def testRandomBlackPatches(self):
preprocessing_options = []
preprocessing_options.append((preprocessor.normalize_image, {
'original_minval': 0,
'original_maxval': 255,
'target_minval': 0,
'target_maxval': 1
}))
preprocessing_options.append((preprocessor.random_black_patches, {
'size_to_image_ratio': 0.5
}))
images = self.createTestImages()
tensor_dict = {fields.InputDataFields.image: images}
blacked_tensor_dict = preprocessor.preprocess(tensor_dict,
preprocessing_options)
blacked_images = blacked_tensor_dict[fields.InputDataFields.image]
images_shape = tf.shape(images)
blacked_images_shape = tf.shape(blacked_images)
with self.test_session() as sess:
(images_shape_, blacked_images_shape_) = sess.run(
[images_shape, blacked_images_shape])
self.assertAllEqual(images_shape_, blacked_images_shape_)
python类image()的实例源码
def _crop_pool_layer(self, bottom, rois, name):
with tf.variable_scope(name) as scope:
batch_ids = tf.squeeze(tf.slice(rois, [0, 0], [-1, 1], name="batch_id"), [1])
# Get the normalized coordinates of bounding boxes
bottom_shape = tf.shape(bottom)
height = (tf.to_float(bottom_shape[1]) - 1.) * np.float32(self._feat_stride[0])
width = (tf.to_float(bottom_shape[2]) - 1.) * np.float32(self._feat_stride[0])
x1 = tf.slice(rois, [0, 1], [-1, 1], name="x1") / width
y1 = tf.slice(rois, [0, 2], [-1, 1], name="y1") / height
x2 = tf.slice(rois, [0, 3], [-1, 1], name="x2") / width
y2 = tf.slice(rois, [0, 4], [-1, 1], name="y2") / height
# Won't be back-propagated to rois anyway, but to save time
bboxes = tf.stop_gradient(tf.concat([y1, x1, y2, x2], axis=1))
pre_pool_size = cfg.POOLING_SIZE * 2
crops = tf.image.crop_and_resize(bottom, bboxes, tf.to_int32(batch_ids), [pre_pool_size, pre_pool_size], name="crops")
return slim.max_pool2d(crops, [2, 2], padding='SAME')
def _crop_pool_layer(self, bottom, rois, name):
with tf.variable_scope(name) as scope:
batch_ids = tf.squeeze(tf.slice(rois, [0, 0], [-1, 1], name="batch_id"), [1])
# Get the normalized coordinates of bounding boxes
bottom_shape = tf.shape(bottom)
height = (tf.to_float(bottom_shape[1]) - 1.) * np.float32(self._feat_stride[0])
width = (tf.to_float(bottom_shape[2]) - 1.) * np.float32(self._feat_stride[0])
x1 = tf.slice(rois, [0, 1], [-1, 1], name="x1") / width
y1 = tf.slice(rois, [0, 2], [-1, 1], name="y1") / height
x2 = tf.slice(rois, [0, 3], [-1, 1], name="x2") / width
y2 = tf.slice(rois, [0, 4], [-1, 1], name="y2") / height
# Won't be back-propagated to rois anyway, but to save time
bboxes = tf.stop_gradient(tf.concat([y1, x1, y2, x2], axis=1))
pre_pool_size = cfg.POOLING_SIZE * 2
crops = tf.image.crop_and_resize(bottom, bboxes, tf.to_int32(batch_ids), [pre_pool_size, pre_pool_size], name="crops")
return slim.max_pool2d(crops, [2, 2], padding='SAME')
def subtract_mean_multi(image_tensors, mean_image_path, channels=NUM_CHANNELS, image_size=512):
mean_image = tf.convert_to_tensor(mean_image_path, dtype=tf.string)
mean_file_contents = tf.read_file(mean_image)
mean_uint8 = tf.image.decode_png(mean_file_contents, channels=channels)
mean_uint8.set_shape([image_size, image_size, channels])
images_mean_free = []
