2011-09-18 04:23:06 +08:00
|
|
|
// The contents of this file are in the public domain. See LICENSE_FOR_EXAMPLE_PROGRAMS.txt
|
|
|
|
/*
|
|
|
|
|
2014-01-11 23:40:39 +08:00
|
|
|
This is an example illustrating the use of dlib's bag-of-visual-word based
|
|
|
|
tools for detecting objects in images. In this example we will create three
|
|
|
|
simple images, each containing some white squares. We will then use the
|
|
|
|
sliding window classifier tools to learn to detect these squares.
|
|
|
|
|
|
|
|
If the objects you want to detect are somewhat rigid in appearance (e.g.
|
|
|
|
faces, pedestrians, etc.) then you should try the methods shown in the
|
|
|
|
fhog_object_detector_ex.cpp example program before trying to use the
|
|
|
|
bag-of-visual-word tools shown in this example.
|
2011-09-18 04:23:06 +08:00
|
|
|
*/
|
|
|
|
|
|
|
|
|
2012-12-08 22:32:13 +08:00
|
|
|
#include <dlib/svm_threaded.h>
|
|
|
|
#include <dlib/gui_widgets.h>
|
|
|
|
#include <dlib/array.h>
|
|
|
|
#include <dlib/array2d.h>
|
|
|
|
#include <dlib/image_keypoint.h>
|
|
|
|
#include <dlib/image_processing.h>
|
2011-09-18 04:23:06 +08:00
|
|
|
|
|
|
|
#include <iostream>
|
|
|
|
#include <fstream>
|
|
|
|
|
|
|
|
|
|
|
|
using namespace std;
|
|
|
|
using namespace dlib;
|
|
|
|
|
|
|
|
// ----------------------------------------------------------------------------------------
|
|
|
|
|
|
|
|
template <
|
|
|
|
typename image_array_type
|
|
|
|
>
|
|
|
|
void make_simple_test_data (
|
|
|
|
image_array_type& images,
|
|
|
|
std::vector<std::vector<rectangle> >& object_locations
|
|
|
|
)
|
|
|
|
/*!
|
|
|
|
ensures
|
|
|
|
- #images.size() == 3
|
|
|
|
- #object_locations.size() == 3
|
|
|
|
- Creates some simple images to test the object detection routines. In particular,
|
|
|
|
this function creates images with white 70x70 squares in them. It also stores
|
|
|
|
the locations of these squares in object_locations.
|
|
|
|
- for all valid i:
|
|
|
|
- object_locations[i] == A list of all the white rectangles present in images[i].
|
|
|
|
!*/
|
|
|
|
{
|
|
|
|
images.clear();
|
|
|
|
object_locations.clear();
|
|
|
|
|
|
|
|
images.resize(3);
|
|
|
|
images[0].set_size(400,400);
|
|
|
|
images[1].set_size(400,400);
|
|
|
|
images[2].set_size(400,400);
|
|
|
|
|
|
|
|
// set all the pixel values to black
|
|
|
|
assign_all_pixels(images[0], 0);
|
|
|
|
assign_all_pixels(images[1], 0);
|
|
|
|
assign_all_pixels(images[2], 0);
|
|
|
|
|
|
|
|
// Now make some squares and draw them onto our black images. All the
|
|
|
|
// squares will be 70 pixels wide and tall.
