本文整理汇总了Python中sklearn.feature_extraction.image.PatchExtractor.transform方法的典型用法代码示例。如果您正苦于以下问题：Python PatchExtractor.transform方法的具体用法？Python PatchExtractor.transform怎么用？Python PatchExtractor.transform使用的例子？那么恭喜您, 这里精选的方法代码示例或许可以为您提供帮助。您也可以进一步了解该方法所在类sklearn.feature_extraction.image.PatchExtractor的用法示例。
在下文中一共展示了PatchExtractor.transform方法的7个代码示例，这些例子默认根据受欢迎程度排序。您可以为喜欢或者感觉有用的代码点赞，您的评价将有助于系统推荐出更棒的Python代码示例。
示例1: test_patch_extractor_all_patches
# 需要导入模块: from sklearn.feature_extraction.image import PatchExtractor [as 别名]
# 或者: from sklearn.feature_extraction.image.PatchExtractor import transform [as 别名]
def test_patch_extractor_all_patches():
faces = face_collection
i_h, i_w = faces.shape[1:3]
p_h, p_w = 8, 8
expected_n_patches = len(faces) * (i_h - p_h + 1) * (i_w - p_w + 1)
extr = PatchExtractor(patch_size=(p_h, p_w), random_state=0)
patches = extr.transform(faces)
assert_true(patches.shape == (expected_n_patches, p_h, p_w))示例2: test_patch_extractor_color
# 需要导入模块: from sklearn.feature_extraction.image import PatchExtractor [as 别名]
# 或者: from sklearn.feature_extraction.image.PatchExtractor import transform [as 别名]
def test_patch_extractor_color():
faces = _make_images(orange_face)
i_h, i_w = faces.shape[1:3]
p_h, p_w = 8, 8
expected_n_patches = len(faces) * (i_h - p_h + 1) * (i_w - p_w + 1)
extr = PatchExtractor(patch_size=(p_h, p_w), random_state=0)
patches = extr.transform(faces)
assert_true(patches.shape == (expected_n_patches, p_h, p_w, 3))示例3: test_patch_extractor_max_patches
# 需要导入模块: from sklearn.feature_extraction.image import PatchExtractor [as 别名]
# 或者: from sklearn.feature_extraction.image.PatchExtractor import transform [as 别名]
def test_patch_extractor_max_patches():
faces = face_collection
i_h, i_w = faces.shape[1:3]
p_h, p_w = 8, 8
max_patches = 100
expected_n_patches = len(faces) * max_patches
extr = PatchExtractor(patch_size=(p_h, p_w), max_patches=max_patches,
random_state=0)
patches = extr.transform(faces)
assert patches.shape == (expected_n_patches, p_h, p_w)
max_patches = 0.5
expected_n_patches = len(faces) * int((i_h - p_h + 1) * (i_w - p_w + 1)
* max_patches)
extr = PatchExtractor(patch_size=(p_h, p_w), max_patches=max_patches,
random_state=0)
patches = extr.transform(faces)
assert patches.shape == (expected_n_patches, p_h, p_w)示例4: generate_data
# 需要导入模块: from sklearn.feature_extraction.image import PatchExtractor [as 别名]
# 或者: from sklearn.feature_extraction.image.PatchExtractor import transform [as 别名]
def generate_data(img_folder, max_patches=0.001):
for fpath in get_img_filepaths(img_folder):
print ('Reading image', fpath)
patch_extractor = PatchExtractor(patch_size=(32,32),
max_patches=max_patches)
img_tensor = imread(fpath, mode='RGB')
# shape : (row, col, channels)
input_matrix = np.array([img_tensor])
# shape : (1, row, col, channels)
input_matrix = input_matrix/255.0 # Casting into 0 to 1 space which DNN models learn faster
patches = patch_extractor.transform(input_matrix)
# shape : (n_samples, row, col, channels)
patches = np.rollaxis(patches, axis=3, start=1)
# shape : (n_samples, channels, row, col)
small_patches = np.array([resize(patch) for patch in patches])
# shape : (n_samples, channels, max_x, max_y)
patches = np.array([p.reshape(p.shape[0] * p.shape[1] * p.shape[2])
for p in patches])
# shape : (n_samples, output_vector_size)
if False:
# Print out values to debug
print ("Shapes of tensors", small_patches.shape, patches.shape)
for i, (small, big) in enumerate(zip(small_patches, patches)):
small_img = np.rollaxis(small, axis=0, start=3)
if not os.path.exists('debug'):
os.makedirs('debug')
imsave('debug/small_patch_{}.jpg'.format(i), small_img)
imsave('debug/big_patch_{}.jpg'.format(i), vec2img(big))
yield small_patches, patches示例5: test_patch_extractor_max_patches_default
# 需要导入模块: from sklearn.feature_extraction.image import PatchExtractor [as 别名]
# 或者: from sklearn.feature_extraction.image.PatchExtractor import transform [as 别名]
def test_patch_extractor_max_patches_default():
faces = face_collection
extr = PatchExtractor(max_patches=100, random_state=0)
patches = extr.transform(faces)
assert_equal(patches.shape, (len(faces) * 100, 19, 25))示例6: extract_patches
# 需要导入模块: from sklearn.feature_extraction.image import PatchExtractor [as 别名]
# 或者: from sklearn.feature_extraction.image.PatchExtractor import transform [as 别名]
def extract_patches(self, patch_size, max_patches=None, random_state=None):
patch_extractor = PatchExtractor(patch_size=patch_size, max_patches=np.int(
max_patches / self.num_images()), random_state=random_state)
return patch_extractor.transform(self._images).astype(np.uint8)示例7: test_patch_extractor_max_patches_default
# 需要导入模块: from sklearn.feature_extraction.image import PatchExtractor [as 别名]
# 或者: from sklearn.feature_extraction.image.PatchExtractor import transform [as 别名]
def test_patch_extractor_max_patches_default():
lenas = lena_collection
extr = PatchExtractor(max_patches=100, random_state=0)
patches = extr.transform(lenas)
assert_equal(patches.shape, (len(lenas) * 100, 12, 12))