for image_tensor in image_tensors:
image_tensor.set_shape([image_size, image_size, channels])
image = tf.cast(image_tensor, tf.float32)
#subtract mean image
image_mean_free = tf.subtract(image, tf.cast(mean_uint8, tf.float32))
images_mean_free.append(image_mean_free)
return images_mean_free
def single_input_image(image_str, mean_image_path, png_with_alpha=False, image_size=512):
mean_image_str = tf.convert_to_tensor(mean_image_path, dtype=tf.string)
file_contents = tf.read_file(image_str)
if png_with_alpha:
uint8image = tf.image.decode_png(file_contents, channels=4)
uint8image.set_shape([image_size, image_size, 4])
else:
uint8image = tf.image.decode_image(file_contents, channels=NUM_CHANNELS)
uint8image.set_shape([image_size, image_size, NUM_CHANNELS])
image = tf.cast(uint8image, tf.float32)
#subtract mean image
mean_file_contents = tf.read_file(mean_image_str)
if png_with_alpha:
mean_uint8 = tf.image.decode_png(mean_file_contents, channels=4)
mean_uint8.set_shape([image_size, image_size, 4])
else:
mean_uint8 = tf.image.decode_image(mean_file_contents, channels=NUM_CHANNELS)
mean_uint8.set_shape([image_size, image_size, NUM_CHANNELS])
image_mean_free = tf.subtract(image, tf.cast(mean_uint8, tf.float32))
return image_mean_free
def _crop_pool_layer(self, bottom, rois, name):
with tf.variable_scope(name) as scope:
batch_ids = tf.squeeze(tf.slice(rois, [0, 0], [-1, 1], name="batch_id"), [1])
# Get the normalized coordinates of bounding boxes
bottom_shape = tf.shape(bottom)
height = (tf.to_float(bottom_shape[1]) - 1.) * np.float32(self._feat_stride[0])
width = (tf.to_float(bottom_shape[2]) - 1.) * np.float32(self._feat_stride[0])
x1 = tf.slice(rois, [0, 1], [-1, 1], name="x1") / width
y1 = tf.slice(rois, [0, 2], [-1, 1], name="y1") / height
x2 = tf.slice(rois, [0, 3], [-1, 1], name="x2") / width
y2 = tf.slice(rois, [0, 4], [-1, 1], name="y2") / height
# Won't be back-propagated to rois anyway, but to save time
bboxes = tf.stop_gradient(tf.concat([y1, x1, y2, x2], axis=1))
pre_pool_size = cfg.POOLING_SIZE * 2
crops = tf.image.crop_and_resize(bottom, bboxes, tf.to_int32(batch_ids), [pre_pool_size, pre_pool_size], name="crops")
return slim.max_pool2d(crops, [2, 2], padding='SAME')
def random_crop(images, height, width):
"""Randomly crops an image/images to a given size.
Args:
images: 4-D Tensor of shape `[batch, height, width, channels]` or
3-D Tensor of shape `[height, width, channels]`.
height: `float`. The height to crop to.
width: `float`. The width to crop to.
Returns:
If `images` was 4-D, a 4-D float Tensor of shape
`[batch, new_height, new_width, channels]`.
If `images` was 3-D, a 3-D float Tensor of shape
`[new_height, new_width, channels]`.
"""
images_shape = get_shape(images)
if len(images_shape) > 4:
ValueError("'image' must have either 3 or 4 dimensions, "
"received `{}`.".format(images_shape))
if len(images_shape) == 4:
return tf.map_fn(lambda img: tf.random_crop(img, [height, width, images_shape[-1]]), images)
return tf.random_crop(images, [height, width, images_shape[-1]])
def adjust_gamma(image, gamma=1, gain=1):
"""Performs Gamma Correction on the input image.
