transforms代码

[docs]class Compose(object):
    """Composes several transforms together.

    Args:
        transforms (list of ``Transform`` objects): list of transforms to compose.

    Example:
        >>> transforms.Compose([
        >>>     transforms.CenterCrop(10),
        >>>     transforms.ToTensor(),
        >>> ])
    """

    def __init__(self, transforms):
        self.transforms = transforms

    def __call__(self, img):
        for t in self.transforms:
            img = t(img)
        return img

    def __repr__(self):
        format_string = self.__class__.__name__ + '('
        for t in self.transforms:
            format_string += '\n'
            format_string += '    {0}'.format(t)
        format_string += '\n)'
        return format_string

transforms.ToTensor()代码，把PIL.Image.Image和numpy.ndarray转换成Tensor，把（H,W,C)转换成（C,H,W)，把（H,W)转换成（1，H,W)，若原数据类型是uint8,则归一化到（0,255）。

[docs]class ToTensor(object):
    """Convert a ``PIL Image`` or ``numpy.ndarray`` to tensor.

    Converts a PIL Image or numpy.ndarray (H x W x C) in the range
    [0, 255] to a torch.FloatTensor of shape (C x H x W) in the range [0.0, 1.0]
    if the PIL Image belongs to one of the modes (L, LA, P, I, F, RGB, YCbCr, RGBA, CMYK, 1)
    or if the numpy.ndarray has dtype = np.uint8

    In the other cases, tensors are returned without scaling.
    """

[docs]    def __call__(self, pic):
        """
        Args:
            pic (PIL Image or numpy.ndarray): Image to be converted to tensor.

        Returns:
            Tensor: Converted image.
        """
        return F.to_tensor(pic)


    def __repr__(self):
        return self.__class__.__name__ + '()'


def to_tensor(pic):
    """Convert a ``PIL Image`` or ``numpy.ndarray`` to tensor.

    See ``ToTensor`` for more details.

    Args:
        pic (PIL Image or numpy.ndarray): Image to be converted to tensor.

    Returns:
        Tensor: Converted image.
    """
    if not(_is_pil_image(pic) or _is_numpy(pic)):
        raise TypeError('pic should be PIL Image or ndarray. Got {}'.format(type(pic)))

    if _is_numpy(pic) and not _is_numpy_image(pic):
        raise ValueError('pic should be 2/3 dimensional. Got {} dimensions.'.format(pic.ndim))

    if isinstance(pic, np.ndarray):
        # handle numpy array
        if pic.ndim == 2:
            pic = pic[:, :, None]

        img = torch.from_numpy(pic.transpose((2, 0, 1)))
        # backward compatibility
        if isinstance(img, torch.ByteTensor):
            return img.float().div(255)
        else:
            return img

    if accimage is not None and isinstance(pic, accimage.Image):
        nppic = np.zeros([pic.channels, pic.height, pic.width], dtype=np.float32)
        pic.copyto(nppic)
        return torch.from_numpy(nppic)

    # handle PIL Image
    if pic.mode == 'I':
        img = torch.from_numpy(np.array(pic, np.int32, copy=False))
    elif pic.mode == 'I;16':
        img = torch.from_numpy(np.array(pic, np.int16, copy=False))
    elif pic.mode == 'F':
        img = torch.from_numpy(np.array(pic, np.float32, copy=False))
    elif pic.mode == '1':
        img = 255 * torch.from_numpy(np.array(pic, np.uint8, copy=False))
    else:
        img = torch.ByteTensor(torch.ByteStorage.from_buffer(pic.tobytes()))
    # PIL image mode: L, LA, P, I, F, RGB, YCbCr, RGBA, CMYK
    if pic.mode == 'YCbCr':
        nchannel = 3
    elif pic.mode == 'I;16':
        nchannel = 1
    else:
        nchannel = len(pic.mode)
    img = img.view(pic.size[1], pic.size[0], nchannel)
    # put it from HWC to CHW format
    # yikes, this transpose takes 80% of the loading time/CPU
    img = img.transpose(0, 1).transpose(0, 2).contiguous()
    if isinstance(img, torch.ByteTensor):
        return img.float().div(255)
    else:
        return img

将ndarray或tensor转换为PIL.Image。mode为Image的数据类型，可以为空。

[docs]class ToPILImage(object):
    """Convert a tensor or an ndarray to PIL Image.

    Converts a torch.*Tensor of shape C x H x W or a numpy ndarray of shape
    H x W x C to a PIL Image while preserving the value range.

    Args:
        mode (`PIL.Image mode`_): color space and pixel depth of input data (optional).
            If ``mode`` is ``None`` (default) there are some assumptions made about the input data:
             - If the input has 4 channels, the ``mode`` is assumed to be ``RGBA``.
             - If the input has 3 channels, the ``mode`` is assumed to be ``RGB``.
             - If the input has 2 channels, the ``mode`` is assumed to be ``LA``.
             - If the input has 1 channel, the ``mode`` is determined by the data type (i.e ``int``, ``float``,
               ``short``).

