o i^@sdZddlmZmZddlZddlmZmZm Z ddl m Z m Z m Z mZmZddlmZmZmZmZmZmZmZmZmZmZmZmZddlmZmZm Z dd l!m"Z"e #e$Z%e"r^ddl&Z&Gd d d eZ'd gZ(dS) z!Image processor class for Nougat.)OptionalUnionN)BaseImageProcessor BatchFeature get_size_dict)get_resize_output_image_sizepadresizeto_channel_dimension_format to_pil_image) IMAGENET_DEFAULT_MEANIMAGENET_DEFAULT_STDChannelDimension ImageInputPILImageResamplingget_image_sizeinfer_channel_dimension_formatis_scaled_imagemake_flat_list_of_imagesto_numpy_array valid_imagesvalidate_preprocess_arguments) TensorTypefilter_out_non_signature_kwargslogging)is_vision_availablec%seZdZdZdgZdddejddddddddf deded ee e e fd ed ed ed edede e e fdedee e ee fdee e ee fddffdd ZdejfddZddZ   d-dejde deedee e efdejf dd Z  d.dejd e e e fdee e efdee e efdejf d!d"Z  d.dejd e e e fdee e efdee e efdejf d#d$Zejddfdejd e e e fd edee e efdee e efdejf d%d&Zejddfdejd e e e fd edee e efdee e efdejf d'd(Zedddddddddddddejdfd)edeedeed ee e e fd eed eed eed eedeedee e e fdeedee e ee fdee e ee fd*ee e efdeedee e efdej j f"d+d,Z!Z"S)/NougatImageProcessora Constructs a Nougat image processor. Args: do_crop_margin (`bool`, *optional*, defaults to `True`): Whether to crop the image margins. do_resize (`bool`, *optional*, defaults to `True`): Whether to resize the image's (height, width) dimensions to the specified `size`. Can be overridden by `do_resize` in the `preprocess` method. size (`dict[str, int]` *optional*, defaults to `{"height": 896, "width": 672}`): Size of the image after resizing. Can be overridden by `size` in the `preprocess` method. resample (`PILImageResampling`, *optional*, defaults to `Resampling.BILINEAR`): Resampling filter to use if resizing the image. Can be overridden by `resample` in the `preprocess` method. do_thumbnail (`bool`, *optional*, defaults to `True`): Whether to resize the image using thumbnail method. do_align_long_axis (`bool`, *optional*, defaults to `False`): Whether to align the long axis of the image with the long axis of `size` by rotating by 90 degrees. do_pad (`bool`, *optional*, defaults to `True`): Whether to pad the images to the largest image size in the batch. do_rescale (`bool`, *optional*, defaults to `True`): Whether to rescale the image by the specified scale `rescale_factor`. Can be overridden by the `do_rescale` parameter in the `preprocess` method. rescale_factor (`int` or `float`, *optional*, defaults to `1/255`): Scale factor to use if rescaling the image. Can be overridden by the `rescale_factor` parameter in the `preprocess` method. do_normalize (`bool`, *optional*, defaults to `True`): Whether to normalize the image. Can be overridden by `do_normalize` in the `preprocess` method. image_mean (`float` or `list[float]`, *optional*, defaults to `IMAGENET_DEFAULT_MEAN`): Mean to use if normalizing the image. This is a float or list of floats the length of the number of channels in the image. Can be overridden by the `image_mean` parameter in the `preprocess` method. image_std (`float` or `list[float]`, *optional*, defaults to `IMAGENET_DEFAULT_STD`): Image standard deviation. pixel_valuesTNFgp?do_crop_margin do_resizesizeresample do_thumbnaildo_align_long_axisdo_pad do_rescalerescale_factor do_normalize image_mean image_stdreturnc  stjdi| |dur|nddd}t|}||_||_||_||_||_||_||_ ||_ | |_ | |_ | dur<| nt |_| durH| |_dSt|_dS)Nii)heightwidth)super__init__rrr r!r"r#r$r%r&r'r(r r)rr*)selfrr r!r"r#r$r%r&r'r(r)r*kwargs __class__r.f/home/ubuntu/.local/lib/python3.10/site-packages/transformers/models/nougat/image_processing_nougat.pyr0[szNougatImageProcessor.__init__imagecCs6tt|}|ddddgf}|ddd}|S)zGThis is a reimplementation of a findNonZero function equivalent to cv2.Nr)np column_stacknonzeroreshape)r1r6non_zero_indicesidxvecr.r.r5python_find_non_zero}sz)NougatImageProcessor.python_find_non_zerocCsftj|ddt}tj|ddt}|d|d}}|d|d}|d|d}||||fS)zHThis is a reimplementation of a BoundingRect function equivalent to cv2.rr7)axisrr7)r:minastypeintmax)r1 coordinates min_values max_valuesx_miny_minr-r,r.r.r5python_bounding_rects  z)NougatImageProcessor.python_bounding_rectgray_threshold data_formatinput_data_formatcCs|durt|}t||d}t|dtj}|}|}||krBt|}t ||t j }|dur>t |||}|S|}|S||||d}||k}| |} | | \} } } } || | | | | | f}t|tj}t ||t j }|durt |||}|S|}|S)a Crops the margin of the