Also known as Power Law Transform. This function transforms the
input image pixelwise according to the equation Out = In**gamma
after scaling each pixel to the range 0 to 1.
(A mirror to tf.image adjust_gamma)
Args:
image : A Tensor.
gamma : A scalar. Non negative real number.
gain : A scalar. The constant multiplier.
Returns:
A Tensor. Gamma corrected output image.
Notes:
For gamma greater than 1, the histogram will shift towards left and
the output image will be darker than the input image.
For gamma less than 1, the histogram will shift towards right and
the output image will be brighter than the input image.
References:
[1] http://en.wikipedia.org/wiki/Gamma_correction
"""
return tf.image.adjust_gamma(image, gamma, gain)
def testNormalizeImage(self):
preprocess_options = [(preprocessor.normalize_image, {
'original_minval': 0,
'original_maxval': 256,
'target_minval': -1,
'target_maxval': 1
})]
images = self.createTestImages()
tensor_dict = {fields.InputDataFields.image: images}
tensor_dict = preprocessor.preprocess(tensor_dict, preprocess_options)
images = tensor_dict[fields.InputDataFields.image]
images_expected = self.expectedImagesAfterNormalization()
with self.test_session() as sess:
(images_, images_expected_) = sess.run(
[images, images_expected])
images_shape_ = images_.shape
images_expected_shape_ = images_expected_.shape
expected_shape = [1, 4, 4, 3]
self.assertAllEqual(images_expected_shape_, images_shape_)
self.assertAllEqual(images_shape_, expected_shape)
self.assertAllClose(images_, images_expected_)
def testRandomPixelValueScale(self):
preprocessing_options = []
preprocessing_options.append((preprocessor.normalize_image, {
'original_minval': 0,
'original_maxval': 255,
'target_minval': 0,
'target_maxval': 1
}))
preprocessing_options.append((preprocessor.random_pixel_value_scale, {}))
images = self.createTestImages()
tensor_dict = {fields.InputDataFields.image: images}
tensor_dict = preprocessor.preprocess(tensor_dict, preprocessing_options)
images_min = tf.to_float(images) * 0.9 / 255.0
images_max = tf.to_float(images) * 1.1 / 255.0
images = tensor_dict[fields.InputDataFields.image]
values_greater = tf.greater_equal(images, images_min)
values_less = tf.less_equal(images, images_max)
values_true = tf.fill([1, 4, 4, 3], True)
with self.test_session() as sess:
(values_greater_, values_less_, values_true_) = sess.run(
[values_greater, values_less, values_true])
self.assertAllClose(values_greater_, values_true_)
self.assertAllClose(values_less_, values_true_)
def testRandomImageScale(self):
preprocess_options = [(preprocessor.random_image_scale, {})]
images_original = self.createTestImages()
tensor_dict = {fields.InputDataFields.image: images_original}
tensor_dict = preprocessor.preprocess(tensor_dict, preprocess_options)
images_scaled = tensor_dict[fields.InputDataFields.image]
images_original_shape = tf.shape(images_original)
images_scaled_shape = tf.shape(images_scaled)
with self.test_session() as sess:
(images_original_shape_, images_scaled_shape_) = sess.run(
[images_original_shape, images_scaled_shape])
self.assertTrue(
images_original_shape_[1] * 0.5 <= images_scaled_shape_[1])
self.assertTrue(
images_original_shape_[1] * 2.0 >= images_scaled_shape_[1])
self.assertTrue(
images_original_shape_[2] * 0.5 <= images_scaled_shape_[2])
self.assertTrue(
images_original_shape_[2] * 2.0 >= images_scaled_shape_[2])
def testRandomAdjustBrightness(self):
preprocessing_options = []
preprocessing_options.append((preprocessor.normalize_image, {
'original_minval': 0,
'original_maxval': 255,
'target_minval': 0,
'target_maxval': 1
}))
preprocessing_options.append((preprocessor.random_adjust_brightness, {}))
images_original = self.createTestImages()
tensor_dict = {fields.InputDataFields.image: images_original}
tensor_dict = preprocessor.preprocess(tensor_dict, preprocessing_options)