    .. _PIL.Image mode: https://pillow.readthedocs.io/en/latest/handbook/concepts.html#concept-modes
    """
    def __init__(self, mode=None):
        self.mode = mode

[docs]    def __call__(self, pic):
        """
        Args:
            pic (Tensor or numpy.ndarray): Image to be converted to PIL Image.

        Returns:
            PIL Image: Image converted to PIL Image.

        """
        return F.to_pil_image(pic, self.mode)


    def __repr__(self):
        format_string = self.__class__.__name__ + '('
        if self.mode is not None:
            format_string += 'mode={0}'.format(self.mode)
        format_string += ')'
        return format_string




def to_pil_image(pic, mode=None):
    """Convert a tensor or an ndarray to PIL Image.

    See :class:`~torchvision.transforms.ToPILImage` for more details.

    Args:
        pic (Tensor or numpy.ndarray): Image to be converted to PIL Image.
        mode (`PIL.Image mode`_): color space and pixel depth of input data (optional).

    .. _PIL.Image mode: https://pillow.readthedocs.io/en/latest/handbook/concepts.html#concept-modes

    Returns:
        PIL Image: Image converted to PIL Image.
    """
    if not(isinstance(pic, torch.Tensor) or isinstance(pic, np.ndarray)):
        raise TypeError('pic should be Tensor or ndarray. Got {}.'.format(type(pic)))

    elif isinstance(pic, torch.Tensor):
        if pic.ndimension() not in {2, 3}:
            raise ValueError('pic should be 2/3 dimensional. Got {} dimensions.'.format(pic.ndimension()))

        elif pic.ndimension() == 2:
            # if 2D image, add channel dimension (CHW)
            pic = pic.unsqueeze(0)

    elif isinstance(pic, np.ndarray):
        if pic.ndim not in {2, 3}:
            raise ValueError('pic should be 2/3 dimensional. Got {} dimensions.'.format(pic.ndim))

        elif pic.ndim == 2:
            # if 2D image, add channel dimension (HWC)
            pic = np.expand_dims(pic, 2)

    npimg = pic
    if isinstance(pic, torch.FloatTensor) and mode != 'F':
        pic = pic.mul(255).byte()
    if isinstance(pic, torch.Tensor):
        npimg = np.transpose(pic.numpy(), (1, 2, 0))

    if not isinstance(npimg, np.ndarray):
        raise TypeError('Input pic must be a torch.Tensor or NumPy ndarray, ' +
                        'not {}'.format(type(npimg)))

    if npimg.shape[2] == 1:
        expected_mode = None
        npimg = npimg[:, :, 0]
        if npimg.dtype == np.uint8:
            expected_mode = 'L'
        elif npimg.dtype == np.int16:
            expected_mode = 'I;16'
        elif npimg.dtype == np.int32:
            expected_mode = 'I'
        elif npimg.dtype == np.float32:
            expected_mode = 'F'
        if mode is not None and mode != expected_mode:
            raise ValueError("Incorrect mode ({}) supplied for input type {}. Should be {}"
                             .format(mode, np.dtype, expected_mode))
        mode = expected_mode

    elif npimg.shape[2] == 2:
        permitted_2_channel_modes = ['LA']
        if mode is not None and mode not in permitted_2_channel_modes:
            raise ValueError("Only modes {} are supported for 2D inputs".format(permitted_2_channel_modes))

        if mode is None and npimg.dtype == np.uint8:
            mode = 'LA'

    elif npimg.shape[2] == 4:
        permitted_4_channel_modes = ['RGBA', 'CMYK', 'RGBX']
        if mode is not None and mode not in permitted_4_channel_modes:
            raise ValueError("Only modes {} are supported for 4D inputs".format(permitted_4_channel_modes))

        if mode is None and npimg.dtype == np.uint8:
            mode = 'RGBA'
    else:
        permitted_3_channel_modes = ['RGB', 'YCbCr', 'HSV']
        if mode is not None and mode not in permitted_3_channel_modes:
            raise ValueError("Only modes {} are supported for 3D inputs".format(permitted_3_channel_modes))
        if mode is None and npimg.dtype == np.uint8:
            mode = 'RGB'

    if mode is None:
        raise TypeError('Input type {} is not supported'.format(npimg.dtype))

    return Image.fromarray(npimg, mode=mode)

transforms.Normalize，创建Normalize对象是大小为C的mean和std列表，调用是（C，H，W）的tensor_image。

[docs]class Normalize(object):
    """Normalize a tensor image with mean and standard deviation.
    Given mean: ``(M1,...,Mn)`` and std: ``(S1,..,Sn)`` for ``n`` channels, this transform
    will normalize each channel of the input ``torch.*Tensor`` i.e.
    ``input[channel] = (input[channel] - mean[channel]) / std[channel]``

    .. note::
        This transform acts out of place, i.e., it does not mutates the input tensor.

