|
}
|