images_bright = tensor_dict[fields.InputDataFields.image]
image_original_shape = tf.shape(images_original)
image_bright_shape = tf.shape(images_bright)
with self.test_session() as sess:
(image_original_shape_, image_bright_shape_) = sess.run(
[image_original_shape, image_bright_shape])
self.assertAllEqual(image_original_shape_, image_bright_shape_)
def testRandomAdjustContrast(self):
preprocessing_options = []
preprocessing_options.append((preprocessor.normalize_image, {
'original_minval': 0,
'original_maxval': 255,
'target_minval': 0,
'target_maxval': 1
}))
preprocessing_options.append((preprocessor.random_adjust_contrast, {}))
images_original = self.createTestImages()
tensor_dict = {fields.InputDataFields.image: images_original}
tensor_dict = preprocessor.preprocess(tensor_dict, preprocessing_options)
images_contrast = tensor_dict[fields.InputDataFields.image]
image_original_shape = tf.shape(images_original)
image_contrast_shape = tf.shape(images_contrast)
with self.test_session() as sess:
(image_original_shape_, image_contrast_shape_) = sess.run(
[image_original_shape, image_contrast_shape])
self.assertAllEqual(image_original_shape_, image_contrast_shape_)
def testRandomAdjustHue(self):
preprocessing_options = []
preprocessing_options.append((preprocessor.normalize_image, {
'original_minval': 0,
'original_maxval': 255,
'target_minval': 0,
'target_maxval': 1
}))
preprocessing_options.append((preprocessor.random_adjust_hue, {}))
images_original = self.createTestImages()
tensor_dict = {fields.InputDataFields.image: images_original}
tensor_dict = preprocessor.preprocess(tensor_dict, preprocessing_options)
images_hue = tensor_dict[fields.InputDataFields.image]
image_original_shape = tf.shape(images_original)
image_hue_shape = tf.shape(images_hue)
with self.test_session() as sess:
(image_original_shape_, image_hue_shape_) = sess.run(
[image_original_shape, image_hue_shape])
self.assertAllEqual(image_original_shape_, image_hue_shape_)
def testRandomDistortColor(self):
preprocessing_options = []
preprocessing_options.append((preprocessor.normalize_image, {
'original_minval': 0,
'original_maxval': 255,
'target_minval': 0,
'target_maxval': 1
}))
preprocessing_options.append((preprocessor.random_distort_color, {}))
images_original = self.createTestImages()
images_original_shape = tf.shape(images_original)
tensor_dict = {fields.InputDataFields.image: images_original}
tensor_dict = preprocessor.preprocess(tensor_dict, preprocessing_options)
images_distorted_color = tensor_dict[fields.InputDataFields.image]
images_distorted_color_shape = tf.shape(images_distorted_color)
with self.test_session() as sess:
(images_original_shape_, images_distorted_color_shape_) = sess.run(
[images_original_shape, images_distorted_color_shape])
self.assertAllEqual(images_original_shape_, images_distorted_color_shape_)
def testRandomResizeMethod(self):
preprocessing_options = []
preprocessing_options.append((preprocessor.normalize_image, {
'original_minval': 0,
'original_maxval': 255,
'target_minval': 0,
'target_maxval': 1
}))
preprocessing_options.append((preprocessor.random_resize_method, {
'target_size': (75, 150)
}))
images = self.createTestImages()
tensor_dict = {fields.InputDataFields.image: images}
resized_tensor_dict = preprocessor.preprocess(tensor_dict,
preprocessing_options)
resized_images = resized_tensor_dict[fields.InputDataFields.image]
resized_images_shape = tf.shape(resized_images)
expected_images_shape = tf.constant([1, 75, 150, 3], dtype=tf.int32)
with self.test_session() as sess:
(expected_images_shape_, resized_images_shape_) = sess.run(
[expected_images_shape, resized_images_shape])
self.assertAllEqual(expected_images_shape_,
resized_images_shape_)
def testResizeToRangeWithDynamicSpatialShape(self):
"""Tests image resizing, checking output sizes."""
