Python object_detection.core.preprocessor._strict_random_crop_image() Examples
The following are 30
code examples of object_detection.core.preprocessor._strict_random_crop_image().
You can vote up the ones you like or vote down the ones you don't like,
and go to the original project or source file by following the links above each example.
You may also want to check out all available functions/classes of the module
object_detection.core.preprocessor
, or try the search function
.
.
Example #1
| Source File: preprocessor_test.py From object_detection_with_tensorflow with MIT License | 5 votes |
|
def testStrictRandomCropImageWithMasks(self):
image = self.createColorfulTestImage()[0]
boxes = self.createTestBoxes()
labels = self.createTestLabels()
masks = tf.random_uniform([2, 200, 400], dtype=tf.float32)
with mock.patch.object(
tf.image,
'sample_distorted_bounding_box'
) as mock_sample_distorted_bounding_box:
mock_sample_distorted_bounding_box.return_value = (
tf.constant([6, 143, 0], dtype=tf.int32),
tf.constant([190, 237, -1], dtype=tf.int32),
tf.constant([[[0.03, 0.3575, 0.98, 0.95]]], dtype=tf.float32))
new_image, new_boxes, new_labels, new_masks = (
preprocessor._strict_random_crop_image(
image, boxes, labels, masks=masks))
with self.test_session() as sess:
new_image, new_boxes, new_labels, new_masks = sess.run(
[new_image, new_boxes, new_labels, new_masks])
expected_boxes = np.array(
[[0.0, 0.0, 0.75789469, 1.0],
[0.23157893, 0.24050637, 0.75789469, 1.0]], dtype=np.float32)
self.assertAllEqual(new_image.shape, [190, 237, 3])
self.assertAllEqual(new_masks.shape, [2, 190, 237])
self.assertAllClose(
new_boxes.flatten(), expected_boxes.flatten())
Example #2
| Source File: preprocessor_test.py From Accident-Detection-on-Indian-Roads with GNU Affero General Public License v3.0 | 5 votes |
|
def testStrictRandomCropImageWithMasks(self):
image = self.createColorfulTestImage()[0]
boxes = self.createTestBoxes()
labels = self.createTestLabels()
masks = tf.random_uniform([2, 200, 400], dtype=tf.float32)
with mock.patch.object(
tf.image,
'sample_distorted_bounding_box'
) as mock_sample_distorted_bounding_box:
mock_sample_distorted_bounding_box.return_value = (
tf.constant([6, 143, 0], dtype=tf.int32),
tf.constant([190, 237, -1], dtype=tf.int32),
tf.constant([[[0.03, 0.3575, 0.98, 0.95]]], dtype=tf.float32))
new_image, new_boxes, new_labels, new_masks = (
preprocessor._strict_random_crop_image(
image, boxes, labels, masks=masks))
with self.test_session() as sess:
new_image, new_boxes, new_labels, new_masks = sess.run(
[new_image, new_boxes, new_labels, new_masks])
expected_boxes = np.array(
[[0.0, 0.0, 0.75789469, 1.0],
[0.23157893, 0.24050637, 0.75789469, 1.0]], dtype=np.float32)
self.assertAllEqual(new_image.shape, [190, 237, 3])
self.assertAllEqual(new_masks.shape, [2, 190, 237])
self.assertAllClose(
new_boxes.flatten(), expected_boxes.flatten())
Example #3
| Source File: preprocessor_test.py From MAX-Object-Detector with Apache License 2.0 | 5 votes |
|
def testStrictRandomCropImageWithKeypoints(self):
image = self.createColorfulTestImage()[0]
boxes = self.createTestBoxes()
labels = self.createTestLabels()
weights = self.createTestGroundtruthWeights()
keypoints = self.createTestKeypoints()
with mock.patch.object(
tf.image,
'sample_distorted_bounding_box'
) as mock_sample_distorted_bounding_box:
mock_sample_distorted_bounding_box.return_value = (
tf.constant([6, 143, 0], dtype=tf.int32),
tf.constant([190, 237, -1], dtype=tf.int32),
tf.constant([[[0.03, 0.3575, 0.98, 0.95]]], dtype=tf.float32))
new_image, new_boxes, new_labels, new_weights, new_keypoints = (
preprocessor._strict_random_crop_image(
image, boxes, labels, weights, keypoints=keypoints))
with self.test_session() as sess:
new_image, new_boxes, new_labels, new_weights, new_keypoints = sess.run(
[new_image, new_boxes, new_labels, new_weights, new_keypoints])
expected_boxes = np.array([
[0.0, 0.0, 0.75789469, 1.0],
[0.23157893, 0.24050637, 0.75789469, 1.0],], dtype=np.float32)
expected_keypoints = np.array([
[[np.nan, np.nan],
[np.nan, np.nan],
[np.nan, np.nan]],
[[0.38947368, 0.07173],
[0.49473682, 0.24050637],
[0.60000002, 0.40928277]]
], dtype=np.float32)
self.assertAllEqual(new_image.shape, [190, 237, 3])
self.assertAllClose(
new_boxes.flatten(), expected_boxes.flatten())
self.assertAllClose(
new_keypoints.flatten(), expected_keypoints.flatten())
Example #4
| Source File: preprocessor_test.py From models with Apache License 2.0 | 5 votes |
|
def testStrictRandomCropImageWithMasks(self):
def graph_fn():
image = self.createColorfulTestImage()[0]
boxes = self.createTestBoxes()
labels = self.createTestLabels()
weights = self.createTestGroundtruthWeights()
masks = tf.random_uniform([2, 200, 400], dtype=tf.float32)
with mock.patch.object(
tf.image,
'sample_distorted_bounding_box'
