def __call__(self, in_data):
# There are five data augmentation steps
# 1. Color augmentation
# 2. Random expansion
# 3. Random cropping
# 4. Resizing with random interpolation
# 5. Random horizontal flipping
# 6. Random vertical flipping
img, bbox, label = in_data
# 1. Color augmentation
img = random_distort(img)
# 2. Random expansion
if np.random.randint(2):
img, param = transforms.random_expand(img,
fill=self.mean,
return_param=True)
bbox = transforms.translate_bbox(bbox,
y_offset=param['y_offset'],
x_offset=param['x_offset'])
# 3. Random cropping
img, param = random_crop_with_bbox_constraints(img,
bbox,
return_param=True)
bbox, param = transforms.crop_bbox(bbox,
y_slice=param['y_slice'],
x_slice=param['x_slice'],
allow_outside_center=False,
return_param=True)
label = label[param['index']]
# 4. Resizing with random interpolatation
_, H, W = img.shape
img = resize_with_random_interpolation(img, (self.size, self.size))
bbox = transforms.resize_bbox(bbox, (H, W), (self.size, self.size))
# 5. Random horizontal flipping
img, params = transforms.random_flip(img,
x_random=True,
return_param=True)
bbox = transforms.flip_bbox(bbox, (self.size, self.size),
x_flip=params['x_flip'])
# 6. Random vertical flipping
img, params = transforms.random_flip(img,
y_random=True,
return_param=True)
bbox = transforms.flip_bbox(bbox, (self.size, self.size),
y_flip=params['y_flip'])
# Preparation for SSD network
img -= self.mean
mb_loc, mb_label = self.coder.encode(bbox, label)
return img, mb_loc, mb_label
def test_flip_bbox(self):
bbox = np.random.uniform(low=0., high=32., size=(10, 4))
out = flip_bbox(bbox, size=(32, 34), x_flip=True)
bbox_expected = bbox.copy()
bbox_expected[:, 0] = 31 - bbox[:, 2]
bbox_expected[:, 2] = 31 - bbox[:, 0]
np.testing.assert_equal(out, bbox_expected)
out = flip_bbox(bbox, size=(32, 34), y_flip=True)
bbox_expected = bbox.copy()
bbox_expected[:, 1] = 33 - bbox[:, 3]
bbox_expected[:, 3] = 33 - bbox[:, 1]
np.testing.assert_equal(out, bbox_expected)
def test_flip_bbox(self):
size = (32, 24)
bbox = generate_random_bbox(10, size, 0, min(size))
out = flip_bbox(bbox, size=size, y_flip=True)
bbox_expected = bbox.copy()
bbox_expected[:, 0] = size[0] - bbox[:, 2]
bbox_expected[:, 2] = size[0] - bbox[:, 0]
np.testing.assert_equal(out, bbox_expected)
out = flip_bbox(bbox, size=size, x_flip=True)
bbox_expected = bbox.copy()
bbox_expected[:, 1] = size[1] - bbox[:, 3]
bbox_expected[:, 3] = size[1] - bbox[:, 1]
np.testing.assert_equal(out, bbox_expected)
def __call__(self, in_data):
# There are five data augmentation steps
# 3. Random cropping
# 4. Resizing with random interpolation
# 5. Random horizontal flipping
img, bbox, label = in_data
# 3. Random cropping
if self.random_crop and np.random.rand() > 0.5:
next_img, param = random_crop_with_bbox_constraints(
img,
bbox,
min_scale=min(self.crop_rate),
max_scale=max(self.crop_rate),
return_param=True)
next_bbox, param = transforms.crop_bbox(bbox,
y_slice=param['y_slice'],
x_slice=param['x_slice'],
allow_outside_center=False,
return_param=True)
if (len(label[param['index']]) != 0):
label = label[param['index']]
img, bbox = next_img, next_bbox
# 4. Resizing with random interpolatation
_, H, W = img.shape
img = transforms.resize(img, (self.size, self.size))
bbox = transforms.resize_bbox(bbox, (H, W), (self.size, self.size))
# 5. Random horizontal flipping
if self.flip:
img, params = transforms.random_flip(img,
x_random=True,
return_param=True)
bbox = transforms.flip_bbox(bbox, (self.size, self.size),
x_flip=params['x_flip'])
img -= self.mean
img /= self.std
_, height, width = img.shape
ymin = bbox[:, 0]
xmin = bbox[:, 1]
ymax = bbox[:, 2]
xmax = bbox[:, 3]
one_hot_label = np.eye(self.n_class)[label]
xs = (xmin + (xmax - xmin) // 2) / width
ws = (xmax - xmin) / width
ys = (ymin + (ymax - ymin) // 2) / height
hs = (ymax - ymin) / height
t = [{
'label': l,
'x': x,
'w': w,
'y': y,
'h': h,
'one_hot_label': hot
} for l, x, w, y, h, hot in zip(label, xs, ws, ys, hs, one_hot_label)]
return img, t
def __call__(self, in_data):
img, bbox, label = in_data
_, H, W = img.shape
# random brightness and contrast
img = random_distort(img)
