Load Caffe framework models

Introduction

In this tutorial you will learn how to use opencv_dnn module for image classification by using GoogLeNet trained network from Caffe model zoo.

We will demonstrate results of this example on the following picture.

Buran space shuttle

Source Code

We will be using snippets from the example application, that can be downloaded here.

#include <opencv2/dnn.hpp>
#include <opencv2/imgproc.hpp>
#include <opencv2/highgui.hpp>
#include <opencv2/core/utils/trace.hpp>
using namespace cv;
using namespace cv::dnn;

#include <fstream>
#include <iostream>
#include <cstdlib>
using namespace std;

/* Find best class for the blob (i. e. class with maximal probability) */
static void getMaxClass(const Mat &probBlob, int *classId, double *classProb)
{
    Mat probMat = probBlob.reshape(1, 1); //reshape the blob to 1x1000 matrix
    Point classNumber;

    minMaxLoc(probMat, NULL, classProb, NULL, &classNumber);
    *classId = classNumber.x;
}

static std::vector<String> readClassNames(const char *filename )
{
    std::vector<String> classNames;

    std::ifstream fp(filename);
    if (!fp.is_open())
    {
        std::cerr << "File with classes labels not found: " << filename << std::endl;
        exit(-1);
    }

    std::string name;
    while (!fp.eof())
    {
        std::getline(fp, name);
        if (name.length())
            classNames.push_back( name.substr(name.find(' ')+1) );
    }

    fp.close();
    return classNames;
}

const char* params
    = "{ help           | false | Sample app for loading googlenet model }"
      "{ proto          | bvlc_googlenet.prototxt | model configuration }"
      "{ model          | bvlc_googlenet.caffemodel | model weights }"
      "{ label          | synset_words.txt | names of ILSVRC2012 classes }"
      "{ image          | space_shuttle.jpg | path to image file }"
      "{ opencl         | false | enable OpenCL }"
;

int main(int argc, char **argv)
{
    CV_TRACE_FUNCTION();

    CommandLineParser parser(argc, argv, params);

    if (parser.get<bool>("help"))
    {
        parser.printMessage();
        return 0;
    }

    String modelTxt = parser.get<string>("proto");
    String modelBin = parser.get<string>("model");
    String imageFile = parser.get<String>("image");
    String classNameFile = parser.get<String>("label");

    Net net;
    try {
        net = dnn::readNetFromCaffe(modelTxt, modelBin);
    }
    catch (const cv::Exception& e) {
        std::cerr << "Exception: " << e.what() << std::endl;
        if (net.empty())
        {
            std::cerr << "Can't load network by using the following files: " << std::endl;
            std::cerr << "prototxt:   " << modelTxt << std::endl;
            std::cerr << "caffemodel: " << modelBin << std::endl;
            std::cerr << "bvlc_googlenet.caffemodel can be downloaded here:" << std::endl;
            std::cerr << "http://dl.caffe.berkeleyvision.org/bvlc_googlenet.caffemodel" << std::endl;
            exit(-1);
        }
    }

    if (parser.get<bool>("opencl"))
    {
        net.setPreferableTarget(DNN_TARGET_OPENCL);
    }

    Mat img = imread(imageFile);
    if (img.empty())
    {
        std::cerr << "Can't read image from the file: " << imageFile << std::endl;
        exit(-1);
    }

    //GoogLeNet accepts only 224x224 BGR-images
    Mat inputBlob = blobFromImage(img, 1.0f, Size(224, 224),
                                  Scalar(104, 117, 123), false);   //Convert Mat to batch of images
    net.setInput(inputBlob, "data");        //set the network input
    Mat prob = net.forward("prob");         //compute output

    cv::TickMeter t;
    for (int i = 0; i < 10; i++)
    {
        CV_TRACE_REGION("forward");
        net.setInput(inputBlob, "data");        //set the network input
        t.start();
        prob = net.forward("prob");                          //compute output
        t.stop();
    }

    int classId;
    double classProb;
    getMaxClass(prob, &classId, &classProb);//find the best class

    std::vector<String> classNames = readClassNames(classNameFile.c_str());
    std::cout << "Best class: #" << classId << " '" << classNames.at(classId) << "'" << std::endl;
    std::cout << "Probability: " << classProb * 100 << "%" << std::endl;
    std::cout << "Time: " << (double)t.getTimeMilli() / t.getCounter() << " ms (average from " << t.getCounter() << " iterations)" << std::endl;

    return 0;
} //main

Explanation

1. Firstly, download GoogLeNet model files: bvlc_googlenet.prototxt and bvlc_googlenet.caffemodel

   Also you need file with names of ILSVRC2012 classes: synset_words.txt.

   Put these files into working dir of this program example.

2. Read and initialize network using path to .prototxt and .caffemodel files

           net = dnn::readNetFromCaffe(modelTxt, modelBin);

3. Check that network was read successfully

           if (net.empty())
           {
               std::cerr << "Can't load network by using the following files: " << std::endl;
               std::cerr << "prototxt:   " << modelTxt << std::endl;
               std::cerr << "caffemodel: " << modelBin << std::endl;
               std::cerr << "bvlc_googlenet.caffemodel can be downloaded here:" << std::endl;
               std::cerr << "http://dl.caffe.berkeleyvision.org/bvlc_googlenet.caffemodel" << std::endl;
               exit(-1);
           }

4. Read input image and convert to the blob, acceptable by GoogleNet

       Mat img = imread(imageFile);
       if (img.empty())
       {
           std::cerr << "Can't read image from the file: " << imageFile << std::endl;
           exit(-1);
       }
   
       //GoogLeNet accepts only 224x224 BGR-images
       Mat inputBlob = blobFromImage(img, 1.0f, Size(224, 224),
                                     Scalar(104, 117, 123), false);   //Convert Mat to batch of images

   We convert the image to a 4-dimensional blob (so-called batch) with 1x3x224x224 shape after applying necessary pre-processing like resizing and mean subtraction using cv::dnn::blobFromImage constructor.

5. Pass the blob to the network

           net.setInput(inputBlob, "data");        //set the network input

   In bvlc_googlenet.prototxt the network input blob named as "data", therefore this blob labeled as ".data" in opencv_dnn API.

   Other blobs labeled as "name_of_layer.name_of_layer_output".

6. Make forward pass

           prob = net.forward("prob");                          //compute output

   During the forward pass output of each network layer is computed, but in this example we need output from "prob" layer only.
7. Determine the best class

       int classId;
       double classProb;
       getMaxClass(prob, &classId, &classProb);//find the best class

   We put the output of "prob" layer, which contain probabilities for each of 1000 ILSVRC2012 image classes, to the prob blob. And find the index of element with maximal value in this one. This index correspond to the class of the image.
8. Print results

       std::vector<String> classNames = readClassNames(classNameFile.c_str());
       std::cout << "Best class: #" << classId << " '" << classNames.at(classId) << "'" << std::endl;
       std::cout << "Probability: " << classProb * 100 << "%" << std::endl;

   For our image we get:

   Best class: #812 'space shuttle'

   Probability: 99.6378%