def create_detector_from_file(cfg_file):
try:
with open(cfg_file, 'r') as fp:
config = json.load(fp)
cropper_cfg = config['cropper']
cropper = Cropper(cropper_cfg['crop_percent'],
cropper_cfg['force_square'])
localizer_cfg = config['localizer']
localizer = Localizer(localizer_cfg['model'],
localizer_cfg['weights'],
localizer_cfg['threshold'])
classifier_cfg = config['classifier']
classifier = Classifier(classifier_cfg['model'],
classifier_cfg['weights'],
classifier_cfg['labels'],
classifier_cfg['threshold'],
classifier_cfg['skip_classes'])
detection_pipeline = DetectionPipeline(localizer, cropper, classifier)
detector_cfg = config['detector']
detector = Detector(detection_pipeline,
detector_cfg['images_path'],
detector_cfg['sounds_path'])
return detector
except:
raise FormatNotSupportedException('The configuration file does not seems to be valid')
def test_expand_and_crop(self):
crop_percent = 0.1
force_square = True
test_image_path = 'tests/data/test.jpg'
image = cv2.imread(test_image_path)
h, w, _ = image.shape
x1 = 100
y1 = 50
width = 200
height = 300
original_coordinates = (x1, y1, x1 + width, y1 + height)
crp = Cropper(crop_percent, force_square)
new_coords, crop_image = crp.expand_and_crop(image,
original_coordinates)
crop_h, crop_w, _ = crop_image.shape
assert crop_h == 330
assert crop_w == 330
def test_detector_image(self):
# Create Localizer
cfg_path = './data/yolo/full/trafficsigns.cfg'
weights_path = './data/yolo/full/trafficsigns.weights'
threshold = 0.24
localizer = Localizer(cfg_path, weights_path, threshold, gpu=0.0)
# Create cropper
crop_percent = 0.25
force_square = True
cropper = Cropper(crop_percent, force_square)
# Create classifier
model_path = './data/classifier/trafficsigns.json'
weights_path = './data/classifier/trafficsigns.h5'
labels_path = './data/classifier/classes.txt'
threshold = 0.5
classifier = Classifier(model_path, weights_path, labels_path,
threshold)
# Create detection pipeline
detection_pipeline = DetectionPipeline(localizer, cropper, classifier)
# Create detector
images_path = './data/classifier/classes/'
sounds_path = './data/sounds/'
detector = Detector(detection_pipeline, images_path, sounds_path)
# Detect on image
img = cv2.imread('./tests/data/test.png')
image, detections = detector.detect_image(img,
show_confidence=True,
return_image=True)
true_detections = [{
'class_id': 15,
'coordinates': [1434, 456, 1590, 612],
'label': 'max-60',
'confidence': 1.0
}]
assert detections == true_detections
def test_detector_video(self):
# Create Localizer
cfg_path = './data/yolo/full/trafficsigns.cfg'
weights_path = './data/yolo/full/trafficsigns.weights'
threshold = 0.24
localizer = Localizer(cfg_path, weights_path, threshold, gpu=0.0)
# Create cropper
crop_percent = 0.25
force_square = True
cropper = Cropper(crop_percent, force_square)
# Create classifier
model_path = './data/classifier/trafficsigns.json'
weights_path = './data/classifier/trafficsigns.h5'
labels_path = './data/classifier/classes.txt'
threshold = 0.5
classifier = Classifier(model_path, weights_path, labels_path,
threshold)
# Create detection pipeline
detection_pipeline = DetectionPipeline(localizer, cropper, classifier)
# Create detector
images_path = './data/classifier/classes/'
sounds_path = './data/sounds/'
detector = Detector(detection_pipeline, images_path, sounds_path)
# Detect on video
video_feed = './tests/data/test.mp4'
output_mp4 = './tests/data/test_output.mp4'
output_csv = './tests/data/test_output.csv'
exit_code = detector.detect_video_feed(video_feed,
output=output_mp4,
output_csv=output_csv,
show_confidence=True,
sound_notifications=True)
assert exit_code == True