# rotate image
# return a tuple whose elements are rotated image, param.
# k (int in param)represents the number of times the image is rotated by 90 degrees.
img, params = transforms.random_rotate(img, return_param=True)
# restore the new hight and width
_, t_H, t_W = img.shape
# rotate bbox based on renewed parameters
bbox = rotate_bbox(bbox, (H, W), params['k'])
img = self.faster_rcnn.prepare(img)
# prepares the image to match the size of the image to be input into the RCNN
_, o_H, o_W = img.shape
# resize the bounding box according to the image resize
bbox = transforms.resize_bbox(bbox, (t_H, t_W), (o_H, o_W))
# horizontally & vertical flip
# simutaneously flip horizontally and vertically of the image
img, params = transforms.random_flip(img,
x_random=True,
y_random=True,
return_param=True)
# flip the bounding box with respect to the parameter
bbox = transforms.flip_bbox(bbox, (o_H, o_W),
x_flip=params['x_flip'],
y_flip=params['y_flip'])
scale = o_H / t_H
return img, bbox, label, scale
def transform(in_data):
img, bbox, label = in_data
img, params = transforms.random_flip(img, x_random=True, return_param=True)
bbox = transforms.flip_bbox(bbox, img.shape[1:], x_flip=params['x_flip'])
return img, bbox, label
def __call__(self, in_data):
if len(in_data) == 4:
img, mask, label, bbox = in_data
else:
img, bbox, label = in_data
# Flipping
img, params = transforms.random_flip(img,
x_random=True,
return_param=True)
x_flip = params['x_flip']
bbox = transforms.flip_bbox(bbox, img.shape[1:], x_flip=x_flip)
# Scaling and mean subtraction
img, scale = scale_img(img, self.min_size, self.max_size)
img -= self.mean
bbox = bbox * scale
if len(in_data) == 4:
mask = transforms.flip(mask, x_flip=x_flip)
mask = transforms.resize(mask.astype(np.float32),
img.shape[1:],
interpolation=PIL.Image.NEAREST).astype(
np.bool)
return img, bbox, label, mask
else:
return img, bbox, label
def __call__(self, in_data):
img, bbox, label = in_data
# Flipping
img, params = transforms.random_flip(img,
x_random=True,
return_param=True)
x_flip = params['x_flip']
bbox = transforms.flip_bbox(bbox, img.shape[1:], x_flip=x_flip)
return img, bbox, label
def __call__(self, in_data):
if len(in_data) == 6:
img, bbox, label, mask, crowd, area = in_data
elif len(in_data) == 4:
img, bbox, label, mask = in_data
else:
raise ValueError
img = img.transpose(2, 0, 1) # H, W, C -> C, H, W
if not self.train:
if len(in_data) == 6:
return img, bbox, label, mask, crowd, area
elif len(in_data) == 4:
return img, bbox, label, mask
else:
raise ValueError
imgs, sizes, scales = self.mask_rcnn.prepare([img])
# print(type(imgs))
# print(type(sizes))
# print(type(scales))
img = imgs[0]
H, W = sizes[0]
scale = scales[0]
_, o_H, o_W = img.shape
if len(bbox) > 0:
bbox = transforms.resize_bbox(bbox, (H, W), (o_H, o_W))
if len(mask) > 0:
mask = transforms.resize(
mask, size=(o_H, o_W), interpolation=0)
# # horizontally flip
# img, params = transforms.random_flip(
# img, x_random=True, return_param=True)
# bbox = transforms.flip_bbox(
# bbox, (o_H, o_W), x_flip=params['x_flip'])
# if mask.ndim == 2:
# mask = transforms.flip(
# mask[None, :, :], x_flip=params['x_flip'])[0]
# else:
# mask = transforms.flip(mask, x_flip=params['x_flip'])
# horizontally and vartically flip
img, params = transforms.random_flip(
img, y_random=True, x_random=True, return_param=True)
bbox = transforms.flip_bbox(
bbox, (o_H, o_W), y_flip=params['y_flip'], x_flip=params['x_flip'])