|
|
|
|
|
|
|
|
std::vector<rectangle> temp;
|
|
|
|
temp.push_back(centered_rect(point(100,100), 70,70));
|
|
|
|
fill_rect(images[0],temp.back(),255); // Paint the square white
|
|
|
|
temp.push_back(centered_rect(point(200,300), 70,70));
|
|
|
|
fill_rect(images[0],temp.back(),255); // Paint the square white
|
|
|
|
object_locations.push_back(temp);
|
|
|
|
|
|
|
|
temp.clear();
|
|
|
|
temp.push_back(centered_rect(point(140,200), 70,70));
|
|
|
|
fill_rect(images[1],temp.back(),255); // Paint the square white
|
|
|
|
temp.push_back(centered_rect(point(303,200), 70,70));
|
|
|
|
fill_rect(images[1],temp.back(),255); // Paint the square white
|
|
|
|
object_locations.push_back(temp);
|
|
|
|
|
|
|
|
temp.clear();
|
|
|
|
temp.push_back(centered_rect(point(123,121), 70,70));
|
|
|
|
fill_rect(images[2],temp.back(),255); // Paint the square white
|
|
|
|
object_locations.push_back(temp);
|
2011-12-27 10:43:07 +08:00
|
|
|
|
|
|
|
// corrupt each image with random noise just to make this a little more
|
|
|
|
// challenging
|
|
|
|
dlib::rand rnd;
|
|
|
|
for (unsigned long i = 0; i < images.size(); ++i)
|
|
|
|
{
|
|
|
|
for (long r = 0; r < images[i].nr(); ++r)
|
|
|
|
{
|
|
|
|
for (long c = 0; c < images[i].nc(); ++c)
|
|
|
|
{
|
2012-01-08 02:11:02 +08:00
|
|
|
images[i][r][c] = put_in_range(0,255,images[i][r][c] + 40*rnd.get_random_gaussian());
|
2011-12-27 10:43:07 +08:00
|
|
|
}
|
|
|
|
}
|
|
|
|
}
|
2011-09-18 04:23:06 +08:00
|
|
|
}
|
|
|
|
|
|
|
|
// ----------------------------------------------------------------------------------------
|
|
|
|
|
2011-11-06 22:45:19 +08:00
|
|
|
int main()
|
2011-09-18 04:23:06 +08:00
|
|
|
{
|
|
|
|
try
|
|
|
|
{
|
|
|
|
// The first thing we do is create the set of 3 images discussed above.
|
2013-03-03 23:40:33 +08:00
|
|
|
dlib::array<array2d<unsigned char> > images;
|
2011-09-18 04:23:06 +08:00
|
|
|
std::vector<std::vector<rectangle> > object_locations;
|
|
|
|
make_simple_test_data(images, object_locations);
|
|
|
|
|
|
|
|
|
|
|
|
/*
|
|
|
|
This next block of code specifies the type of sliding window classifier we will
|
|
|
|
be using to detect the white squares. The most important thing here is the
|
|
|
|
scan_image_pyramid template. Instances of this template represent the core
|
|
|
|
of a sliding window classifier. To go into more detail, the sliding window
|
|
|
|
classifiers used by this object have three parts:
|
|
|
|
1. The underlying feature extraction. See the dlib documentation for a detailed
|
|
|
|
discussion of how the hashed_feature_image and hog_image feature extractors
|
|
|
|
work. However, to understand this example, all you need to know is that the
|
|
|
|
feature extractor associates a vector with each location in an image. This
|
|
|
|
vector is supposed to capture information which describes how parts of the
|
2012-01-02 06:55:56 +08:00
|
|
|
image look. Importantly, it should do this in a way that is relevant to the
|
|
|
|
problem you are trying to solve.
|
2011-09-18 04:23:06 +08:00
|
|
|
|
|
|
|
2. A detection template. This is a rectangle which defines the shape of a
|
2012-08-27 03:07:28 +08:00
|
|
|
sliding window (i.e. the object_box), as well as a set of rectangular feature
|
|
|
|
extraction regions inside it. This set of regions defines the spatial
|
|
|
|
structure of the overall feature extraction within a sliding window. In
|
|
|
|
particular, each location of a sliding window has a feature vector
|
2011-09-18 04:23:06 +08:00
|
|
|
associated with it. This feature vector is defined as follows:
|
2012-08-27 03:07:28 +08:00
|
|
|
- Let N denote the number of feature extraction zones.
|
|
|
|
- Let M denote the dimensionality of the vectors output by Feature_extractor_type
|
2011-09-18 04:23:06 +08:00
|
|
|
objects.
|
|
|
|
- Let F(i) == the M dimensional vector which is the sum of all vectors
|
2012-08-27 03:07:28 +08:00
|
|
|
given by our Feature_extractor_type object inside the ith feature extraction
|
|
|
|
zone.
|
2011-09-18 04:23:06 +08:00
|
|
|
- Then the feature vector for a sliding window is an M*N dimensional vector
|
|
|
|
[F(1) F(2) F(3) ... F(N)] (i.e. it is a concatenation of the N vectors).
|
|
|
|
This feature vector can be thought of as a collection of N "bags of features",
|
2012-08-27 03:07:28 +08:00
|
|
|
each bag coming from a spatial location determined by one of the rectangular
|
|
|
|
feature extraction zones.
