| 39 | // ---------------------------------------------------------------------------------------- |
| 40 | |
| 41 | int main(int argc, char** argv) |
| 42 | { |
| 43 | |
| 44 | try |
| 45 | { |
| 46 | // In this example we are going to train a face detector based on the |
| 47 | // small faces dataset in the examples/faces directory. So the first |
| 48 | // thing we do is load that dataset. This means you need to supply the |
| 49 | // path to this faces folder as a command line argument so we will know |
| 50 | // where it is. |
| 51 | if (argc != 2) |
| 52 | { |
| 53 | cout << "Give the path to the examples/faces directory as the argument to this" << endl; |
| 54 | cout << "program. For example, if you are in the examples folder then execute " << endl; |
| 55 | cout << "this program by running: " << endl; |
| 56 | cout << " ./fhog_object_detector_ex faces" << endl; |
| 57 | cout << endl; |
| 58 | return 0; |
| 59 | } |
| 60 | const std::string faces_directory = argv[1]; |
| 61 | // The faces directory contains a training dataset and a separate |
| 62 | // testing dataset. The training data consists of 4 images, each |
| 63 | // annotated with rectangles that bound each human face. The idea is |
| 64 | // to use this training data to learn to identify human faces in new |
| 65 | // images. |
| 66 | // |
| 67 | // Once you have trained an object detector it is always important to |
| 68 | // test it on data it wasn't trained on. Therefore, we will also load |
| 69 | // a separate testing set of 5 images. Once we have a face detector |
| 70 | // created from the training data we will see how well it works by |
| 71 | // running it on the testing images. |
| 72 | // |
| 73 | // So here we create the variables that will hold our dataset. |
| 74 | // images_train will hold the 4 training images and face_boxes_train |
| 75 | // holds the locations of the faces in the training images. So for |
| 76 | // example, the image images_train[0] has the faces given by the |
| 77 | // rectangles in face_boxes_train[0]. |
| 78 | dlib::array<array2d<unsigned char> > images_train, images_test; |
| 79 | std::vector<std::vector<rectangle> > face_boxes_train, face_boxes_test; |
| 80 | |
| 81 | // Now we load the data. These XML files list the images in each |
| 82 | // dataset and also contain the positions of the face boxes. Obviously |
| 83 | // you can use any kind of input format you like so long as you store |
| 84 | // the data into images_train and face_boxes_train. But for convenience |
| 85 | // dlib comes with tools for creating and loading XML image dataset |
| 86 | // files. Here you see how to load the data. To create the XML files |
| 87 | // you can use the imglab tool which can be found in the tools/imglab |
| 88 | // folder. It is a simple graphical tool for labeling objects in images |
| 89 | // with boxes. To see how to use it read the tools/imglab/README.txt |
| 90 | // file. |
| 91 | load_image_dataset(images_train, face_boxes_train, faces_directory+"/training.xml"); |
| 92 | load_image_dataset(images_test, face_boxes_test, faces_directory+"/testing.xml"); |
| 93 | |
| 94 | // Now we do a little bit of pre-processing. The first thing we do is |
| 95 | // increase the size of the images by a factor of two. We do this |
| 96 | // because it will allow us to detect smaller faces than otherwise would |
| 97 | // be practical (since the faces are all now twice as big). Note that, |
| 98 | // in addition to resizing the images, these functions also make the |
nothing calls this directly
no test coverage detected