if mask.ndim == 2:
mask = transforms.flip(
mask[None, :, :], y_flip=params['y_flip'], x_flip=params['x_flip'])[0]
else:
mask = transforms.flip(mask, y_flip=params['y_flip'], x_flip=params['x_flip'])
return img, bbox, label, mask, scale, sizes
def __call__(self, in_data):
# 5段階のステップでデータの水増しを行う
# 1. 色の拡張
# 2. ランダムな拡大
# 3. ランダムなトリミング
# 4. ランダムな補完の再補正
# 5. ランダムな水平反転
img, bbox, label = in_data
# 1. 色の拡張
# 明るさ，コントラスト，彩度，色相を組み合わせ，データ拡張をする
img = random_distort(img)
# 2. ランダムな拡大
if np.random.randint(2):
# キャンバスの様々な座標に入力画像を置いて，様々な比率の画像を生成し，bounding boxを更新
img, param = transforms.random_expand(img,
fill=self.mean,
return_param=True)
bbox = transforms.translate_bbox(bbox,
y_offset=param['y_offset'],
x_offset=param['x_offset'])
# 3. ランダムなトリミング
img, param = random_crop_with_bbox_constraints(img,
bbox,
return_param=True)
# トリミングされた画像内にbounding boxが入るように調整
bbox, param = transforms.crop_bbox(bbox,
y_slice=param['y_slice'],
x_slice=param['x_slice'],
allow_outside_center=False,
return_param=True)
label = label[param['index']]
# 4. ランダムな補完の再補正
## 画像とbounding boxのリサイズ
_, H, W = img.shape
img = resize_with_random_interpolation(img, (self.size, self.size))
bbox = transforms.resize_bbox(bbox, (H, W), (self.size, self.size))
# 5. ランダムな水平反転
## 画像とbounding boxをランダムに水平方向に反転
img, params = transforms.random_flip(img,
x_random=True,
return_param=True)
bbox = transforms.flip_bbox(bbox, (self.size, self.size),
x_flip=params['x_flip'])
# SSDのネットワークに入力するための準備の処理
img -= self.mean
## SSDに入力するためのloc(デフォルトbounding boxのオフセットとスケール)と
## mb_label(クラスを表す配列)を出力
mb_loc, mb_label = self.coder.encode(bbox, label)
return img, mb_loc, mb_label
def __call__(self, in_data):
img, bbox, label = in_data
_, H, W = img.shape
img = preprocess(img, self.min_size, self.max_size)
_, o_H, o_W = img.shape
scale = o_H / H
bbox = resize_bbox(bbox, (H, W), (o_H, o_W))
# horizontally flip
img, params = random_flip(img, x_random=True, return_param=True)
bbox = flip_bbox(bbox, (o_H, o_W), x_flip=params['x_flip'])
return img, bbox, label, [scale, scale]
def __call__(self, in_data):
img, bbox, label = in_data
_, H, W = img.shape
img = self.faster_rcnn.prepare(img)
_, o_H, o_W = img.shape
scale = o_H / H
bbox = transforms.resize_bbox(bbox, (H, W), (o_H, o_W))
# horizontally flip
img, params = transforms.random_flip(img, x_random=True, return_param=True)
bbox = transforms.flip_bbox(bbox, (o_H, o_W), x_flip=params["x_flip"])
return img, bbox, label, scale
def transform(in_data):
img, bbox, label = in_data
_, H, W = img.shape
img = faster_rcnn.prepare(img)
_, o_H, o_W = img.shape
scale = o_H / H
bbox = transforms.resize_bbox(bbox, (W, H), (o_W, o_H))
# horizontally flip
img, params = transforms.random_flip(
img, x_random=True, return_param=True)
bbox = transforms.flip_bbox(bbox, (o_W, o_H), params['x_flip'])
return img, bbox, label, scale
def __call__(self, in_data):
# There are five data augmentation steps
# 1. Color augmentation
# 2. Random expansion
# 3. Random cropping
# 4. Resizing with random interpolation
# 5. Random horizontal flipping
img, bbox, label = in_data
bbox = np.array(bbox).astype(np.float32)
if len(bbox) == 0:
warnings.warn("No bounding box detected", RuntimeWarning)
img = resize_with_random_interpolation(img, (self.size, self.size))
mb_loc, mb_label = self.coder.encode(bbox, label)
return img, mb_loc, mb_label
# 1. Color augmentation
img = random_distort(img)