) as mock_sample_distorted_bounding_box:
mock_sample_distorted_bounding_box.return_value = (
tf.constant([6, 143, 0], dtype=tf.int32),
tf.constant([190, 237, -1], dtype=tf.int32),
tf.constant([[[0.03, 0.3575, 0.98, 0.95]]], dtype=tf.float32))
new_image, new_boxes, new_labels, new_weights, new_masks = (
preprocessor._strict_random_crop_image(
image, boxes, labels, weights, masks=masks))
return [new_image, new_boxes, new_labels, new_weights, new_masks]
(new_image, new_boxes, _, _,
new_masks) = self.execute_cpu(graph_fn, [])
expected_boxes = np.array(
[[0.0, 0.0, 0.75789469, 1.0],
[0.23157893, 0.24050637, 0.75789469, 1.0]], dtype=np.float32)
self.assertAllEqual(new_image.shape, [190, 237, 3])
self.assertAllEqual(new_masks.shape, [2, 190, 237])
self.assertAllClose(
new_boxes.flatten(), expected_boxes.flatten())
Example #5
| Source File: preprocessor_test.py From motion-rcnn with MIT License | 5 votes |
|
def testStrictRandomCropImageWithMasks(self):
image = self.createColorfulTestImage()[0]
boxes = self.createTestBoxes()
labels = self.createTestLabels()
masks = tf.random_uniform([2, 200, 400], dtype=tf.float32)
with mock.patch.object(
tf.image,
'sample_distorted_bounding_box'
) as mock_sample_distorted_bounding_box:
mock_sample_distorted_bounding_box.return_value = (
tf.constant([6, 143, 0], dtype=tf.int32),
tf.constant([190, 237, -1], dtype=tf.int32),
tf.constant([[[0.03, 0.3575, 0.98, 0.95]]], dtype=tf.float32))
(new_image, new_boxes, new_labels,
new_masks) = preprocessor._strict_random_crop_image(
image, boxes, labels, masks=masks)
with self.test_session() as sess:
new_image, new_boxes, new_labels, new_masks = sess.run([
new_image, new_boxes, new_labels, new_masks])
expected_boxes = np.array([
[0.0, 0.0, 0.75789469, 1.0],
[0.23157893, 0.24050637, 0.75789469, 1.0],
], dtype=np.float32)
self.assertAllEqual(new_image.shape, [190, 237, 3])
self.assertAllEqual(new_masks.shape, [2, 190, 237])
self.assertAllClose(
new_boxes.flatten(), expected_boxes.flatten())
Example #6
| Source File: preprocessor_test.py From motion-rcnn with MIT License | 5 votes |
|
def testStrictRandomCropImageWithKeypoints(self):
image = self.createColorfulTestImage()[0]
boxes = self.createTestBoxes()
labels = self.createTestLabels()
keypoints = self.createTestKeypoints()
with mock.patch.object(
tf.image,
'sample_distorted_bounding_box'
) as mock_sample_distorted_bounding_box:
mock_sample_distorted_bounding_box.return_value = (
tf.constant([6, 143, 0], dtype=tf.int32),
tf.constant([190, 237, -1], dtype=tf.int32),
tf.constant([[[0.03, 0.3575, 0.98, 0.95]]], dtype=tf.float32))
(new_image, new_boxes, new_labels,
new_keypoints) = preprocessor._strict_random_crop_image(
image, boxes, labels, keypoints=keypoints)
with self.test_session() as sess:
new_image, new_boxes, new_labels, new_keypoints = sess.run([
new_image, new_boxes, new_labels, new_keypoints])
expected_boxes = np.array([
[0.0, 0.0, 0.75789469, 1.0],
[0.23157893, 0.24050637, 0.75789469, 1.0],
], dtype=np.float32)
expected_keypoints = np.array([
[[np.nan, np.nan],
[np.nan, np.nan],
[np.nan, np.nan]],
[[0.38947368, 0.07173],
[0.49473682, 0.24050637],
[0.60000002, 0.40928277]]
], dtype=np.float32)
self.assertAllEqual(new_image.shape, [190, 237, 3])
self.assertAllClose(
new_boxes.flatten(), expected_boxes.flatten())
self.assertAllClose(
new_keypoints.flatten(), expected_keypoints.flatten())
Example #7
| Source File: preprocessor_test.py From mtl-ssl with Apache License 2.0 | 5 votes |
|
def testStrictRandomCropImageWithMasks(self):
image = self.createColorfulTestImage()[0]
boxes = self.createTestBoxes()
labels = self.createTestLabels()
masks = tf.random_uniform([2, 200, 400], dtype=tf.float32)
with mock.patch.object(
tf.image,
'sample_distorted_bounding_box'
) as mock_sample_distorted_bounding_box:
mock_sample_distorted_bounding_box.return_value = (
tf.constant([6, 143, 0], dtype=tf.int32),
tf.constant([190, 237, -1], dtype=tf.int32),
tf.constant([[[0.03, 0.3575, 0.98, 0.95]]], dtype=tf.float32))
(new_image, new_boxes, new_labels,
new_masks) = preprocessor._strict_random_crop_image(
image, boxes, labels, masks=masks)
with self.test_session() as sess:
new_image, new_boxes, new_labels, new_masks = sess.run([
new_image, new_boxes, new_labels, new_masks])
expected_boxes = np.array([
[0.0, 0.0, 0.75789469, 1.0],
[0.23157893, 0.24050637, 0.75789469, 1.0],
], dtype=np.float32)
self.assertAllEqual(new_image.shape, [190, 237, 3])
self.assertAllEqual(new_masks.shape, [2, 190, 237])
self.assertAllClose(
new_boxes.flatten(), expected_boxes.flatten())
Example #8
| Source File: preprocessor_test.py From mtl-ssl with Apache License 2.0 | 5 votes |
|
def testStrictRandomCropImageWithKeypoints(self):
image = self.createColorfulTestImage()[0]
boxes = self.createTestBoxes()
labels = self.createTestLabels()
keypoints = self.createTestKeypoints()