image. Gray pixels are considered margin (i.e., pixels with a value below the threshold). Args: image (`np.ndarray`): The image to be cropped. gray_threshold (`int`, *optional*, defaults to `200`) Value below which pixels are considered to be gray. data_format (`ChannelDimension`, *optional*): The channel dimension format of the output image. If unset, will use the inferred format from the input. input_data_format (`ChannelDimension`, *optional*): The channel dimension format of the input image. If unset, will use the inferred format from the input. NrPL)rr r:arrayconvertrDuint8rFrCr rLASTr@rLcrop)r1r6rNrOrPdatamax_valmin_valgraycoordsrJrKr-r,r.r.r5 crop_margins:    z NougatImageProcessor.crop_marginc Cst||d\}}|d|d}}|durt|}|tjkr!d} n|tjkr)d} ntd|||kr8||ks@||krH||krHtj|d| d }|durSt|||d }|S) a Align the long axis of the image to the longest axis of the specified size. Args: image (`np.ndarray`): The image to be aligned. size (`dict[str, int]`): The size `{"height": h, "width": w}` to align the long axis to. data_format (`str` or `ChannelDimension`, *optional*): The data format of the output image. If unset, the same format as the input image is used. input_data_format (`ChannelDimension` or `str`, *optional*): The channel dimension format of the input image. If not provided, it will be inferred. Returns: `np.ndarray`: The aligned image.  channel_dimr,r-NrA)r7r9zUnsupported data format: r)axesinput_channel_dim) rrrrWFIRST ValueErrorr:rot90r ) r1r6r!rOrP input_height input_width output_height output_widthrot_axesr.r.r5align_long_axiss  z$NougatImageProcessor.align_long_axiscCsr|d|d}}t||d\}}||} ||} | d} | d} | | } | | }| | f| |ff}t||||dS)a Pad the image to the specified size at the top, bottom, left and right. Args: image (`np.ndarray`): The image to be padded. size (`dict[str, int]`): The size `{"height": h, "width": w}` to pad the image to. data_format (`str` or `ChannelDimension`, *optional*): The data format of the output image. If unset, the same format as the input image is used. input_data_format (`ChannelDimension` or `str`, *optional*): The channel dimension format of the input image. If not provided, it will be inferred. r,r-r_r9rOrP)rr )r1r6r!rOrPrirjrgrh delta_width delta_heightpad_toppad_left pad_bottom pad_rightpaddingr.r.r5 pad_imageszNougatImageProcessor.pad_imagec Kst||d\}}|d|d} } t|| } t|| } | |kr%| |kr%|S||kr2t|| |} n ||kr>t|| |} t|f| | f|d||d|S)as Resize the image to make a thumbnail. The image is resized so that no dimension is larger than any corresponding dimension of the specified size. Args: image (`np.ndarray`): The image to be resized. size (`dict[str, int]`): The size `{"height": h, "width": w}` to resize the image to. resample (`PILImageResampling`, *optional*, defaults to `PILImageResampling.BICUBIC`): The resampling filter to use. data_format (`Optional[Union[str, ChannelDimension]]`, *optional*): The data format of the output image. If unset, the same format as the input image is used. input_data_format (`ChannelDimension` or `str`, *optional*): The channel dimension format of the input image. If not provided, it will be inferred. r_r,r-g@)r!r" reducing_gaprOrP)rrCrEr ) r1r6r!r"rOrPr2rgrhrirjr,r-r.r.r5 thumbnails*  zNougatImageProcessor.thumbnailc KsHt|}t|d|d}t||d|d}t|f||||d|} | S)a Resizes `image` to `(height, width)` specified by `size` using the PIL library. Args: image (`np.ndarray`): Image to resize. size (`dict[str, int]`): Size of the output image. resample (`PILImageResampling`, *optional*, defaults to `PILImageResampling.BICUBIC`): Resampling filter to use when resiizing the image. data_format (`str` or `ChannelDimension`, *optional*): The channel dimension format of the image. If not provided, it will be the same as the input image. input_data_format (`ChannelDimension` or `str`, *optional*): The channel dimension format of the input image. If not provided, it will be inferred. r,r-F)r!default_to_squarerP)r!r"rOrP)rrCrr ) r1r6r!r"rOrPr2 shortest_edge output_size resized_imager.r.r5r Hs zNougatImageProcessor.resizeimagesreturn_tensorsc s$|dur|nj}|dur|nj}durnjdur!nj|dur*|nj}|dur3|nj}|dur<|nj}| durE| nj} durNnj| durW| nj } dur`nj durinj t |}t |sxtdt| | |ddd|D}| rt|drtddurt|d|rfdd|D}|rfd d|D}|rɇfd d|D}|rևfd d|D}|rfd d|D}| rfd d|D}| rfdd|D}fdd|D}d|i}t||dS)a