# 2. Random expansion
if np.random.randint(2):
img, param = transforms.random_expand(
img, fill=self.mean, return_param=True)
bbox = transforms.translate_bbox(
bbox, y_offset=param['y_offset'], x_offset=param['x_offset'])
# 3. Random cropping
img, param = random_crop_with_bbox_constraints(
img, bbox, return_param=True)
bbox, param = transforms.crop_bbox(
bbox, y_slice=param['y_slice'], x_slice=param['x_slice'],
allow_outside_center=False, return_param=True)
label = label[param['index']]
# 4. Resizing with random interpolatation
_, H, W = img.shape
img = resize_with_random_interpolation(img, (self.size, self.size))
bbox = transforms.resize_bbox(bbox, (H, W), (self.size, self.size))
# 5. Random horizontal flipping
img, params = transforms.random_flip(
img, x_random=True, return_param=True)
bbox = transforms.flip_bbox(
bbox, (self.size, self.size), x_flip=params['x_flip'])
mb_loc, mb_label = self.coder.encode(bbox, label)
return img, mb_loc, mb_label
def __call__(self, in_data):
img, bbox, label = in_data
_, H, W = img.shape
img = self.faster_rcnn.prepare(img)
_, o_H, o_W = img.shape
scale = o_H / H
bbox = transforms.resize_bbox(bbox, (H, W), (o_H, o_W))
# horizontally flip
img, params = transforms.random_flip(
img, x_random=True, return_param=True)
bbox = transforms.flip_bbox(
bbox, (o_H, o_W), x_flip=params['x_flip'])
return img, bbox, label, scale
def __call__(self, in_data):
img, mask, label = in_data
bbox = mask_to_bbox(mask)
_, orig_H, orig_W = img.shape
img = self.fcis.prepare(img)
_, H, W = img.shape
scale = H / orig_H
mask = transforms.resize(mask.astype(np.float32), (H, W))
bbox = transforms.resize_bbox(bbox, (orig_H, orig_W), (H, W))
img, params = transforms.random_flip(img,
x_random=True,
return_param=True)
mask = transforms.flip(mask, x_flip=params['x_flip'])
bbox = transforms.flip_bbox(bbox, (H, W), x_flip=params['x_flip'])
return img, mask, label, bbox, scale
def transform(in_data):
img, bbox = in_data
img -= np.array([103.939, 116.779, 123.68])[:, None, None]
# Resize bounding box to a shape
# with the smaller edge at least at length 600
input_shape = img.shape[1:]
output_shape = _shape_soft_min_hard_max(input_shape, 600, 1200)
img = transforms.resize(img, output_shape)
bbox = transforms.resize_bbox(bbox, input_shape, output_shape)
# horizontally flip
img, flips = transforms.random_flip(img,
horizontal_flip=True,
return_flip=True)
h_flip = flips['h']
bbox = transforms.flip_bbox(bbox, output_shape, h_flip)
return img, bbox
def __call__(self, in_data):
# There are five data augmentation steps
# 1. Color augmentation
# 2. Random expansion
# 3. Random cropping
# 4. Resizing with random interpolation
# 5. Random horizontal flipping
img, bbox, label = in_data
# 1. Color augmentation
img = random_distort(img)
# 2. Random expansion
if np.random.randint(2):
img, param = transforms.random_expand(
img, fill=self.mean, return_param=True)
bbox = transforms.translate_bbox(
bbox, y_offset=param['y_offset'], x_offset=param['x_offset'])
# 3. Random cropping
img, param = random_crop_with_bbox_constraints(
img, bbox, return_param=True)
bbox, param = transforms.crop_bbox(
bbox, y_slice=param['y_slice'], x_slice=param['x_slice'],
allow_outside_center=False, return_param=True)
label = label[param['index']]
# 4. Resizing with random interpolatation
_, H, W = img.shape
img = resize_with_random_interpolation(img, (self.size, self.size))
bbox = transforms.resize_bbox(bbox, (H, W), (self.size, self.size))