with mock.patch.object(
tf.image,
'sample_distorted_bounding_box'
) as mock_sample_distorted_bounding_box:
mock_sample_distorted_bounding_box.return_value = (
tf.constant([6, 143, 0], dtype=tf.int32),
tf.constant([190, 237, -1], dtype=tf.int32),
tf.constant([[[0.03, 0.3575, 0.98, 0.95]]], dtype=tf.float32))
(new_image, new_boxes, new_labels,
new_keypoints) = preprocessor._strict_random_crop_image(
image, boxes, labels, keypoints=keypoints)
with self.test_session() as sess:
new_image, new_boxes, new_labels, new_keypoints = sess.run([
new_image, new_boxes, new_labels, new_keypoints])
expected_boxes = np.array([
[0.0, 0.0, 0.75789469, 1.0],
[0.23157893, 0.24050637, 0.75789469, 1.0],
], dtype=np.float32)
expected_keypoints = np.array([
[[np.nan, np.nan],
[np.nan, np.nan],
[np.nan, np.nan]],
[[0.38947368, 0.07173],
[0.49473682, 0.24050637],
[0.60000002, 0.40928277]]
], dtype=np.float32)
self.assertAllEqual(new_image.shape, [190, 237, 3])
self.assertAllClose(
new_boxes.flatten(), expected_boxes.flatten())
self.assertAllClose(
new_keypoints.flatten(), expected_keypoints.flatten())
Example #9
| Source File: preprocessor_test.py From multilabel-image-classification-tensorflow with MIT License | 5 votes |
|
def testStrictRandomCropImageWithGroundtruthWeights(self):
image = self.createColorfulTestImage()[0]
boxes = self.createTestBoxes()
labels = self.createTestLabels()
weights = self.createTestGroundtruthWeights()
with mock.patch.object(
tf.image,
'sample_distorted_bounding_box'
) as mock_sample_distorted_bounding_box:
mock_sample_distorted_bounding_box.return_value = (
tf.constant([6, 143, 0], dtype=tf.int32),
tf.constant([190, 237, -1], dtype=tf.int32),
tf.constant([[[0.03, 0.3575, 0.98, 0.95]]], dtype=tf.float32))
new_image, new_boxes, new_labels, new_groundtruth_weights = (
preprocessor._strict_random_crop_image(
image, boxes, labels, weights))
with self.test_session() as sess:
new_image, new_boxes, new_labels, new_groundtruth_weights = (
sess.run(
[new_image, new_boxes, new_labels, new_groundtruth_weights])
)
expected_boxes = np.array(
[[0.0, 0.0, 0.75789469, 1.0],
[0.23157893, 0.24050637, 0.75789469, 1.0]], dtype=np.float32)
self.assertAllEqual(new_image.shape, [190, 237, 3])
self.assertAllEqual(new_groundtruth_weights, [1.0, 0.5])
self.assertAllClose(
new_boxes.flatten(), expected_boxes.flatten())
Example #10
| Source File: preprocessor_test.py From multilabel-image-classification-tensorflow with MIT License | 5 votes |
|
def testStrictRandomCropImageWithMasks(self):
image = self.createColorfulTestImage()[0]
boxes = self.createTestBoxes()
labels = self.createTestLabels()
weights = self.createTestGroundtruthWeights()
masks = tf.random_uniform([2, 200, 400], dtype=tf.float32)
with mock.patch.object(
tf.image,
'sample_distorted_bounding_box'
) as mock_sample_distorted_bounding_box:
mock_sample_distorted_bounding_box.return_value = (
tf.constant([6, 143, 0], dtype=tf.int32),
tf.constant([190, 237, -1], dtype=tf.int32),
tf.constant([[[0.03, 0.3575, 0.98, 0.95]]], dtype=tf.float32))
new_image, new_boxes, new_labels, new_weights, new_masks = (
preprocessor._strict_random_crop_image(
image, boxes, labels, weights, masks=masks))
with self.test_session() as sess:
new_image, new_boxes, new_labels, new_weights, new_masks = sess.run(
[new_image, new_boxes, new_labels, new_weights, new_masks])
expected_boxes = np.array(
[[0.0, 0.0, 0.75789469, 1.0],
[0.23157893, 0.24050637, 0.75789469, 1.0]], dtype=np.float32)
self.assertAllEqual(new_image.shape, [190, 237, 3])
self.assertAllEqual(new_masks.shape, [2, 190, 237])
self.assertAllClose(
new_boxes.flatten(), expected_boxes.flatten())
Example #11
| Source File: preprocessor_test.py From multilabel-image-classification-tensorflow with MIT License | 5 votes |
|
def testStrictRandomCropImageWithKeypoints(self):
image = self.createColorfulTestImage()[0]
boxes = self.createTestBoxes()
labels = self.createTestLabels()
weights = self.createTestGroundtruthWeights()
keypoints = self.createTestKeypoints()
with mock.patch.object(
tf.image,
'sample_distorted_bounding_box'
) as mock_sample_distorted_bounding_box:
mock_sample_distorted_bounding_box.return_value = (
tf.constant([6, 143, 0], dtype=tf.int32),
tf.constant([190, 237, -1], dtype=tf.int32),
tf.constant([[[0.03, 0.3575, 0.98, 0.95]]], dtype=tf.float32))
new_image, new_boxes, new_labels, new_weights, new_keypoints = (
preprocessor._strict_random_crop_image(
image, boxes, labels, weights, keypoints=keypoints))
with self.test_session() as sess:
new_image, new_boxes, new_labels, new_weights, new_keypoints = sess.run(
[new_image, new_boxes, new_labels, new_weights, new_keypoints])
expected_boxes = np.array([
[0.0, 0.0, 0.75789469, 1.0],
[0.23157893, 0.24050637, 0.75789469, 1.0],], dtype=np.float32)
expected_keypoints = np.array([
[[np.nan, np.nan],
