def test_normalize_bboxes():
bboxes = [[15, 25, 100, 200], [15, 25, 100, 200, 99]]
normalized_bboxes_1 = normalize_bboxes(bboxes, 200, 400)
normalized_bboxes_2 = [
normalize_bbox(bboxes[0], 200, 400),
normalize_bbox(bboxes[1], 200, 400)
]
assert normalized_bboxes_1 == normalized_bboxes_2
def crop_bbox_by_coords(bbox, crop_coords, crop_height, crop_width, rows, cols):
"""Crop a bounding box using the provided coordinates of bottom-left and top-right corners in pixels and the
required height and width of the crop.
"""
bbox = denormalize_bbox(bbox, rows, cols)
x_min, y_min, x_max, y_max = bbox
x1, y1, x2, y2 = crop_coords
cropped_bbox = [x_min - x1, y_min - y1, x_max - x1, y_max - y1]
return normalize_bbox(cropped_bbox, crop_height, crop_width)
def _convert_to_yolo(self, bbox, rows, cols): if self.source_format == 'pascal_voc': (x_min, y_min, x_max, y_max), tail = bbox[:4], bbox[4:] return normalize_bbox(( (x_max + x_min) / 2, (y_max + y_min) / 2, (x_max - x_min), (y_max - y_min) ) + tail, rows, cols) elif self.source_format == 'coco': (x, y, width, height), tail = bbox[:4], bbox[4:] return normalize_bbox(( x + width / 2, y + height / 2, width, height ) + tail, rows, cols) return bbox
def apply_to_bbox(self,
bbox,
pad_top=0,
pad_bottom=0,
pad_left=0,
pad_right=0,
rows=0,
cols=0,
**params):
x_min, y_min, x_max, y_max = denormalize_bbox(bbox, rows, cols)
bbox = x_min + pad_left, y_min + pad_top, x_max + pad_left, y_max + pad_top
return normalize_bbox(bbox, rows + pad_top + pad_bottom,
cols + pad_left + pad_right)
def apply_to_bbox(self, bbox, **params):
"""
Removes the bounding boxes which are covered by the applied cutout
:param bbox: A single bounding box coordinates in pascal_voc format
:returns transformed bbox's coordinates
"""
# Denormalize the bbox coordinates
bbox = denormalize_bbox(bbox, self.img_height, self.img_width)
x_min, y_min, x_max, y_max = tuple(map(int, bbox))
bbox_size = (x_max - x_min) * (y_max - y_min) # width * height
overlapping_size = np.sum(
(self.image[y_min:y_max, x_min:x_max, 0] == self.fill_value)
& (self.image[y_min:y_max, x_min:x_max, 1] == self.fill_value)
& (self.image[y_min:y_max, x_min:x_max, 2] == self.fill_value))
# Remove the bbox if it has more than some threshold of content is inside the cutout patch
if overlapping_size / bbox_size > self.bbox_removal_threshold:
return normalize_bbox((0, 0, 0, 0), self.img_height,
self.img_width)
return normalize_bbox(bbox, self.img_height, self.img_width)
def apply_to_bboxes(self, bboxes, **params):
norm_cap_bboxes = [
normalize_bbox(bbox, rows=self.img_h, cols=self.img_w)
for bbox in self.cap_bboxes
]
bboxes = np.array(bboxes)
norm_cap_bboxes = np.array(norm_cap_bboxes)
if len(bboxes) > 0 and len(norm_cap_bboxes) > 0:
bbx_result = np.concatenate((bboxes, norm_cap_bboxes), axis=0)
elif len(norm_cap_bboxes) > 0:
bbx_result = norm_cap_bboxes
else:
bbx_result = bboxes
return bbx_result
def test_denormalize_normalize_bbox(bbox):
denormalized_bbox = denormalize_bbox(bbox, 200, 400)
normalized_bbox = normalize_bbox(denormalized_bbox, 200, 400)
assert normalized_bbox == bbox
def test_normalize_denormalize_bbox(bbox):
normalized_bbox = normalize_bbox(bbox, 200, 400)
denormalized_bbox = denormalize_bbox(normalized_bbox, 200, 400)
assert denormalized_bbox == bbox
def test_normalize_bbox(bbox, expected):
normalized_bbox = normalize_bbox(bbox, 200, 400)
assert normalized_bbox == expected