# 5. Random horizontal flipping
img, params = transforms.random_flip(
img, x_random=True, return_param=True)
bbox = transforms.flip_bbox(
bbox, (self.size, self.size), x_flip=params['x_flip'])
# Preparation for SSD network
img -= self.mean
mb_loc, mb_label = self.coder.encode(bbox, label)
return img, mb_loc, mb_label
def __call__(self, in_data):
img, bbox, label = in_data
_, H, W = img.shape
# random brightness and contrast
img = random_distort(img)
img = self.faster_rcnn.prepare(img)
_, o_H, o_W = img.shape
bbox = transforms.resize_bbox(bbox, (H, W), (o_H, o_W))
# horizontally & vertical flip
img, params = transforms.random_flip(
img, x_random=True, y_random=True, return_param=True)
bbox = transforms.flip_bbox(
bbox, (o_H, o_W), x_flip=params['x_flip'], y_flip=params['y_flip'])
scale = o_H / t_H
return img, bbox, label, scale
def __call__(self, in_data):
if len(in_data) == 5:
img, label, bbox, mask, i = in_data
elif len(in_data) == 4:
img, bbox, label, i = in_data
_, H, W = img.shape
img = self.net.prepare(img)
_, o_H, o_W = img.shape
scale = o_H / H
if len(bbox) == 0:
return img, [], [], 1
bbox = resize_bbox(bbox, (H, W), (o_H, o_W))
mask = resize(mask, (o_H, o_W))
#horizontal flip
img, params = transforms.random_flip(img,
x_random=True,
return_param=True)
bbox = transforms.flip_bbox(bbox, (o_H, o_W), x_flip=params['x_flip'])
mask = transforms.flip(mask, x_flip=params['x_flip'])
cv2.imwrite("gt_roi.png", mask[0] * 255)
return img, bbox, label, scale, mask
def __call__(self, in_data):
if len(in_data) == 6:
img, bbox, label, mask, crowd, area = in_data
elif len(in_data) == 4:
img, bbox, label, mask = in_data
else:
raise ValueError
img = img.transpose(2, 0, 1) # H, W, C -> C, H, W
if not self.train:
if len(in_data) == 6:
return img, bbox, label, mask, crowd, area
elif len(in_data) == 4:
return img, bbox, label, mask
else:
raise ValueError
_, H, W = img.shape
img = self.mask_rcnn.prepare(img)
_, o_H, o_W = img.shape
scale = o_H / H
if len(bbox) > 0:
bbox = transforms.resize_bbox(bbox, (H, W), (o_H, o_W))
if len(mask) > 0:
mask = transforms.resize(
mask, size=(o_H, o_W), interpolation=0)
# horizontally flip
img, params = transforms.random_flip(
img, x_random=True, return_param=True)
bbox = transforms.flip_bbox(
bbox, (o_H, o_W), x_flip=params['x_flip'])
if mask.ndim == 2:
mask = transforms.flip(
mask[None, :, :], x_flip=params['x_flip'])[0]
else:
mask = transforms.flip(mask, x_flip=params['x_flip'])
return img, bbox, label, mask, scale
def __call__(self, in_data):
if len(in_data)==5:
img, label, bbox, mask, i = in_data
elif len(in_data)==4:
img, bbox, label, i= in_data
label = [self.labelids.index(l) + 1 for l in label]
_, H, W = img.shape
if chainer.config.train:
img = self.net.prepare(img)
_, o_H, o_W = img.shape
scale = o_H / H
if len(bbox)==0:
return img, [],[],1
bbox = resize_bbox(bbox, (H, W), (o_H, o_W))
mask = resize(mask,(o_H, o_W))
if chainer.config.train:
#horizontal flip
img, params = transforms.random_flip(
img, x_random=True, return_param=True)
bbox = transforms.flip_bbox(
bbox, (o_H, o_W), x_flip=params['x_flip'])
mask = transforms.flip(mask, x_flip=params['x_flip'])
return img, bbox, label, scale, mask, i
def __call__(self, in_data):
img, bbox, label = in_data
img = random_distort(img)
if np.random.randint(2):
img, param = transforms.random_expand(img,
fill=self.mean,
return_param=True)
bbox = transforms.translate_bbox(bbox,