def test_expand_coordinates_top_side(self):
crop_percent = 0.1
force_square = False
image_size = (1920, 1080)
x1 = 500
y1 = 0
width = 200
height = 100
original_coordinates = (x1, y1, x1 + width, y1 + height)
crp = Cropper(crop_percent, force_square)
expanded_coordinates = crp.expand_coordinates(original_coordinates,
image_size)
expected_coordinates = (400, 0, 800, 250)
assert expanded_coordinates == expected_coordinates
def test_expand_coordinates_right_side(self):
crop_percent = 0.1
force_square = False
image_size = (600, 800)
x1 = 500
y1 = 300
width = 100
height = 200
original_coordinates = (x1, y1, x1 + width, y1 + height)
crp = Cropper(crop_percent, force_square)
expanded_coordinates = crp.expand_coordinates(original_coordinates,
image_size)
expected_coordinates = (350, 200, 600, 600)
assert expanded_coordinates == expected_coordinates
def test_expand_coordinates_square_bottom_side(self):
crop_percent = 0.1
force_square = True
image_size = (1920, 1080)
x1 = 500
y1 = 1000
width = 100
height = 80
original_coordinates = (x1, y1, x1 + width, y1 + height)
crp = Cropper(crop_percent, force_square)
expanded_coordinates = crp.expand_coordinates(original_coordinates,
image_size)
expected_coordinates = (495, 985, 605, 1080)
assert expanded_coordinates == expected_coordinates
def test_expand_coordinates_square_left_side(self):
crop_percent = 0.1
force_square = True
image_size = (1920, 1080)
x1 = 0
y1 = 500
width = 100
height = 200
original_coordinates = (x1, y1, x1 + width, y1 + height)
crp = Cropper(crop_percent, force_square)
expanded_coordinates = crp.expand_coordinates(original_coordinates,
image_size)
expected_coordinates = (0, 490, 160, 710)
assert expanded_coordinates == expected_coordinates
def test_detector_image_output_json(self):
# Create Localizer
cfg_path = './data/yolo/full/trafficsigns.cfg'
weights_path = './data/yolo/full/trafficsigns.weights'
threshold = 0.24
localizer = Localizer(cfg_path, weights_path, threshold, gpu=0.0)
# Create cropper
crop_percent = 0.25
force_square = True
cropper = Cropper(crop_percent, force_square)
# Create classifier
model_path = './data/classifier/trafficsigns.json'
weights_path = './data/classifier/trafficsigns.h5'
labels_path = './data/classifier/classes.txt'
threshold = 0.5
classifier = Classifier(model_path, weights_path, labels_path,
threshold)
# Create detection pipeline
detection_pipeline = DetectionPipeline(localizer, cropper, classifier)
# Create detector
images_path = './data/classifier/classes/'
sounds_path = './data/sounds/'
detector = Detector(detection_pipeline, images_path, sounds_path)
# Detect on image
img = cv2.imread('./tests/data/test.png')
output_filename = './tests/data/detector_image.json'
detections = detector.detect_image(img, output=output_filename)
assert os.path.exists(output_filename) == True
def test_detector_pipeline(self):
# Create Localizer
cfg_path = './data/yolo/full/trafficsigns.cfg'
weights_path = './data/yolo/full/trafficsigns.weights'
threshold = 0.24
localizer = Localizer(cfg_path, weights_path, threshold, gpu=0.0)
# Create cropper
crop_percent = 0.25
force_square = True
cropper = Cropper(crop_percent, force_square)
# Create classifier
model_path = './data/classifier/trafficsigns.json'
weights_path = './data/classifier/trafficsigns.h5'
labels_path = './data/classifier/classes.txt'
threshold = 0.5
classifier = Classifier(model_path, weights_path, labels_path, threshold)
# Create detection pipeline
pipeline = DetectionPipeline(localizer, cropper, classifier)
# Detect on image
img = cv2.imread('./tests/data/test.png')
detections = pipeline.detect_objects_in_image(img)
print(detections)
assert 0 == 0