[np.nan, np.nan],
[np.nan, np.nan]],
[[0.38947368, 0.07173],
[0.49473682, 0.24050637],
[0.60000002, 0.40928277]]
], dtype=np.float32)
self.assertAllEqual(new_image.shape, [190, 237, 3])
self.assertAllClose(
new_boxes.flatten(), expected_boxes.flatten())
self.assertAllClose(
new_keypoints.flatten(), expected_keypoints.flatten())
Example #12
| Source File: preprocessor_test.py From models with Apache License 2.0 | 5 votes |
|
def testStrictRandomCropImageWithGroundtruthWeights(self):
def graph_fn():
image = self.createColorfulTestImage()[0]
boxes = self.createTestBoxes()
labels = self.createTestLabels()
weights = self.createTestGroundtruthWeights()
with mock.patch.object(
tf.image,
'sample_distorted_bounding_box'
) as mock_sample_distorted_bounding_box:
mock_sample_distorted_bounding_box.return_value = (
tf.constant([6, 143, 0], dtype=tf.int32),
tf.constant([190, 237, -1], dtype=tf.int32),
tf.constant([[[0.03, 0.3575, 0.98, 0.95]]], dtype=tf.float32))
new_image, new_boxes, new_labels, new_groundtruth_weights = (
preprocessor._strict_random_crop_image(
image, boxes, labels, weights))
return [new_image, new_boxes, new_labels, new_groundtruth_weights]
(new_image, new_boxes, _,
new_groundtruth_weights) = self.execute_cpu(graph_fn, [])
expected_boxes = np.array(
[[0.0, 0.0, 0.75789469, 1.0],
[0.23157893, 0.24050637, 0.75789469, 1.0]], dtype=np.float32)
self.assertAllEqual(new_image.shape, [190, 237, 3])
self.assertAllEqual(new_groundtruth_weights, [1.0, 0.5])
self.assertAllClose(
new_boxes.flatten(), expected_boxes.flatten())
Example #13
| Source File: preprocessor_test.py From AniSeg with Apache License 2.0 | 5 votes |
|
def testStrictRandomCropImageWithKeypoints(self):
image = self.createColorfulTestImage()[0]
boxes = self.createTestBoxes()
labels = self.createTestLabels()
keypoints = self.createTestKeypoints()
with mock.patch.object(
tf.image,
'sample_distorted_bounding_box'
) as mock_sample_distorted_bounding_box:
mock_sample_distorted_bounding_box.return_value = (
tf.constant([6, 143, 0], dtype=tf.int32),
tf.constant([190, 237, -1], dtype=tf.int32),
tf.constant([[[0.03, 0.3575, 0.98, 0.95]]], dtype=tf.float32))
new_image, new_boxes, new_labels, new_keypoints = (
preprocessor._strict_random_crop_image(
image, boxes, labels, keypoints=keypoints))
with self.test_session() as sess:
new_image, new_boxes, new_labels, new_keypoints = sess.run(
[new_image, new_boxes, new_labels, new_keypoints])
expected_boxes = np.array([
[0.0, 0.0, 0.75789469, 1.0],
[0.23157893, 0.24050637, 0.75789469, 1.0],], dtype=np.float32)
expected_keypoints = np.array([
[[np.nan, np.nan],
[np.nan, np.nan],
[np.nan, np.nan]],
[[0.38947368, 0.07173],
[0.49473682, 0.24050637],
[0.60000002, 0.40928277]]
], dtype=np.float32)
self.assertAllEqual(new_image.shape, [190, 237, 3])
self.assertAllClose(
new_boxes.flatten(), expected_boxes.flatten())
self.assertAllClose(
new_keypoints.flatten(), expected_keypoints.flatten())
Example #14
| Source File: preprocessor_test.py From AniSeg with Apache License 2.0 | 5 votes |
|
def testStrictRandomCropImageWithMasks(self):
image = self.createColorfulTestImage()[0]
boxes = self.createTestBoxes()
labels = self.createTestLabels()
masks = tf.random_uniform([2, 200, 400], dtype=tf.float32)
with mock.patch.object(
tf.image,
'sample_distorted_bounding_box'
) as mock_sample_distorted_bounding_box:
mock_sample_distorted_bounding_box.return_value = (
tf.constant([6, 143, 0], dtype=tf.int32),
tf.constant([190, 237, -1], dtype=tf.int32),
tf.constant([[[0.03, 0.3575, 0.98, 0.95]]], dtype=tf.float32))
new_image, new_boxes, new_labels, new_masks = (
preprocessor._strict_random_crop_image(
image, boxes, labels, masks=masks))
with self.test_session() as sess:
new_image, new_boxes, new_labels, new_masks = sess.run(
[new_image, new_boxes, new_labels, new_masks])
expected_boxes = np.array(
[[0.0, 0.0, 0.75789469, 1.0],
[0.23157893, 0.24050637, 0.75789469, 1.0]], dtype=np.float32)
self.assertAllEqual(new_image.shape, [190, 237, 3])
self.assertAllEqual(new_masks.shape, [2, 190, 237])
self.assertAllClose(
new_boxes.flatten(), expected_boxes.flatten())
Example #15
| Source File: preprocessor_test.py From AniSeg with Apache License 2.0 | 5 votes |
|
def testStrictRandomCropImageWithLabelScores(self):
image = self.createColorfulTestImage()[0]
boxes = self.createTestBoxes()
labels = self.createTestLabels()
label_scores = self.createTestLabelScores()
with mock.patch.object(
tf.image,
'sample_distorted_bounding_box'
) as mock_sample_distorted_bounding_box:
mock_sample_distorted_bounding_box.return_value = (
tf.constant([6, 143, 0], dtype=tf.int32),
tf.constant([190, 237, -1], dtype=tf.int32),
tf.constant([[[0.03, 0.3575, 0.98, 0.95]]], dtype=tf.float32))
new_image, new_boxes, new_labels, new_label_scores = (
preprocessor._strict_random_crop_image(
image, boxes, labels, label_scores))
