sgG \dZddlmZmZmZmZddlZddlm Z m Z m Z ddl m Z mZmZddlmZmZmZmZmZmZmZmZmZmZmZddlmZmZmZm Z m!Z!m"Z"e rddl#Z#erddl$Z$e!jJe&Z'd Z( dd ejRd ee*d ee*d eeee*ffdZ+Gdde Z,y)z%Image processor class for LayoutLMv3.)DictIterableOptionalUnionN)BaseImageProcessor BatchFeature get_size_dict)resizeto_channel_dimension_format to_pil_image) IMAGENET_STANDARD_MEANIMAGENET_STANDARD_STDChannelDimension ImageInputPILImageResamplinginfer_channel_dimension_formatis_scaled_imagemake_list_of_imagesto_numpy_array valid_imagesvalidate_preprocess_arguments) TensorTypefilter_out_non_signature_kwargsis_pytesseract_availableis_vision_availableloggingrequires_backendsctd|d|z ztd|d|z ztd|d|z ztd|d|z zgS)Nirr)int)boxwidthheights m/var/www/html/venv/lib/python3.12/site-packages/transformers/models/layoutlmv3/image_processing_layoutlmv3.py normalize_boxr'8s` DCFUN #$ DCFVO $% DCFUN #$ DCFVO $%  imagelangtesseract_configinput_data_formatct||}|j\}}tj||d|}|d|d|d|d|df\}} } } } t |D cgc]\} }|j r| }} }t |D cgc] \} }| |vs |}} }t | D cgc] \} }| |vs |} } }t | D cgc] \} }| |vs |} } }t | D cgc] \} }| |vs |} } }t | D cgc] \} }| |vs |} } }g}t | | | | D]$\}}}}||||z||zg}|j|&g}|D]}|jt||| t|t|k(sJd ||fScc}} wcc}} wcc}} wcc}} wcc}} wcc}} w) zdApplies Tesseract OCR on a document image, and returns recognized words + normalized bounding boxes.r,dict)r* output_typeconfigtextlefttopr$r%z-Not as many words as there are bounding boxes) r size pytesseract image_to_data enumeratestripzipappendr'len)r)r*r+r, pil_image image_width image_heightdatawordsr3r4r$r%idxwordirrelevant_indicescoord actual_boxesxywh actual_boxnormalized_boxesr#s r&apply_tesseractrMAsU6GHI )K  $ $YTvVf gD&*6lDL$u+tT[}^bck^l&l#E4eV09/?T)#ttzz|#TT#,U#3 Uic4sBT7TT UE U$-dO Ujc5sBT7TE UD U#,S> SZS%S@R5R5 SC S%.u%5 WzsEDV9VU WE W&/&7 Y U3FX;Xe YF YL$UF3( 1aAE1q5) J'( O c; MNO u:-. ._0__ . " "")U U U S W YsH F(9F( F.F.0 F4=F4 F:F:4 GG G#Gc!ReZdZdZdgZddej ddddddddf dedee e fd ed ed e d ed e e e e fde e e e fdedee dee ddffd Zej ddfdej"dee e fd edee e efdee e efdej"f dZeddddddddddddej*dfdededee e fd ede d ed e e e e fde e e e fdedee dee dee e efdedee e efdej2j2fdZxZS)LayoutLMv3ImageProcessora Constructs a LayoutLMv3 image processor. Args: do_resize (`bool`, *optional*, defaults to `True`): Whether to resize the image's (height, width) dimensions to `(size["height"], size["width"])`. Can be overridden by `do_resize` in `preprocess`. size (`Dict[str, int]` *optional*, defaults to `{"height": 224, "width": 224}`): Size of the image after resizing. Can be overridden by `size` in `preprocess`. resample (`PILImageResampling`, *optional*, defaults to `PILImageResampling.BILINEAR`): Resampling filter to use if resizing the image. Can be overridden by `resample` in `preprocess`. do_rescale (`bool`, *optional*, defaults to `True`): Whether to rescale the image's pixel values by the specified `rescale_value`. Can be overridden by `do_rescale` in `preprocess`. rescale_factor (`float`, *optional*, defaults to 1 / 255): Value by which the image's pixel values are rescaled. Can be overridden by `rescale_factor` in `preprocess`. do_normalize (`bool`, *optional*, defaults to `True`): Whether to normalize the image. Can be overridden by the `do_normalize` parameter in the `preprocess` method. image_mean (`Iterable[float]` or `float`, *optional*, defaults to `IMAGENET_STANDARD_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 (`Iterable[float]` or `float`, *optional*, defaults to `IMAGENET_STANDARD_STD`): Standard deviation 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_std` parameter in the `preprocess` method. apply_ocr (`bool`, *optional*, defaults to `True`): Whether to apply the Tesseract OCR engine to get words + normalized bounding boxes. Can be overridden by the `apply_ocr` parameter in the `preprocess` method. ocr_lang (`str`, *optional*): The language, specified by its ISO code, to be used by the Tesseract OCR engine. By default, English is used. Can be overridden by the `ocr_lang` parameter in the `preprocess` method. tesseract_config (`str`, *optional*): Any additional custom configuration flags that are forwarded to the `config` parameter when calling Tesseract. For example: '--psm 6'. Can be overridden by the `tesseract_config` parameter in the `preprocess` method. pixel_valuesTNgp? do_resizer5resample do_rescale rescale_value do_normalize image_mean image_std apply_ocrocr_langr+returnc  t |di| ||nddd}t|}||_||_||_||_||_||_||nt|_ ||nt|_ | |_ | |_| |_y)N)r%r$)super__init__r rRr5rSrTrescale_factorrVrrWrrXrYrZr+)selfrRr5rSrTrUrVrWrXrYrZr+kwargs __class__s r&r`z!LayoutLMv3ImageProcessor.