def image_preprocess(image, image_size: int):
input_processor = dataloader.DetectionInputProcessor(image, image_size)
input_processor.normalize_image()
input_processor.set_scale_factors_to_output_size()
image = input_processor.resize_and_crop_image()
image_scale = input_processor.image_scale_to_original
return image, image_scale
def map_fn(image):
input_processor = dataloader.DetectionInputProcessor(
image, image_size)
input_processor.normalize_image(mean_rgb, stddev_rgb)
input_processor.set_scale_factors_to_output_size()
image = input_processor.resize_and_crop_image()
image_scale = input_processor.image_scale_to_original
return image, image_scale
def image_preprocess(image, image_size: Union[int, Tuple[int, int]]):
"""Preprocess image for inference.
Args:
image: input image, can be a tensor or a numpy arary.
image_size: single integer of image size for square image or tuple of two
integers, in the format of (image_height, image_width).
Returns:
(image, scale): a tuple of processed image and its scale.
"""
input_processor = dataloader.DetectionInputProcessor(image, image_size)
input_processor.normalize_image()
input_processor.set_scale_factors_to_output_size()
image = input_processor.resize_and_crop_image()
image_scale = input_processor.image_scale_to_original
return image, image_scale
def image_preprocess(image, image_size: int):
"""Preprocess image for inference.
Args:
image: input image, can be a tensor or a numpy arary.
image_size: integer of image size.
Returns:
(image, scale): a tuple of processed image and its scale.
"""
input_processor = dataloader.DetectionInputProcessor(image, image_size)
input_processor.normalize_image()
input_processor.set_scale_factors_to_output_size()
image = input_processor.resize_and_crop_image()
image_scale = input_processor.image_scale_to_original
return image, image_scale
def image_preprocess(image, image_size, mean_rgb, stddev_rgb):
"""Preprocess image for inference.
Args:
image: input image, can be a tensor or a numpy arary.
image_size: single integer of image size for square image or tuple of two
integers, in the format of (image_height, image_width).
mean_rgb: Mean value of RGB, can be a list of float or a float value.
stddev_rgb: Standard deviation of RGB, can be a list of float or a float
value.
Returns:
(image, scale): a tuple of processed image and its scale.
"""
input_processor = dataloader.DetectionInputProcessor(image, image_size)
input_processor.normalize_image(mean_rgb, stddev_rgb)
input_processor.set_scale_factors_to_output_size()
image = input_processor.resize_and_crop_image()
image_scale = input_processor.image_scale_to_original
return image, image_scale
def _preprocessing(self, raw_images, image_size, mode=None):
"""Preprocess images before feeding to the network."""
if not mode:
return raw_images, None
image_size = utils.parse_image_size(image_size)
if mode != 'infer':
# We only support inference for now.
raise ValueError('preprocessing must be infer or empty')
scales, images = [], []
if isinstance(raw_images, tf.Tensor):
batch_size = raw_images.shape[0]
else:
batch_size = len(raw_images)
for i in range(batch_size):
input_processor = dataloader.DetectionInputProcessor(
raw_images[i], image_size)
input_processor.normalize_image()
input_processor.set_scale_factors_to_output_size()
images.append(input_processor.resize_and_crop_image())
scales.append(input_processor.image_scale_to_original)
return tf.stack(images), tf.stack(scales)