with self.test_session() as sess:
new_image, new_boxes, new_labels, new_label_scores = (
sess.run(
[new_image, new_boxes, new_labels, new_label_scores])
)
expected_boxes = np.array(
[[0.0, 0.0, 0.75789469, 1.0],
[0.23157893, 0.24050637, 0.75789469, 1.0]], dtype=np.float32)
self.assertAllEqual(new_image.shape, [190, 237, 3])
self.assertAllEqual(new_label_scores, [1.0, 0.5])
self.assertAllClose(
new_boxes.flatten(), expected_boxes.flatten())
Example #16
| Source File: preprocessor_test.py From MAX-Object-Detector with Apache License 2.0 | 5 votes |
|
def testStrictRandomCropImageWithGroundtruthWeights(self):
image = self.createColorfulTestImage()[0]
boxes = self.createTestBoxes()
labels = self.createTestLabels()
weights = self.createTestGroundtruthWeights()
with mock.patch.object(
tf.image,
'sample_distorted_bounding_box'
) as mock_sample_distorted_bounding_box:
mock_sample_distorted_bounding_box.return_value = (
tf.constant([6, 143, 0], dtype=tf.int32),
tf.constant([190, 237, -1], dtype=tf.int32),
tf.constant([[[0.03, 0.3575, 0.98, 0.95]]], dtype=tf.float32))
new_image, new_boxes, new_labels, new_groundtruth_weights = (
preprocessor._strict_random_crop_image(
image, boxes, labels, weights))
with self.test_session() as sess:
new_image, new_boxes, new_labels, new_groundtruth_weights = (
sess.run(
[new_image, new_boxes, new_labels, new_groundtruth_weights])
)
expected_boxes = np.array(
[[0.0, 0.0, 0.75789469, 1.0],
[0.23157893, 0.24050637, 0.75789469, 1.0]], dtype=np.float32)
self.assertAllEqual(new_image.shape, [190, 237, 3])
self.assertAllEqual(new_groundtruth_weights, [1.0, 0.5])
self.assertAllClose(
new_boxes.flatten(), expected_boxes.flatten())
Example #17
| Source File: preprocessor_test.py From object_detection_with_tensorflow with MIT License | 5 votes |
|
def testStrictRandomCropImageWithLabelScores(self):
image = self.createColorfulTestImage()[0]
boxes = self.createTestBoxes()
labels = self.createTestLabels()
label_scores = self.createTestLabelScores()
with mock.patch.object(
tf.image,
'sample_distorted_bounding_box'
) as mock_sample_distorted_bounding_box:
mock_sample_distorted_bounding_box.return_value = (
tf.constant([6, 143, 0], dtype=tf.int32),
tf.constant([190, 237, -1], dtype=tf.int32),
tf.constant([[[0.03, 0.3575, 0.98, 0.95]]], dtype=tf.float32))
new_image, new_boxes, new_labels, new_label_scores = (
preprocessor._strict_random_crop_image(
image, boxes, labels, label_scores))
with self.test_session() as sess:
new_image, new_boxes, new_labels, new_label_scores = (
sess.run(
[new_image, new_boxes, new_labels, new_label_scores])
)
expected_boxes = np.array(
[[0.0, 0.0, 0.75789469, 1.0],
[0.23157893, 0.24050637, 0.75789469, 1.0]], dtype=np.float32)
self.assertAllEqual(new_image.shape, [190, 237, 3])
self.assertAllEqual(new_label_scores, [1.0, 0.5])
self.assertAllClose(
new_boxes.flatten(), expected_boxes.flatten())
Example #18
| Source File: preprocessor_test.py From object_detection_with_tensorflow with MIT License | 5 votes |
|
def testStrictRandomCropImageWithKeypoints(self):
image = self.createColorfulTestImage()[0]
boxes = self.createTestBoxes()
labels = self.createTestLabels()
keypoints = self.createTestKeypoints()
with mock.patch.object(
tf.image,
'sample_distorted_bounding_box'
) as mock_sample_distorted_bounding_box:
mock_sample_distorted_bounding_box.return_value = (
tf.constant([6, 143, 0], dtype=tf.int32),
tf.constant([190, 237, -1], dtype=tf.int32),
tf.constant([[[0.03, 0.3575, 0.98, 0.95]]], dtype=tf.float32))
new_image, new_boxes, new_labels, new_keypoints = (
preprocessor._strict_random_crop_image(
image, boxes, labels, keypoints=keypoints))
with self.test_session() as sess:
new_image, new_boxes, new_labels, new_keypoints = sess.run(
[new_image, new_boxes, new_labels, new_keypoints])
expected_boxes = np.array([
[0.0, 0.0, 0.75789469, 1.0],
[0.23157893, 0.24050637, 0.75789469, 1.0],], dtype=np.float32)
expected_keypoints = np.array([
[[np.nan, np.nan],
[np.nan, np.nan],
[np.nan, np.nan]],
[[0.38947368, 0.07173],
[0.49473682, 0.24050637],
[0.60000002, 0.40928277]]
], dtype=np.float32)
self.assertAllEqual(new_image.shape, [190, 237, 3])
self.assertAllClose(
new_boxes.flatten(), expected_boxes.flatten())
self.assertAllClose(
new_keypoints.flatten(), expected_keypoints.flatten())
Example #19
| Source File: preprocessor_test.py From object_detection_with_tensorflow with MIT License | 5 votes |
|
def testStrictRandomCropImageWithMasks(self):
image = self.createColorfulTestImage()[0]
boxes = self.createTestBoxes()
labels = self.createTestLabels()
masks = tf.random_uniform([2, 200, 400], dtype=tf.float32)
with mock.patch.object(
tf.image,
'sample_distorted_bounding_box'
) as mock_sample_distorted_bounding_box:
mock_sample_distorted_bounding_box.return_value = (
tf.constant([6, 143, 0], dtype=tf.int32),
