Python albumentations.ShiftScaleRotate() Examples

def get_training_augmentation(resize_to=(320,640), crop_size=(288,576)):
    print('[get_training_augmentation] crop_size:', crop_size, ', resize_to:', resize_to) 

    train_transform = [
        albu.HorizontalFlip(p=0.5),
        albu.VerticalFlip(p=0.5),
        albu.ShiftScaleRotate(scale_limit=0.20, rotate_limit=10, shift_limit=0.1, p=0.5, border_mode=cv2.BORDER_CONSTANT, value=0),
        albu.GridDistortion(p=0.5),
        albu.Resize(*resize_to),
        albu.RandomCrop(*crop_size),
        albu.ChannelShuffle(),
        albu.InvertImg(),
        albu.ToGray(),
        albu.Normalize(),
    ]

    return albu.Compose(train_transform) 

def augment(image, boxes):
    h, w, _ = image.shape
    labels, boxes_coord = boxes[:, 0], boxes[:, 1:]
    labels = labels.tolist()
    boxes_coord = boxes_coord * h     # 得到原图尺寸下的坐标（未归一化的坐标）
    boxes_coord[:, 0] = np.clip(boxes_coord[:, 0]-boxes_coord[:, 2]/2, a_min=0, a_max=None)   # 确保x_min和y_min有效
    boxes_coord[:, 1] = np.clip(boxes_coord[:, 1]-boxes_coord[:, 3]/2, a_min=0, a_max=None)
    boxes_coord = boxes_coord.tolist()      # [x_min, y_min, width, height]

    # 在这里设置数据增强的方法
    aug = A.Compose([
        A.HorizontalFlip(p=0.5),
        # A.HueSaturationValue(hue_shift_limit=10, sat_shift_limit=10, val_shift_limit=10, p=0.5),
        # A.ShiftScaleRotate(shift_limit=0.1, scale_limit=0.1, rotate_limit=5, border_mode=0, p=0.5)
    ], bbox_params={'format':'coco', 'label_fields': ['category_id']})

    augmented = aug(image=image, bboxes=boxes_coord, category_id=labels)

    # 经过aug之后，如果把boxes变没了，则返回原来的图片
    if augmented['bboxes']:
        image = augmented['image']

        boxes_coord = np.array(augmented['bboxes']) # x_min, y_min, w, h → x, y, w, h
        boxes_coord[:, 0] = boxes_coord[:, 0] + boxes_coord[:, 2]/2
        boxes_coord[:, 1] = boxes_coord[:, 1] + boxes_coord[:, 3]/2
        boxes_coord = boxes_coord / h
        labels = np.array(augmented['category_id'])[:, None]
        boxes = np.concatenate((labels, boxes_coord), 1)

    return image, boxes 

def augment(image, boxes):
    h, w, _ = image.shape
    labels, boxes_coord = boxes[:, 0], boxes[:, 1:]
    labels = labels.tolist()
    boxes_coord = boxes_coord * h     # 得到原图尺寸下的坐标（未归一化的坐标）
    boxes_coord[:, 0] = np.clip(boxes_coord[:, 0]-boxes_coord[:, 2]/2, a_min=0, a_max=None)   # 确保x_min和y_min有效
    boxes_coord[:, 1] = np.clip(boxes_coord[:, 1]-boxes_coord[:, 3]/2, a_min=0, a_max=None)
    boxes_coord = boxes_coord.tolist()      # [x_min, y_min, width, height]

    # 在这里设置数据增强的方法
    aug = A.Compose([
        A.HorizontalFlip(p=0.5),
        # A.HueSaturationValue(hue_shift_limit=10, sat_shift_limit=10, val_shift_limit=10, p=0.5),
        # A.ShiftScaleRotate(shift_limit=0.1, scale_limit=0.1, rotate_limit=5, border_mode=0, p=0.5)
    ], bbox_params={'format':'coco', 'label_fields': ['category_id']})

    augmented = aug(image=image, bboxes=boxes_coord, category_id=labels)

    # 经过aug之后，如果把boxes变没了，则返回原来的图片
    if augmented['bboxes']:
        image = augmented['image']

        boxes_coord = np.array(augmented['bboxes']) # x_min, y_min, w, h → x, y, w, h
        boxes_coord[:, 0] = boxes_coord[:, 0] + boxes_coord[:, 2]/2
        boxes_coord[:, 1] = boxes_coord[:, 1] + boxes_coord[:, 3]/2
        boxes_coord = boxes_coord / h
        labels = np.array(augmented['category_id'])[:, None]
        boxes = np.concatenate((labels, boxes_coord), 1)

    return image, boxes 

def main():
    image = cv2.imread("images/image_1.jpg")

    keypoints = cv2.goodFeaturesToTrack(
        cv2.cvtColor(image, cv2.COLOR_RGB2GRAY), maxCorners=100, qualityLevel=0.5, minDistance=5
    ).squeeze(1)

    bboxes = [(kp[0] - 10, kp[1] - 10, kp[0] + 10, kp[1] + 10) for kp in keypoints]

    disp_image = visualize(image, keypoints, bboxes)
    plt.figure(figsize=(10, 10))
    plt.imshow(cv2.cvtColor(disp_image, cv2.COLOR_RGB2BGR))
    plt.tight_layout()
    plt.show()

    aug = A.Compose(
        [A.ShiftScaleRotate(scale_limit=0.1, shift_limit=0.2, rotate_limit=10, always_apply=True)],
        bbox_params=A.BboxParams(format="pascal_voc", label_fields=["bbox_labels"]),
        keypoint_params=A.KeypointParams(format="xy"),
    )

    for _i in range(10):
        data = aug(image=image, keypoints=keypoints, bboxes=bboxes, bbox_labels=np.ones(len(bboxes)))

        aug_image = data["image"]
        aug_image = visualize(aug_image, data["keypoints"], data["bboxes"])

        plt.figure(figsize=(10, 10))
        plt.imshow(cv2.cvtColor(aug_image, cv2.COLOR_RGB2BGR))
        plt.tight_layout()
        plt.show() 

def test_shift_scale_rotate_interpolation(interpolation):
    image = np.random.randint(low=0, high=256, size=(100, 100, 3), dtype=np.uint8)
    mask = np.random.randint(low=0, high=2, size=(100, 100), dtype=np.uint8)
    aug = A.ShiftScaleRotate(
        shift_limit=(0.2, 0.2), scale_limit=(1.1, 1.1), rotate_limit=(45, 45), interpolation=interpolation, p=1
    )
    data = aug(image=image, mask=mask)
    expected_image = F.shift_scale_rotate(
        image, angle=45, scale=2.1, dx=0.2, dy=0.2, interpolation=interpolation, border_mode=cv2.BORDER_REFLECT_101
    )
    expected_mask = F.shift_scale_rotate(
        mask, angle=45, scale=2.1, dx=0.2, dy=0.2, interpolation=cv2.INTER_NEAREST, border_mode=cv2.BORDER_REFLECT_101
    )
    assert np.array_equal(data["image"], expected_image)
    assert np.array_equal(data["mask"], expected_mask)