y_offset=param['y_offset'],
x_offset=param['x_offset'])
img, param = random_crop_with_bbox_constraints(img,
bbox,
return_param=True)
bbox, param = transforms.crop_bbox(bbox,
y_slice=param['y_slice'],
x_slice=param['x_slice'],
allow_outside_center=False,
return_param=True)
label = label[param['index']]
_, H, W = img.shape
img = resize_with_random_interpolation(img, (self.size, self.size))
bbox = transforms.resize_bbox(bbox, (H, W), (self.size, self.size))
img, params = transforms.random_flip(img,
x_random=True,
return_param=True)
bbox = transforms.flip_bbox(bbox, (self.size, self.size),
x_flip=params['x_flip'])
img -= self.mean
mb_loc, mb_label = self.coder.encode(bbox, label)
return img, mb_loc, mb_label
def __call__(self, in_data):
# There are five data augmentation steps
# 1. Color augmentation
# 2. Random expansion
# 3. Random cropping
# 4. Resizing with random interpolation
# 5. Random horizontal flipping
# mask = None
img, bbox, label, mask = in_data
# TODO: show information
# self._show_img(img)
# self._show_mask(mask)
# 1. Color augmentation
img = random_distort(img)
# self._show_img(img)
# 2. Random expansion
if np.random.randint(2):
img, param = transforms.random_expand(
img, fill=self.mean, return_param=True)
bbox = transforms.translate_bbox(
bbox, y_offset=param['y_offset'], x_offset=param['x_offset'])
if mask is not None:
_, new_height, new_width = img.shape
param['new_height'] = new_height
param['new_width'] = new_width
mask = self._random_expand_mask(mask, param)
# self._show_img(img)
# self._show_mask(mask)
# 3. Random cropping
img, param = random_crop_with_bbox_constraints(
img, bbox, return_param=True)
# self._show_img(img)
mask = self._fixed_crop_mask(mask, param['y_slice'], param['x_slice'])
# self._show_mask(mask)
bbox, param = transforms.crop_bbox(
bbox, y_slice=param['y_slice'], x_slice=param['x_slice'],
allow_outside_center=False, return_param=True)
label = label[param['index']]
# 4. Resizing with random interpolatation
_, H, W = img.shape
img = resize_with_random_interpolation(img, (self.size, self.size))
# self._show_img(img)
if mask is not None:
if mask.size == 0:
raise RuntimeError
mask = self._resize_with_nearest(mask, (self.size, self.size))
# self._show_mask(mask)
bbox = transforms.resize_bbox(bbox, (H, W), (self.size, self.size))
# 5. Random horizontal flipping
img, params = transforms.random_flip(
img, x_random=True, return_param=True)
bbox = transforms.flip_bbox(
bbox, (self.size, self.size), x_flip=params['x_flip'])
if mask is not None:
mask = self._random_flip_mask(mask, x_flip=params['x_flip'], y_flip=params['y_flip'])
# self._show_img(img)
# self._show_mask(mask)
# Preparation for SSD network
img -= self.mean
mb_loc, mb_label = self.coder.encode(bbox, label)
if mask is None:
mask = np.ones([self.size, self.size], dtype=np.int32) * -1
# print("Dtype is :"+str(mask.dtype))
data_type = str(mask.dtype)
target_type = 'int32'
if data_type != target_type:
mask = mask.astype(np.int32)
if img is None:
raise RuntimeError
return img, mb_loc, mb_label, mask
def __call__(self, in_data):
"""in_data includes three datas.
Args:
img(array): Shape is (3, H, W). range is [0, 255].
bbox(array): Shape is (N, 4). (y_min, x_min, y_max, x_max).
range is [0, max size of boxes].
label(array): Classes of bounding boxes.
Returns:
img(array): Shape is (3, out_H, out_W). range is [0, 1].
interpolation value equals to self.value.