|
2011-09-18 04:23:06 +08:00
|
|
|
|
|
|
|
3. A weight vector and a threshold value. The dot product between the weight
|
|
|
|
vector and the feature vector for a sliding window location gives the score
|
|
|
|
of the window. If this score is greater than the threshold value then the
|
|
|
|
window location is output as a detection. You don't need to determine these
|
|
|
|
parameters yourself. They are automatically populated by the
|
|
|
|
structural_object_detection_trainer.
|
|
|
|
|
2013-10-04 09:36:51 +08:00
|
|
|
The sliding window classifiers described above are applied to every level of an
|
|
|
|
image pyramid. So you need to tell scan_image_pyramid what kind of pyramid you want
|
|
|
|
to use. In this case we are using pyramid_down<2> which downsamples each pyramid
|
|
|
|
layer by half (if you want to use a finer image pyramid then just change the
|
|
|
|
template argument to a larger value. For example, using pyramid_down<5> would
|
|
|
|
downsample each layer by a ratio of 5 to 4).
|
2012-08-27 03:07:28 +08:00
|
|
|
|
|
|
|
Finally, some of the feature extraction zones are allowed to move freely within the
|
|
|
|
object box. This means that when we are sliding the classifier over an image, some
|
|
|
|
feature extraction zones are stationary (i.e. always in the same place relative to
|
|
|
|
the object box) while others are allowed to move anywhere within the object box. In
|
|
|
|
particular, the movable regions are placed at the locations that maximize the score
|
|
|
|
of the classifier. Note further that each of the movable feature extraction zones
|
|
|
|
must pass a threshold test for it to be included. That is, if the score that a
|
|
|
|
movable zone would contribute to the overall score for a sliding window location is
|
|
|
|
not positive then that zone is not included in the feature vector (i.e. its part of
|
|
|
|
the feature vector is set to zero. This way the length of the feature vector stays
|
|
|
|
constant). This movable region construction allows us to represent objects with
|
|
|
|
parts that move around relative to the object box. For example, a human has hands
|
|
|
|
but they aren't always in the same place relative to a person's bounding box.
|
|
|
|
However, to keep this example program simple, we will only be using stationary
|
|
|
|
feature extraction regions.
|
2011-09-18 04:23:06 +08:00
|
|
|
*/
|
|
|
|
typedef hashed_feature_image<hog_image<3,3,1,4,hog_signed_gradient,hog_full_interpolation> > feature_extractor_type;
|
2013-10-04 09:36:51 +08:00
|
|
|
typedef scan_image_pyramid<pyramid_down<2>, feature_extractor_type> image_scanner_type;
|
2011-09-18 04:23:06 +08:00
|
|
|
image_scanner_type scanner;
|
|
|
|
|
2011-12-27 23:27:59 +08:00
|
|
|
// The hashed_feature_image in the scanner needs to be supplied with a hash function capable
|
|
|
|
// of hashing the outputs of the hog_image. Calling this function will set it up for us. The
|
2013-08-06 12:33:50 +08:00
|
|
|
// 10 here indicates that it will hash HOG vectors into the range [0, pow(2,10)). Therefore,
|
2011-12-27 10:43:07 +08:00
|
|
|
// the feature vectors output by the hashed_feature_image will have dimension pow(2,10).
|
|
|
|
setup_hashed_features(scanner, images, 10);
|
2013-08-06 12:33:50 +08:00
|
|
|
// We should also tell the scanner to use the uniform feature weighting scheme
|
|
|
|
// since it works best on the data in this example. If you don't call this
|
|
|
|
// function then it will use a slightly different weighting scheme which can give
|
|
|
|
// improved results on many normal image types.
|
|
|
|
use_uniform_feature_weights(scanner);
|
2011-09-18 04:23:06 +08:00
|
|
|
|
2011-12-27 10:43:07 +08:00
|
|
|
// We also need to setup the detection templates the scanner will use. It is important that
|
|
|
|
// we add detection templates which are capable of matching all the output boxes we want to learn.