    Args:
        mean (sequence): Sequence of means for each channel.
        std (sequence): Sequence of standard deviations for each channel.
        inplace(bool,optional): Bool to make this operation in-place.

    """

    def __init__(self, mean, std, inplace=False):
        self.mean = mean
        self.std = std
        self.inplace = inplace

[docs]    def __call__(self, tensor):
        """
        Args:
            tensor (Tensor): Tensor image of size (C, H, W) to be normalized.

        Returns:
            Tensor: Normalized Tensor image.
        """
        return F.normalize(tensor, self.mean, self.std, self.inplace)


    def __repr__(self):
        return self.__class__.__name__ + '(mean={0}, std={1})'.format(self.mean, self.std)



def normalize(tensor, mean, std, inplace=False):
    """Normalize a tensor image with mean and standard deviation.

    .. note::
        This transform acts out of place by default, i.e., it does not mutates the input tensor.

    See :class:`~torchvision.transforms.Normalize` for more details.

    Args:
        tensor (Tensor): Tensor image of size (C, H, W) to be normalized.
        mean (sequence): Sequence of means for each channel.
        std (sequence): Sequence of standard deviations for each channel.
        inplace(bool,optional): Bool to make this operation inplace.

    Returns:
        Tensor: Normalized Tensor image.
    """
    if not _is_tensor_image(tensor):
        raise TypeError('tensor is not a torch image.')

    if not inplace:
        tensor = tensor.clone()

    dtype = tensor.dtype
    mean = torch.as_tensor(mean, dtype=dtype, device=tensor.device)
    std = torch.as_tensor(std, dtype=dtype, device=tensor.device)
    #将mean和std转换成（C，1,1)
    tensor.sub_(mean[:, None, None]).div_(std[:, None, None])
    return tensor

transform.LinearTransformation。先把图像flatten成（1，C*H*W)的形状的向量，再减去mean_vector (Tensor): tensor （1， D = C x H x W）乘以transformation_matrix (Tensor): tensor （D = C x H x W，D = C x H x W），再flat_tensor.view(tensor.size())。

[docs]class LinearTransformation(object):
    """Transform a tensor image with a square transformation matrix and a mean_vector computed
    offline.
    Given transformation_matrix and mean_vector, will flatten the torch.*Tensor and
    subtract mean_vector from it which is then followed by computing the dot
    product with the transformation matrix and then reshaping the tensor to its
    original shape.

    Applications:
        whitening transformation: Suppose X is a column vector zero-centered data.
        Then compute the data covariance matrix [D x D] with torch.mm(X.t(), X),
        perform SVD on this matrix and pass it as transformation_matrix.

    Args:
        transformation_matrix (Tensor): tensor [D x D], D = C x H x W
        mean_vector (Tensor): tensor [D], D = C x H x W
    """

    def __init__(self, transformation_matrix, mean_vector):
        if transformation_matrix.size(0) != transformation_matrix.size(1):
            raise ValueError("transformation_matrix should be square. Got " +
                             "[{} x {}] rectangular matrix.".format(*transformation_matrix.size()))

        if mean_vector.size(0) != transformation_matrix.size(0):
            raise ValueError("mean_vector should have the same length {}".format(mean_vector.size(0)) +
                             " as any one of the dimensions of the transformation_matrix [{} x {}]"
                             .format(transformation_matrix.size()))

        self.transformation_matrix = transformation_matrix
        self.mean_vector = mean_vector

    def __call__(self, tensor):
        """
        Args:
            tensor (Tensor): Tensor image of size (C, H, W) to be whitened.

        Returns:
            Tensor: Transformed image.
        """
        if tensor.size(0) * tensor.size(1) * tensor.size(2) != self.transformation_matrix.size(0):
            raise ValueError("tensor and transformation matrix have incompatible shape." +
                             "[{} x {} x {}] != ".format(*tensor.size()) +
                             "{}".format(self.transformation_matrix.size(0)))
        flat_tensor = tensor.view(1, -1) - self.mean_vector
        transformed_tensor = torch.mm(flat_tensor, self.transformation_matrix)
        tensor = transformed_tensor.view(tensor.size())
        return tensor

    def __repr__(self):
        format_string = self.__class__.__name__ + '(transformation_matrix='
        format_string += (str(self.transformation_matrix.tolist()) + ')')
        format_string += (", (mean_vector=" + str(self.mean_vector.tolist()) + ')')
        return format_string

transforms.RandomErasing。输入torch_image，（C，H，W），随机是否擦除，随机擦除面积，随机擦除长宽比例，随机位置，随机通道，随机值擦除。

randomerase=transforms.RandomErasing(p=1,scale=(0.2,0.5),value=(255,255,0))
im=Image.open(r'C:\Users\Administrator\Desktop\panda.jpg')
#im_data=torch.from_numpy(np.array(im))
im_data_t=transforms.ToTensor()(im)
erased=randomerase(im_data_t)
erased_im=transforms.ToPILImage()(erased)