tf.constant([190, 237, -1], dtype=tf.int32),
tf.constant([[[0.03, 0.3575, 0.98, 0.95]]], dtype=tf.float32))
new_image, new_boxes, new_labels, new_masks = (
preprocessor._strict_random_crop_image(
image, boxes, labels, masks=masks))
with self.test_session() as sess:
new_image, new_boxes, new_labels, new_masks = sess.run(
[new_image, new_boxes, new_labels, new_masks])
expected_boxes = np.array(
[[0.0, 0.0, 0.75789469, 1.0],
[0.23157893, 0.24050637, 0.75789469, 1.0]], dtype=np.float32)
self.assertAllEqual(new_image.shape, [190, 237, 3])
self.assertAllEqual(new_masks.shape, [2, 190, 237])
self.assertAllClose(
new_boxes.flatten(), expected_boxes.flatten())
Example #20
| Source File: preprocessor_test.py From object_detection_with_tensorflow with MIT License | 5 votes |
|
def testStrictRandomCropImageWithLabelScores(self):
image = self.createColorfulTestImage()[0]
boxes = self.createTestBoxes()
labels = self.createTestLabels()
label_scores = self.createTestLabelScores()
with mock.patch.object(
tf.image,
'sample_distorted_bounding_box'
) as mock_sample_distorted_bounding_box:
mock_sample_distorted_bounding_box.return_value = (
tf.constant([6, 143, 0], dtype=tf.int32),
tf.constant([190, 237, -1], dtype=tf.int32),
tf.constant([[[0.03, 0.3575, 0.98, 0.95]]], dtype=tf.float32))
new_image, new_boxes, new_labels, new_label_scores = (
preprocessor._strict_random_crop_image(
image, boxes, labels, label_scores))
with self.test_session() as sess:
new_image, new_boxes, new_labels, new_label_scores = (
sess.run(
[new_image, new_boxes, new_labels, new_label_scores])
)
expected_boxes = np.array(
[[0.0, 0.0, 0.75789469, 1.0],
[0.23157893, 0.24050637, 0.75789469, 1.0]], dtype=np.float32)
self.assertAllEqual(new_image.shape, [190, 237, 3])
self.assertAllEqual(new_label_scores, [1.0, 0.5])
self.assertAllClose(
new_boxes.flatten(), expected_boxes.flatten())
Example #21
| Source File: preprocessor_test.py From Elphas with Apache License 2.0 | 5 votes |
|
def testStrictRandomCropImageWithKeypoints(self):
image = self.createColorfulTestImage()[0]
boxes = self.createTestBoxes()
labels = self.createTestLabels()
keypoints = self.createTestKeypoints()
with mock.patch.object(
tf.image,
'sample_distorted_bounding_box'
) as mock_sample_distorted_bounding_box:
mock_sample_distorted_bounding_box.return_value = (
tf.constant([6, 143, 0], dtype=tf.int32),
tf.constant([190, 237, -1], dtype=tf.int32),
tf.constant([[[0.03, 0.3575, 0.98, 0.95]]], dtype=tf.float32))
new_image, new_boxes, new_labels, new_keypoints = (
preprocessor._strict_random_crop_image(
image, boxes, labels, keypoints=keypoints))
with self.test_session() as sess:
new_image, new_boxes, new_labels, new_keypoints = sess.run(
[new_image, new_boxes, new_labels, new_keypoints])
expected_boxes = np.array([
[0.0, 0.0, 0.75789469, 1.0],
[0.23157893, 0.24050637, 0.75789469, 1.0],], dtype=np.float32)
expected_keypoints = np.array([
[[np.nan, np.nan],
[np.nan, np.nan],
[np.nan, np.nan]],
[[0.38947368, 0.07173],
[0.49473682, 0.24050637],
[0.60000002, 0.40928277]]
], dtype=np.float32)
self.assertAllEqual(new_image.shape, [190, 237, 3])
self.assertAllClose(
new_boxes.flatten(), expected_boxes.flatten())
self.assertAllClose(
new_keypoints.flatten(), expected_keypoints.flatten())
Example #22
| Source File: preprocessor_test.py From Elphas with Apache License 2.0 | 5 votes |
|
def testStrictRandomCropImageWithMasks(self):
image = self.createColorfulTestImage()[0]
boxes = self.createTestBoxes()
labels = self.createTestLabels()
masks = tf.random_uniform([2, 200, 400], dtype=tf.float32)
with mock.patch.object(
tf.image,
'sample_distorted_bounding_box'
) as mock_sample_distorted_bounding_box:
mock_sample_distorted_bounding_box.return_value = (
tf.constant([6, 143, 0], dtype=tf.int32),
tf.constant([190, 237, -1], dtype=tf.int32),
tf.constant([[[0.03, 0.3575, 0.98, 0.95]]], dtype=tf.float32))
new_image, new_boxes, new_labels, new_masks = (
preprocessor._strict_random_crop_image(
image, boxes, labels, masks=masks))
with self.test_session() as sess:
new_image, new_boxes, new_labels, new_masks = sess.run(
[new_image, new_boxes, new_labels, new_masks])
expected_boxes = np.array(
[[0.0, 0.0, 0.75789469, 1.0],
[0.23157893, 0.24050637, 0.75789469, 1.0]], dtype=np.float32)
self.assertAllEqual(new_image.shape, [190, 237, 3])
self.assertAllEqual(new_masks.shape, [2, 190, 237])
self.assertAllClose(
new_boxes.flatten(), expected_boxes.flatten())
Example #23
| Source File: preprocessor_test.py From Elphas with Apache License 2.0 | 5 votes |
|
def testStrictRandomCropImageWithLabelScores(self):
image = self.createColorfulTestImage()[0]
boxes = self.createTestBoxes()
labels = self.createTestLabels()
label_scores = self.createTestLabelScores()
with mock.patch.object(
tf.image,
'sample_distorted_bounding_box'
) as mock_sample_distorted_bounding_box:
mock_sample_distorted_bounding_box.return_value = (
tf.constant([6, 143, 0], dtype=tf.int32),
tf.constant([190, 237, -1], dtype=tf.int32),
tf.constant([[[0.03, 0.3575, 0.98, 0.95]]], dtype=tf.float32))
new_image, new_boxes, new_labels, new_label_scores = (
preprocessor._strict_random_crop_image(
image, boxes, labels, label_scores))
with self.test_session() as sess:
new_image, new_boxes, new_labels, new_label_scores = (
sess.run(
[new_image, new_boxes, new_labels, new_label_scores])
)
expected_boxes = np.array(
[[0.0, 0.0, 0.75789469, 1.0],
[0.23157893, 0.24050637, 0.75789469, 1.0]], dtype=np.float32)
self.assertAllEqual(new_image.shape, [190, 237, 3])
self.assertAllEqual(new_label_scores, [1.0, 0.5])
self.assertAllClose(
new_boxes.flatten(), expected_boxes.flatten())
Example #24
| Source File: preprocessor_test.py From MBMD with MIT License | 5 votes |
|
def testStrictRandomCropImageWithKeypoints(self):
image = self.createColorfulTestImage()[0]
boxes = self.createTestBoxes()
labels = self.createTestLabels()
keypoints = self.createTestKeypoints()
with mock.patch.object(
tf.image,
'sample_distorted_bounding_box'
) as mock_sample_distorted_bounding_box:
mock_sample_distorted_bounding_box.return_value = (
tf.constant([6, 143, 0], dtype=tf.int32),
tf.constant([190, 237, -1], dtype=tf.int32),
tf.constant([[[0.03, 0.3575, 0.98, 0.95]]], dtype=tf.float32))
(new_image, new_boxes, new_labels,
new_keypoints) = preprocessor._strict_random_crop_image(
image, boxes, labels, keypoints=keypoints)
with self.test_session() as sess:
new_image, new_boxes, new_labels, new_keypoints = sess.run([
new_image, new_boxes, new_labels, new_keypoints])
expected_boxes = np.array([
[0.0, 0.0, 0.75789469, 1.0],
[0.23157893, 0.24050637, 0.75789469, 1.0],
], dtype=np.float32)
expected_keypoints = np.array([
[[np.nan, np.nan],
[np.nan, np.nan],
[np.nan, np.nan]],
[[0.38947368, 0.07173],
[0.49473682, 0.24050637],
[0.60000002, 0.40928277]]
], dtype=np.float32)
self.assertAllEqual(new_image.shape, [190, 237, 3])
self.assertAllClose(
new_boxes.flatten(), expected_boxes.flatten())
self.assertAllClose(
new_keypoints.flatten(), expected_keypoints.flatten())
Example #25
| Source File: preprocessor_test.py From MBMD with MIT License | 5 votes |
|
def testStrictRandomCropImageWithMasks(self):
image = self.createColorfulTestImage()[0]
boxes = self.createTestBoxes()
labels = self.createTestLabels()
masks = tf.random_uniform([2, 200, 400], dtype=tf.float32)
with mock.patch.object(
tf.image,
'sample_distorted_bounding_box'
) as mock_sample_distorted_bounding_box:
mock_sample_distorted_bounding_box.return_value = (
tf.constant([6, 143, 0], dtype=tf.int32),
tf.constant([190, 237, -1], dtype=tf.int32),
tf.constant([[[0.03, 0.3575, 0.98, 0.95]]], dtype=tf.float32))
(new_image, new_boxes, new_labels,
new_masks) = preprocessor._strict_random_crop_image(
image, boxes, labels, masks=masks)
with self.test_session() as sess:
new_image, new_boxes, new_labels, new_masks = sess.run([
new_image, new_boxes, new_labels, new_masks])
expected_boxes = np.array([
[0.0, 0.0, 0.75789469, 1.0],
[0.23157893, 0.24050637, 0.75789469, 1.0],
], dtype=np.float32)
self.assertAllEqual(new_image.shape, [190, 237, 3])
self.assertAllEqual(new_masks.shape, [2, 190, 237])
self.assertAllClose(
new_boxes.flatten(), expected_boxes.flatten())
Example #26
| Source File: preprocessor_test.py From object_detection_kitti with Apache License 2.0 | 5 votes |
|
def testStrictRandomCropImageWithKeypoints(self):
image = self.createColorfulTestImage()[0]
boxes = self.createTestBoxes()
labels = self.createTestLabels()
keypoints = self.createTestKeypoints()
with mock.patch.object(
tf.image,
'sample_distorted_bounding_box'
) as mock_sample_distorted_bounding_box:
mock_sample_distorted_bounding_box.return_value = (
tf.constant([6, 143, 0], dtype=tf.int32),
tf.constant([190, 237, -1], dtype=tf.int32),
tf.constant([[[0.03, 0.3575, 0.98, 0.95]]], dtype=tf.float32))
(new_image, new_boxes, new_labels,
new_keypoints) = preprocessor._strict_random_crop_image(
image, boxes, labels, keypoints=keypoints)
with self.test_session() as sess:
new_image, new_boxes, new_labels, new_keypoints = sess.run([
new_image, new_boxes, new_labels, new_keypoints])
expected_boxes = np.array([
[0.0, 0.0, 0.75789469, 1.0],
[0.23157893, 0.24050637, 0.75789469, 1.0],
], dtype=np.float32)
expected_keypoints = np.array([
[[np.nan, np.nan],
[np.nan, np.nan],
[np.nan, np.nan]],
[[0.38947368, 0.07173],
[0.49473682, 0.24050637],
[0.60000002, 0.40928277]]
], dtype=np.float32)
self.assertAllEqual(new_image.shape, [190, 237, 3])
self.assertAllClose(
new_boxes.flatten(), expected_boxes.flatten())
self.assertAllClose(
new_keypoints.flatten(), expected_keypoints.flatten())
Example #27
| Source File: preprocessor_test.py From object_detection_kitti with Apache License 2.0 | 5 votes |