"""
# There are five data augmentation steps
# 1. Color augmentation
# 2. Random expansion
# 3. Random cropping
# 4. Resizing with random interpolation
# 5. Random horizontal flipping
if self.count % 10 == 0 and self.count % self.batchsize == 0 and self.count != 0:
self.i += 1
i = self.i % len(self.dim)
self.output_shape = (self.dim[i], self.dim[i])
self.count += 1
img, bbox, label = in_data
# 1. Color augmentation
img = random_distort(img,
brightness_delta=32,
contrast_low=0.5,
contrast_high=1.5,
saturation_low=0.5,
saturation_high=1.5,
hue_delta=25)
# Normalize. range is [0, 1]
img /= 255.0
_, H, W = img.shape
scale = np.random.uniform(0.25, 2)
random_expand = np.random.uniform(0.8, 1.2, 2)
net_h, net_w = self.output_shape
out_h = net_h * scale # random_expand[0]
out_w = net_w * scale # random_expand[1]
if H > W:
out_w = out_h * (float(W) / H) * np.random.uniform(0.8, 1.2)
elif H < W:
out_h = out_w * (float(H) / W) * np.random.uniform(0.8, 1.2)
out_h = int(out_h)
out_w = int(out_w)
img = resize_with_random_interpolation(img, (out_h, out_w))
bbox = transforms.resize_bbox(bbox, (H, W), (out_h, out_w))
if out_h < net_h and out_w < net_w:
img, param = expand(img,
out_h=net_h,
out_w=net_w,
fill=self.value,
return_param=True)
bbox = transforms.translate_bbox(bbox,
y_offset=param['y_offset'],
x_offset=param['x_offset'])
else:
out_h = net_h if net_h > out_h else int(out_h * 1.05)
out_w = net_w if net_w > out_w else int(out_w * 1.05)
img, param = expand(img,
out_h=out_h,
out_w=out_w,
fill=self.value,
return_param=True)
bbox = transforms.translate_bbox(bbox,
y_offset=param['y_offset'],
x_offset=param['x_offset'])
img, param = crop_with_bbox_constraints(img,
bbox,
return_param=True,
crop_height=net_h,
crop_width=net_w)
bbox, param = transforms.crop_bbox(bbox,
y_slice=param['y_slice'],
x_slice=param['x_slice'],
allow_outside_center=False,
return_param=True)
label = label[param['index']]
# 5. Random horizontal flipping # OK
img, params = transforms.random_flip(img,
x_random=True,
return_param=True)
bbox = transforms.flip_bbox(bbox,
self.output_shape,
x_flip=params['x_flip'])
# Preparation for Yolov2 network. scale=[0, 1]
bbox[:, ::2] /= self.output_shape[0] # y
bbox[:, 1::2] /= self.output_shape[1] # x
num_bbox = len(bbox)
len_max = max(num_bbox, self.max_target)
out_bbox = np.zeros((len_max, 4), dtype='f')
out_bbox[:num_bbox] = bbox[:num_bbox]
out_label = np.zeros((len_max), dtype='i')
out_label[:num_bbox] = label
out_bbox = out_bbox[:self.max_target]
out_label = out_label[:self.max_target]
num_array = min(num_bbox, self.max_target)
gmap = create_map_anchor_gt(bbox, self.anchors, self.output_shape,
self.downscale, self.n_boxes, len_max)
gmap = gmap[:self.max_target]
img = np.clip(img, 0, 1)
return img, out_bbox, out_label, gmap, np.array([num_array], dtype='i')
def __call__(self, in_data):
image, label, bbox, mask = in_data
_, H, W = image.shape
cell_size = 32 * self.downscale
# Horizontal flip
if self.options['x_flip']:
if np.random.randint(2) == 0:
image = transforms.flip(image, x_flip=True)
bbox = transforms.flip_bbox(bbox, (H, W), x_flip=True)
mask = transforms.flip(mask, x_flip=True)
# Random rotation (90 or 270 degrees)
if self.options['rotate90']:
assert H == W, 'Height and width must match when `rotate90` is set.'
if np.random.randint(2) == 0: # Rotate?
if np.random.randint(2) == 0: # Counter-clockwise?
bbox = rotate_bbox(bbox, 1, (H, W))
image = np.rot90(image, 1, axes=(1, 2))
mask = np.rot90(mask, 1, axes=(1, 2))
else:
bbox = rotate_bbox(bbox, 3, (H, W))
image = np.rot90(image, 3, axes=(1, 2))
mask = np.rot90(mask, 3, axes=(1, 2))
_, H, W = image.shape
# Zoom in / zoom out
if self.options['zoom'] > 1:
assert self.options[
'scale'] <= 1.0, "`scale` shouldn't be set if `zoom` is set."
max_log_zoom = np.log(self.options['zoom'])
log_zoom = np.random.random() * 2 * max_log_zoom - max_log_zoom
zoom = np.exp(log_zoom)
if zoom > 1:
# Zoom in
y_size, x_size = int(H / zoom), int(W / zoom)
y_offset = np.random.randint(H - y_size + 1)
x_offset = np.random.randint(W - x_size + 1)
y_slice = slice(y_offset, y_offset + y_size)
x_slice = slice(x_offset, x_offset + x_size)
bbox = transforms.crop_bbox(bbox, y_slice, x_slice)
bbox *= zoom
image = transforms.resize(image[:, y_slice, x_slice], (H, W))
mask = transforms.resize(mask[:, y_slice, x_slice], (H, W),
interpolation=Image.NEAREST)
elif zoom < 1:
# Zoom out
y_size, x_size = int(H / zoom), int(W / zoom)
y_offset = np.random.randint(y_size - H + 1)
x_offset = np.random.randint(x_size - W + 1)
bbox = transforms.translate_bbox(bbox, y_offset, x_offset)
new_image = np.zeros((1, y_size, x_size), dtype=np.float32)
new_image[:, y_offset:y_offset + H,
x_offset:x_offset + W] = image
new_mask = np.zeros((1, y_size, x_size), dtype=np.float32)
new_mask[:, y_offset:y_offset + H,
x_offset:x_offset + W] = mask
bbox *= zoom
image = transforms.resize(new_image, (H, W))
mask = transforms.resize(new_mask, (H, W),
interpolation=Image.NEAREST)
# Random scale
if self.options['scale'] > 1.0:
assert self.options[
'crop'], '`crop` must be set if `scale` is set.'