|
|
|
|
// For example, if object_locations contained a rectangle with a height to width ratio of 10 but
|
|
|
|
// we only added square detection templates then it would be impossible to detect this non-square
|
2012-03-21 06:12:37 +08:00
|
|
|
// rectangle. The setup_grid_detection_templates_verbose() routine will take care of this for us by
|
|
|
|
// looking at the contents of object_locations and automatically picking an appropriate set. Also,
|
2012-08-27 03:07:28 +08:00
|
|
|
// the final arguments indicate that we want our detection templates to have 4 feature extraction
|
|
|
|
// regions laid out in a 2x2 regular grid inside each sliding window.
|
2012-03-21 06:12:37 +08:00
|
|
|
setup_grid_detection_templates_verbose(scanner, object_locations, 2, 2);
|
2011-09-18 04:23:06 +08:00
|
|
|
|
|
|
|
|
|
|
|
// Now that we have defined the kind of sliding window classifier system we want and stored
|
|
|
|
// the details into the scanner object we are ready to use the structural_object_detection_trainer
|
|
|
|
// to learn the weight vector and threshold needed to produce a complete object detector.
|
|
|
|
structural_object_detection_trainer<image_scanner_type> trainer(scanner);
|
|
|
|
trainer.set_num_threads(4); // Set this to the number of processing cores on your machine.
|
|
|
|
|
|
|
|
|
|
|
|
// There are a variety of other useful parameters to the structural_object_detection_trainer.
|
|
|
|
// Examples of the ones you are most likely to use follow (see dlib documentation for what they do):
|
2011-12-27 10:43:07 +08:00
|
|
|
//trainer.set_match_eps(0.80);
|
2011-09-18 04:23:06 +08:00
|
|
|
//trainer.set_c(1.0);
|
|
|
|
//trainer.set_loss_per_missed_target(1);
|
|
|
|
//trainer.set_loss_per_false_alarm(1);
|
|
|
|
|
|
|
|
|
|
|
|
// Do the actual training and save the results into the detector object.
|
|
|
|
object_detector<image_scanner_type> detector = trainer.train(images, object_locations);
|
|
|
|
|
|
|
|
// We can easily test the new detector against our training data. This print statement will indicate that it
|
2013-07-25 18:38:23 +08:00
|
|
|
// has perfect precision and recall on this simple task. It will also print the average precision (AP).
|
|
|
|
cout << "Test detector (precision,recall,AP): " << test_object_detection_function(detector, images, object_locations) << endl;
|
2011-09-18 04:23:06 +08:00
|
|
|
|
|
|
|
// The cross validation should also indicate perfect precision and recall.
|
2013-07-25 18:38:23 +08:00
|
|
|
cout << "3-fold cross validation (precision,recall,AP): "
|
2011-09-18 04:23:06 +08:00
|
|
|
<< cross_validate_object_detection_trainer(trainer, images, object_locations, 3) << endl;
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
2014-02-23 05:07:17 +08:00
|
|
|
// Let's display the output of the detector along with our training images.
|
2011-09-18 04:23:06 +08:00
|
|
|
image_window win;
|
|
|
|
for (unsigned long i = 0; i < images.size(); ++i)
|
|
|
|
{
|
|
|
|
// Run the detector on images[i]
|
|
|
|
const std::vector<rectangle> rects = detector(images[i]);
|
|
|
|
cout << "Number of detections: "<< rects.size() << endl;
|
|
|
|
|
|
|
|
// Put the image and detections into the window.
|
|
|
|
win.clear_overlay();
|
|
|
|
win.set_image(images[i]);
|
2012-01-22 05:52:54 +08:00
|
|
|
win.add_overlay(rects, rgb_pixel(255,0,0));
|
2011-09-18 04:23:06 +08:00
|
|
|
|
|
|
|
cout << "Hit enter to see the next image.";
|
|
|
|
cin.get();
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
// Finally, note that the detector can be serialized to disk just like other dlib objects.
|
2014-05-09 09:07:56 +08:00
|
|
|
serialize("object_detector.dat") << detector;
|
2011-09-18 04:23:06 +08:00
|
|
|
|
|
|
|
// Recall from disk.
|
2014-05-09 09:07:56 +08:00
|
|
|
deserialize("object_detector.dat") >> detector;
|
2011-09-18 04:23:06 +08:00
|
|
|
}
|
|
|
|
catch (exception& e)
|
|
|
|
{
|
|
|
|
cout << "\nexception thrown!" << endl;
|
|
|
|
cout << e.what() << endl;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
// ----------------------------------------------------------------------------------------
|
|
|
|
|