[docs]class RandomErasing(object):
    """ Randomly selects a rectangle region in an image and erases its pixels.
        'Random Erasing Data Augmentation' by Zhong et al.
        See https://arxiv.org/pdf/1708.04896.pdf
    Args:
         p: probability that the random erasing operation will be performed.
         scale: range of proportion of erased area against input image.
         ratio: range of aspect ratio of erased area.
         value: erasing value. Default is 0. If a single int, it is used to
            erase all pixels. If a tuple of length 3, it is used to erase
            R, G, B channels respectively.
            If a str of 'random', erasing each pixel with random values.
         inplace: boolean to make this transform inplace. Default set to False.

    Returns:
        Erased Image.
    # Examples:
        >>> transform = transforms.Compose([
        >>> transforms.RandomHorizontalFlip(),
        >>> transforms.ToTensor(),
        >>> transforms.Normalize((0.485, 0.456, 0.406), (0.229, 0.224, 0.225)),
        >>> transforms.RandomErasing(),
        >>> ])
    """

    def __init__(self, p=0.5, scale=(0.02, 0.33), ratio=(0.3, 3.3), value=0, inplace=False):
        assert isinstance(value, (numbers.Number, str, tuple, list))
        if (scale[0] > scale[1]) or (ratio[0] > ratio[1]):
            warnings.warn("range should be of kind (min, max)")
        if scale[0] < 0 or scale[1] > 1:
            raise ValueError("range of scale should be between 0 and 1")
        if p < 0 or p > 1:
            raise ValueError("range of random erasing probability should be between 0 and 1")

        self.p = p
        self.scale = scale
        self.ratio = ratio
        self.value = value
        self.inplace = inplace

    @staticmethod
    def get_params(img, scale, ratio, value=0):
        """Get parameters for ``erase`` for a random erasing.

        Args:
            img (Tensor): Tensor image of size (C, H, W) to be erased.
            scale: range of proportion of erased area against input image.
            ratio: range of aspect ratio of erased area.

        Returns:
            tuple: params (i, j, h, w, v) to be passed to ``erase`` for random erasing.
        """
        img_c, img_h, img_w = img.shape
        area = img_h * img_w

        for attempt in range(10):
            erase_area = random.uniform(scale[0], scale[1]) * area
            aspect_ratio = random.uniform(ratio[0], ratio[1])

            h = int(round(math.sqrt(erase_area * aspect_ratio)))
            w = int(round(math.sqrt(erase_area / aspect_ratio)))

            if h < img_h and w < img_w:
                i = random.randint(0, img_h - h)
                j = random.randint(0, img_w - w)
                if isinstance(value, numbers.Number):
                    v = value
                elif isinstance(value, torch._six.string_classes):
                    v = torch.empty([img_c, h, w], dtype=torch.float32).normal_()
                elif isinstance(value, (list, tuple)):
                    #转换成相同维度，和矩形框相同大小
                    v = torch.tensor(value, dtype=torch.float32).view(-1, 1, 1).expand(-1, h, w)
                return i, j, h, w, v

        # Return original image
        return 0, 0, img_h, img_w, img

    def __call__(self, img):
        """
        Args:
            img (Tensor): Tensor image of size (C, H, W) to be erased.

        Returns:
            img (Tensor): Erased Tensor image.
        """
        if random.uniform(0, 1) < self.p:
            x, y, h, w, v = self.get_params(img, scale=self.scale, ratio=self.ratio, value=self.value)
            return F.erase(img, x, y, h, w, v, self.inplace)
        return img


def erase(img, i, j, h, w, v, inplace=False):
    """ Erase the input Tensor Image with given value.

    Args:
        img (Tensor Image): Tensor image of size (C, H, W) to be erased
        i (int): i in (i,j) i.e coordinates of the upper left corner.
        j (int): j in (i,j) i.e coordinates of the upper left corner.
        h (int): Height of the erased region.
        w (int): Width of the erased region.
        v: Erasing value.
        inplace(bool, optional): For in-place operations. By default is set False.

    Returns:
        Tensor Image: Erased image.
    """
    if not isinstance(img, torch.Tensor):
        raise TypeError('img should be Tensor Image. Got {}'.format(type(img)))

    if not inplace:
        img = img.clone()

    img[:, i:i + h, j:j + w] = v
    return img