|
def testStrictRandomCropImageWithMasks(self):
image = self.createColorfulTestImage()[0]
boxes = self.createTestBoxes()
labels = self.createTestLabels()
masks = tf.random_uniform([2, 200, 400], dtype=tf.float32)
with mock.patch.object(
tf.image,
'sample_distorted_bounding_box'
) as mock_sample_distorted_bounding_box:
mock_sample_distorted_bounding_box.return_value = (
tf.constant([6, 143, 0], dtype=tf.int32),
tf.constant([190, 237, -1], dtype=tf.int32),
tf.constant([[[0.03, 0.3575, 0.98, 0.95]]], dtype=tf.float32))
(new_image, new_boxes, new_labels,
new_masks) = preprocessor._strict_random_crop_image(
image, boxes, labels, masks=masks)
with self.test_session() as sess:
new_image, new_boxes, new_labels, new_masks = sess.run([
new_image, new_boxes, new_labels, new_masks])
expected_boxes = np.array([
[0.0, 0.0, 0.75789469, 1.0],
[0.23157893, 0.24050637, 0.75789469, 1.0],
], dtype=np.float32)
self.assertAllEqual(new_image.shape, [190, 237, 3])
self.assertAllEqual(new_masks.shape, [2, 190, 237])
self.assertAllClose(
new_boxes.flatten(), expected_boxes.flatten())
Example #28
| Source File: preprocessor_test.py From hands-detection with MIT License | 5 votes |
|
def testStrictRandomCropImageWithKeypoints(self):
image = self.createColorfulTestImage()[0]
boxes = self.createTestBoxes()
labels = self.createTestLabels()
keypoints = self.createTestKeypoints()
with mock.patch.object(
tf.image,
'sample_distorted_bounding_box'
) as mock_sample_distorted_bounding_box:
mock_sample_distorted_bounding_box.return_value = (
tf.constant([6, 143, 0], dtype=tf.int32),
tf.constant([190, 237, -1], dtype=tf.int32),
tf.constant([[[0.03, 0.3575, 0.98, 0.95]]], dtype=tf.float32))
(new_image, new_boxes, new_labels,
new_keypoints) = preprocessor._strict_random_crop_image(
image, boxes, labels, keypoints=keypoints)
with self.test_session() as sess:
new_image, new_boxes, new_labels, new_keypoints = sess.run([
new_image, new_boxes, new_labels, new_keypoints])
expected_boxes = np.array([
[0.0, 0.0, 0.75789469, 1.0],
[0.23157893, 0.24050637, 0.75789469, 1.0],
], dtype=np.float32)
expected_keypoints = np.array([
[[np.nan, np.nan],
[np.nan, np.nan],
[np.nan, np.nan]],
[[0.38947368, 0.07173],
[0.49473682, 0.24050637],
[0.60000002, 0.40928277]]
], dtype=np.float32)
self.assertAllEqual(new_image.shape, [190, 237, 3])
self.assertAllClose(
new_boxes.flatten(), expected_boxes.flatten())
self.assertAllClose(
new_keypoints.flatten(), expected_keypoints.flatten())
Example #29
| Source File: preprocessor_test.py From hands-detection with MIT License | 5 votes |
|
def testStrictRandomCropImageWithMasks(self):
image = self.createColorfulTestImage()[0]
boxes = self.createTestBoxes()
labels = self.createTestLabels()
masks = tf.random_uniform([2, 200, 400], dtype=tf.float32)
with mock.patch.object(
tf.image,
'sample_distorted_bounding_box'
) as mock_sample_distorted_bounding_box:
mock_sample_distorted_bounding_box.return_value = (
tf.constant([6, 143, 0], dtype=tf.int32),
tf.constant([190, 237, -1], dtype=tf.int32),
tf.constant([[[0.03, 0.3575, 0.98, 0.95]]], dtype=tf.float32))
(new_image, new_boxes, new_labels,
new_masks) = preprocessor._strict_random_crop_image(
image, boxes, labels, masks=masks)
with self.test_session() as sess:
new_image, new_boxes, new_labels, new_masks = sess.run([
new_image, new_boxes, new_labels, new_masks])
expected_boxes = np.array([
[0.0, 0.0, 0.75789469, 1.0],
[0.23157893, 0.24050637, 0.75789469, 1.0],
], dtype=np.float32)
self.assertAllEqual(new_image.shape, [190, 237, 3])
self.assertAllEqual(new_masks.shape, [2, 190, 237])
self.assertAllClose(
new_boxes.flatten(), expected_boxes.flatten())
Example #30
| Source File: preprocessor_test.py From DOTA_models with Apache License 2.0 | 5 votes |
|
def testStrictRandomCropImageWithMasks(self):
image = self.createColorfulTestImage()[0]
boxes = self.createTestBoxes()
labels = self.createTestLabels()
masks = tf.random_uniform([2, 200, 400], dtype=tf.float32)
with mock.patch.object(
tf.image,
'sample_distorted_bounding_box'
) as mock_sample_distorted_bounding_box:
mock_sample_distorted_bounding_box.return_value = (
tf.constant([6, 143, 0], dtype=tf.int32),
tf.constant([190, 237, -1], dtype=tf.int32),
tf.constant([[[0.03, 0.3575, 0.98, 0.95]]], dtype=tf.float32))
(new_image, new_boxes, new_labels,
new_masks) = preprocessor._strict_random_crop_image(
image, boxes, labels, masks=masks)
with self.test_session() as sess:
new_image, new_boxes, new_labels, new_masks = sess.run([
new_image, new_boxes, new_labels, new_masks])
expected_boxes = np.array([
[0.0, 0.0, 0.75789469, 1.0],
[0.23157893, 0.24050637, 0.75789469, 1.0],
], dtype=np.float32)
self.assertAllEqual(new_image.shape, [190, 237, 3])
self.assertAllEqual(new_masks.shape, [2, 190, 237])
self.assertAllClose(
new_boxes.flatten(), expected_boxes.flatten())
