def _read_voc_annot(annot_path: str) -> Tuple[Sequence[int],
Sequence[tf.Tensor]]:
# Reads a voc annotation and returns
# a list of tuples containing the ground
# truth boxes and its respective label
root = ET.parse(annot_path).getroot()
image_size = (int(root.findtext('size/height')), # type: ignore[arg-type]
int(root.findtext('size/width'))) # type: ignore[arg-type]
boxes = root.findall('object')
bbs = []
labels = []
for b in boxes:
bb = b.find('bndbox')
bb = (int(bb.findtext('xmin')), # type: ignore
int(bb.findtext('ymin')), # type: ignore
int(bb.findtext('xmax')), # type: ignore
int(bb.findtext('ymax'))) # type: ignore
bbs.append(bb)
labels.append(LABEL_2_IDX[b.findtext('name')]) # type: ignore
bbs = tf.stack(bbs)
bbs = bb_utils.normalize_bndboxes(bbs, image_size)
return labels, bbs
def _load_labelme_instance(
images_base_path: Union[str, Path],
annot_path: Union[str, Path],
im_input_size: Tuple[int, int],
class2idx: Mapping[str, int]) -> ObjectDetectionInstance:
# Reads a labelme annotation and returns
# a list of tuples containing the ground
# truth boxes and its respective label
with Path(annot_path).open() as f:
annot = json.load(f)
bbs = []
labels = []
image_path = Path(images_base_path) / annot['imagePath']
image = io_utils.load_image(str(image_path),
im_input_size,
normalize_image=True)
w, h = annot['imageWidth'], annot['imageHeight']
for shape in annot['shapes']:
shape_type = shape['shape_type']
if shape_type == 'rectangle':
points = _load_bbox_from_rectangle(shape['points'])
elif shape_type == 'polygon':
points = _load_bbox_from_polygon(shape['points'])
else:
raise ValueError(
f'Unexpected shape type: {shape_type} in file {annot_path}')
label = shape['label']
bbs.append(points)
labels.append(class2idx[label])
boxes = tf.stack(bbs)
boxes = bb_utils.normalize_bndboxes(boxes, (h, w))
return image, (tf.stack(labels), boxes)