Preprocess an image or batch of images. Args: images (`ImageInput`): Image to preprocess. Expects a single or batch of images with pixel values ranging from 0 to 255. do_crop_margin (`bool`, *optional*, defaults to `self.do_crop_margin`): Whether to crop the image margins. do_resize (`bool`, *optional*, defaults to `self.do_resize`): Whether to resize the image. size (`dict[str, int]`, *optional*, defaults to `self.size`): Size of the image after resizing. Shortest edge of the image is resized to min(size["height"], size["width"]) with the longest edge resized to keep the input aspect ratio. resample (`int`, *optional*, defaults to `self.resample`): Resampling filter to use if resizing the image. This can be one of the enum `PILImageResampling`. Only has an effect if `do_resize` is set to `True`. do_thumbnail (`bool`, *optional*, defaults to `self.do_thumbnail`): Whether to resize the image using thumbnail method. do_align_long_axis (`bool`, *optional*, defaults to `self.do_align_long_axis`): Whether to align the long axis of the image with the long axis of `size` by rotating by 90 degrees. do_pad (`bool`, *optional*, defaults to `self.do_pad`): Whether to pad the images to the largest image size in the batch. do_rescale (`bool`, *optional*, defaults to `self.do_rescale`): Whether to rescale the image by the specified scale `rescale_factor`. rescale_factor (`int` or `float`, *optional*, defaults to `self.rescale_factor`): Scale factor to use if rescaling the image. do_normalize (`bool`, *optional*, defaults to `self.do_normalize`): Whether to normalize the image. image_mean (`float` or `list[float]`, *optional*, defaults to `self.image_mean`): Image mean to use for normalization. image_std (`float` or `list[float]`, *optional*, defaults to `self.image_std`): Image standard deviation to use for normalization. return_tensors (`str` or `TensorType`, *optional*): The type of tensors to return. Can be one of: - Unset: Return a list of `np.ndarray`. - `TensorType.TENSORFLOW` or `'tf'`: Return a batch of type `tf.Tensor`. - `TensorType.PYTORCH` or `'pt'`: Return a batch of type `torch.Tensor`. - `TensorType.NUMPY` or `'np'`: Return a batch of type `np.ndarray`. - `TensorType.JAX` or `'jax'`: Return a batch of type `jax.numpy.ndarray`. data_format (`ChannelDimension` or `str`, *optional*, defaults to `ChannelDimension.FIRST`): The channel dimension format for the output image. Can be one of: - `ChannelDimension.FIRST`: image in (num_channels, height, width) format. - `ChannelDimension.LAST`: image in (height, width, num_channels) format. - Unset: defaults to the channel dimension format of the input image. input_data_format (`ChannelDimension` or `str`, *optional*): The channel dimension format for the input image. If unset, the channel dimension format is inferred from the input image. Can be one of: - `"channels_first"` or `ChannelDimension.FIRST`: image in (num_channels, height, width) format. - `"channels_last"` or `ChannelDimension.LAST`: image in (height, width, num_channels) format. - `"none"` or `ChannelDimension.NONE`: image in (height, width) format. NzkInvalid image type. Must be of type PIL.Image.Image, numpy.ndarray, torch.Tensor, tf.Tensor or jax.ndarray.)r&r'r(r)r*r r!r"cSsg|]}t|qSr.)r.0r6r.r.r5 sz3NougatImageProcessor.preprocess..rzIt looks like you are trying to rescale already rescaled images. If the input images have pixel values between 0 and 1, set `do_rescale=False` to avoid rescaling them again.csg|] }j|dqS)rQ)r^r~)rPr1r.r5rscg|] }j|dqS))r!rP)rlr~rPr1r!r.r5rcsg|] }j|dqS))r6r!r"rP)r r~)rPr"r1r!r.r5rcr)r6r!rP)rwr~rr.r5rrcrr)rur~rr.r5rrcsg|] }j|dqS))r6scalerP)rescaler~)rPr'r1r.r5rscsg|] }j|dqS))r6meanstdrP) normalizer~)r)r*rPr1r.r5rrcsg|] }t|dqS)rb)r r~rmr.r5rsr)rY tensor_type)rr r!r"r#r$r%r&r'r(r)r*rrrerrlogger warning_oncerr)r1r|rr r!r"r#r$r%r&r'r(r)r*r}rOrPrYr.)rOr)r*rPr"r'r1r!r5 preprocessosvG    zNougatImageProcessor.preprocess)rMNN)NN)#__name__ __module__ __qualname____doc__model_input_namesrBILINEARboolrdictstrrErfloatlistr0r:ndarrayr@rLrr^rlruBICUBICrwr rrdrrPILImager __classcell__r.r.r3r5r6sN"     "  9  3  (  7  '     r))rtypingrrnumpyr:image_processing_utilsrrrimage_transformsrr r r r image_utilsr rrrrrrrrrrrutilsrrrutils.import_utilsr get_loggerrrrr__all__r.r.r.r5s 8   U