max_log_scale = np.log(self.options['scale'])
log_scale = np.random.random() * 2 * max_log_scale - max_log_scale
scale = np.exp(log_scale)
image = transforms.resize(image, (int(H * scale), int(W * scale)))
mask = transforms.resize(mask, (int(H * scale), int(W * scale)),
interpolation=Image.NEAREST)
_, H, W = image.shape
bbox *= scale
# Random crop
if self.options['crop']:
y_margin = (H - 1) % cell_size + 1
x_margin = (W - 1) % cell_size + 1
y_offset = np.random.randint(y_margin)
x_offset = np.random.randint(x_margin)
y_size = H - y_margin
x_size = W - x_margin
y_slice = slice(y_offset, y_offset + y_size)
x_slice = slice(x_offset, x_offset + x_size)
image = image[:, y_slice, x_slice]
bbox = transforms.crop_bbox(bbox, y_slice, x_slice)
mask = mask[:, y_slice, x_slice]
# Change window width
if self.options['window_width'] > 1.0:
image = (image - 128) * self.options['window_width'] + 128
# Change contrast
if self.options['contrast']:
image += np.random.randint(self.options['contrast'] * 2 +
1) - self.options['contrast']
image = np.clip(image, 0, 255)
# save_data(image, bbox, mask)
return image, label, bbox, mask
def __call__(self, in_data):
img, bbox, label = in_data
_, H, W = img.shape
# random brightness and contrast
img = random_distort(img)
# rotate image
# return a tuple whose elements are rotated image, param.
# k (int in param)represents the number of times the image is rotated by 90 degrees.
img, params = transforms.random_rotate(img, return_param=True)
# restore the new hight and width
_, t_H, t_W = img.shape
# rotate bbox based on renewed parameters
bbox = rotate_bbox(bbox, (H, W), params['k'])
# Random expansion:This method randomly place the input image on
# a larger canvas. The size of the canvas is (rH,rW), r is a random ratio drawn from [1,max_ratio].
# The canvas is filled by a value fill except for the region where the original image is placed.
if np.random.randint(2):
fill_value = img.mean(axis=1).mean(axis=1).reshape(-1, 1, 1)
img, param = transforms.random_expand(img,
max_ratio=2,
fill=fill_value,
return_param=True)
bbox = transforms.translate_bbox(bbox,
y_offset=param['y_offset'],
x_offset=param['x_offset'])
# Random crop
# crops the image with bounding box constraints
img, param = random_crop_with_bbox_constraints(img,
bbox,
min_scale=0.5,
max_aspect_ratio=1.5,
return_param=True)
# this translates bounding boxes to fit within the cropped area of an image, bounding boxes whose centers are outside of the cropped area are removed.
bbox, param = transforms.crop_bbox(bbox,
y_slice=param['y_slice'],
x_slice=param['x_slice'],
allow_outside_center=False,
return_param=True)
#assigning new labels to the bounding boxes after cropping
label = label[param['index']]
# if the bounding boxes are all removed,
if bbox.shape[0] == 0:
img, bbox, label = in_data
# update the height and width of the image
_, t_H, t_W = img.shape
img = self.faster_rcnn.prepare(img)
# prepares the image to match the size of the image to be input into the RCNN
_, o_H, o_W = img.shape
# resize the bounding box according to the image resize
bbox = transforms.resize_bbox(bbox, (t_H, t_W), (o_H, o_W))
# horizontally & vertical flip
# simutaneously flip horizontally and vertically of the image
img, params = transforms.random_flip(img,
x_random=True,
y_random=True,
return_param=True)
# flip the bounding box with respect to the parameter
bbox = transforms.flip_bbox(bbox, (o_H, o_W),
x_flip=params['x_flip'],
y_flip=params['y_flip'])
scale = o_H / t_H
return img, bbox, label, scale