in_shape_list = [[60, 40, 3], [15, 30, 3], [15, 50, 3]]
min_dim = 50
max_dim = 100
expected_shape_list = [[75, 50, 3], [50, 100, 3], [30, 100, 3]]
for in_shape, expected_shape in zip(in_shape_list, expected_shape_list):
in_image = tf.placeholder(tf.float32, shape=(None, None, 3))
out_image = preprocessor.resize_to_range(
in_image, min_dimension=min_dim, max_dimension=max_dim)
out_image_shape = tf.shape(out_image)
with self.test_session() as sess:
out_image_shape = sess.run(out_image_shape,
feed_dict={in_image:
np.random.randn(*in_shape)})
self.assertAllEqual(out_image_shape, expected_shape)
def testResizeToRangeSameMinMax(self):
"""Tests image resizing, checking output sizes."""
in_shape_list = [[312, 312, 3], [299, 299, 3]]
min_dim = 320
max_dim = 320
expected_shape_list = [[320, 320, 3], [320, 320, 3]]
for in_shape, expected_shape in zip(in_shape_list, expected_shape_list):
in_image = tf.random_uniform(in_shape)
out_image = preprocessor.resize_to_range(
in_image, min_dimension=min_dim, max_dimension=max_dim)
out_image_shape = tf.shape(out_image)
with self.test_session() as sess:
out_image_shape = sess.run(out_image_shape)
self.assertAllEqual(out_image_shape, expected_shape)
def _add_gt_image(self):
# add back mean
image = self._image + cfg.PIXEL_MEANS
# BGR to RGB (opencv uses BGR)
resized = tf.image.resize_bilinear(image, tf.to_int32(self._im_info[:2] / self._im_info[2]))
self._gt_image = tf.reverse(resized, axis=[-1])
def _add_gt_image_summary(self):
# use a customized visualization function to visualize the boxes
if self._gt_image is None:
self._add_gt_image()
image = tf.py_func(draw_bounding_boxes,
[self._gt_image, self._gt_boxes, self._im_info],
tf.float32, name="gt_boxes")
return tf.summary.image('GROUND_TRUTH', image)
def _roi_pool_layer(self, bootom, rois, name):
with tf.variable_scope(name) as scope:
return tf.image.roi_pooling(bootom, rois,
pooled_height=cfg.POOLING_SIZE,
pooled_width=cfg.POOLING_SIZE,
spatial_scale=1. / 16.)[0]
def _crop_pool_layer(self, bottom, rois, name):
with tf.variable_scope(name) as scope:
batch_ids = tf.squeeze(tf.slice(rois, [0, 0], [-1, 1], name="batch_id"), [1])
# Get the normalized coordinates of bounding boxes
bottom_shape = tf.shape(bottom)
height = (tf.to_float(bottom_shape[1]) - 1.) * np.float32(self._feat_stride[0])
width = (tf.to_float(bottom_shape[2]) - 1.) * np.float32(self._feat_stride[0])
x1 = tf.slice(rois, [0, 1], [-1, 1], name="x1") / width
y1 = tf.slice(rois, [0, 2], [-1, 1], name="y1") / height
x2 = tf.slice(rois, [0, 3], [-1, 1], name="x2") / width
y2 = tf.slice(rois, [0, 4], [-1, 1], name="y2") / height
# Won't be back-propagated to rois anyway, but to save time
bboxes = tf.stop_gradient(tf.concat([y1, x1, y2, x2], axis=1))
pre_pool_size = cfg.POOLING_SIZE * 2
crops = tf.image.crop_and_resize(bottom, bboxes, tf.to_int32(batch_ids), [pre_pool_size, pre_pool_size], name="crops")
return slim.max_pool2d(crops, [2, 2], padding='SAME')
def _build_network(self, is_training=True):
# select initializers
if cfg.TRAIN.TRUNCATED:
initializer = tf.truncated_normal_initializer(mean=0.0, stddev=0.01)
initializer_bbox = tf.truncated_normal_initializer(mean=0.0, stddev=0.001)
else:
initializer = tf.random_normal_initializer(mean=0.0, stddev=0.01)
initializer_bbox = tf.random_normal_initializer(mean=0.0, stddev=0.001)
net_conv = self._image_to_head(is_training)
with tf.variable_scope(self._scope, self._scope):
# build the anchors for the image
self._anchor_component()
# region proposal network
rois = self._region_proposal(net_conv, is_training, initializer)
# region of interest pooling
if cfg.POOLING_MODE == 'crop':
pool5 = self._crop_pool_layer(net_conv, rois, "pool5")
else:
raise NotImplementedError
fc7 = self._head_to_tail(pool5, is_training)
with tf.variable_scope(self._scope, self._scope):
# region classification
cls_prob, bbox_pred = self._region_classification(fc7, is_training,
initializer, initializer_bbox)
self._score_summaries.update(self._predictions)
return rois, cls_prob, bbox_pred
def test_image(self, sess, image, im_info):
feed_dict = {self._image: image,
self._im_info: im_info}
cls_score, cls_prob, bbox_pred, rois = sess.run([self._predictions["cls_score"],
self._predictions['cls_prob'],
self._predictions['bbox_pred'],
self._predictions['rois']],
feed_dict=feed_dict)
return cls_score, cls_prob, bbox_pred, rois
def _add_gt_image(self):
# add back mean
image = self._image + cfg.PIXEL_MEANS
# BGR to RGB (opencv uses BGR)
resized = tf.image.resize_bilinear(image, tf.to_int32(self._im_info[:2] / self._im_info[2]))
self._gt_image = tf.reverse(resized, axis=[-1])
def _add_gt_image_summary(self):
# use a customized visualization function to visualize the boxes
if self._gt_image is None:
self._add_gt_image()
image = tf.py_func(draw_bounding_boxes,
[self._gt_image, self._gt_boxes, self._im_info],
tf.float32, name="gt_boxes")
return tf.summary.image('GROUND_TRUTH', image)
def _roi_pool_layer(self, bootom, rois, name):
with tf.variable_scope(name) as scope:
return tf.image.roi_pooling(bootom, rois,
pooled_height=cfg.POOLING_SIZE,
pooled_width=cfg.POOLING_SIZE,
spatial_scale=1. / 16.)[0]
def test_image(self, sess, image, im_info):
feed_dict = {self._image: image,
self._im_info: im_info}
cls_score, cls_prob, bbox_pred, rois = sess.run([self._predictions["cls_score"],
self._predictions['cls_prob'],
self._predictions['bbox_pred'],
self._predictions['rois']],
feed_dict=feed_dict)
return cls_score, cls_prob, bbox_pred, rois
def _add_gt_image(self):
# add back mean
image = self._image + cfg.PIXEL_MEANS
# BGR to RGB (opencv uses BGR)
resized = tf.image.resize_bilinear(image, tf.to_int32(self._im_info[:2] / self._im_info[2]))
self._gt_image = tf.reverse(resized, axis=[-1])
def _add_gt_image_summary(self):
# use a customized visualization function to visualize the boxes
if self._gt_image is None:
self._add_gt_image()
image = tf.py_func(draw_bounding_boxes,
[self._gt_image, self._gt_boxes, self._im_info],
tf.float32, name="gt_boxes")
return tf.summary.image('GROUND_TRUTH', image)