__init__s "6"'tc-JT""   $+((2(>*DZ&/&;AV"  0r(r) data_formatr,c t|}d|vsd|vrtd|j|d|df}t|f||||d|S)a Resize an image to `(size["height"], size["width"])`. Args: image (`np.ndarray`): Image to resize. size (`Dict[str, int]`): Dictionary in the format `{"height": int, "width": int}` specifying the size of the output image. resample (`PILImageResampling`, *optional*, defaults to `PILImageResampling.BILINEAR`): `PILImageResampling` filter to use when resizing the image e.g. `PILImageResampling.BILINEAR`. data_format (`ChannelDimension` or `str`, *optional*): The channel dimension format for the output image. If unset, the channel dimension format of the input image is used. Can be one of: - `"channels_first"` or `ChannelDimension.FIRST`: image in (num_channels, height, width) format. - `"channels_last"` or `ChannelDimension.LAST`: image in (height, width, num_channels) format. - `"none"` or `ChannelDimension.NONE`: image in (height, width) format. input_data_format (`ChannelDimension` or `str`, *optional*): The channel dimension format for the input image. If unset, the channel dimension format is inferred from the input image. Can be one of: - `"channels_first"` or `ChannelDimension.FIRST`: image in (num_channels, height, width) format. - `"channels_last"` or `ChannelDimension.LAST`: image in (height, width, num_channels) format. - `"none"` or `ChannelDimension.NONE`: image in (height, width) format. Returns: `np.ndarray`: The resized image. r%r$zFThe `size` dictionary must contain the keys `height` and `width`. Got )r5rSrer,)r ValueErrorkeysr )rbr)r5rSrer,rc output_sizes r&r zLayoutLMv3ImageProcessor.resizesyFT" 4 7$#6efjfofofqerst tH~tG}5   #/     r(imagesrareturn_tensorsc ||n |j}||n |j}t|}||n |j}||n |j}||n |j }||n |j }||n |j}| | n |j} | | n |j} | | n |j} | | n |j} t|}t|s tdt||||| ||||Dcgc] }t!|}}t#|dr|rt$j'd|t)|d}| rKt+|dg}g}|D]6}t-|| | |\}}|j/||j/|8|r"|Dcgc]}|j1||||}}|r!|Dcgc]}|j3|||}}|r"|Dcgc]}|j5||| | }}|Dcgc]}t7||| }}t9d |i| }| r |d <|d<|Scc}wcc}wcc}wcc}wcc}w)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. If passing in images with pixel values between 0 and 1, set `do_rescale=False`. do_resize (`bool`, *optional*, defaults to `self.do_resize`): Whether to resize the image. size (`Dict[str, int]`, *optional*, defaults to `self.size`): Desired size of the output image after applying `resize`. resample (`int`, *optional*, defaults to `self.resample`): Resampling filter to use if resizing the image. This can be one of the `PILImageResampling` filters. Only has an effect if `do_resize` is set to `True`. do_rescale (`bool`, *optional*, defaults to `self.do_rescale`): Whether to rescale the image pixel values between [0, 1]. rescale_factor (`float`, *optional*, defaults to `self.rescale_factor`): Rescale factor to apply to the image pixel values. Only has an effect if `do_rescale` is set to `True`. do_normalize (`bool`, *optional*, defaults to `self.do_normalize`): Whether to normalize the image. image_mean (`float` or `Iterable[float]`, *optional*, defaults to `self.image_mean`): Mean values to be used for normalization. Only has an effect if `do_normalize` is set to `True`. image_std (`float` or `Iterable[float]`, *optional*, defaults to `self.image_std`): Standard deviation values to be used for normalization. Only has an effect if `do_normalize` is set to `True`. apply_ocr (`bool`, *optional*, defaults to `self.apply_ocr`): Whether to apply the Tesseract OCR engine to get words + normalized bounding boxes. ocr_lang (`str`, *optional*, defaults to `self.ocr_lang`): The language, specified by its ISO code, to be used by the Tesseract OCR engine. By default, English is used. tesseract_config (`str`, *optional*, defaults to `self.tesseract_config`): Any additional custom configuration flags that are forwarded to the `config` parameter when calling Tesseract. 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. 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. zkInvalid image type. Must be of type PIL.Image.Image, numpy.ndarray, torch.Tensor, tf.Tensor or jax.ndarray.)rTrarVrWrXrRr5rSrzIt 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.r6r.)r)r5rSr,)r)scaler,)r)meanstdr,)input_channel_dimrP)r@ tensor_typerAboxes)rRr5r rSrTrarVrWrXrYrZr+rrrgrrrlogger warning_oncerrrMr;r rescale normalizer r )rbrjrRr5rSrTrarVrWrXrYrZr+rkrer,r) words_batch boxes_batchrArrr@s r& preprocessz#LayoutLMv3ImageProcessor.preprocesssL"+!6IDNN 'tTYYT"'38#-#9Zt +9+E4K^K^'3'?|TEVEV #-#9Zt !*!6IDNN !*!6IDNN '38/?/K+QUQfQf$V,F#:  &!)%! 6<vay I   dM 2KK *.uh@Pduv u""5)""5) * $ %dXYj kF $ 5Rc dF $U^opF ou ej '{N_ `  .&!9~V 'DM'DM c=.    s4H-H24H7H<9I)__name__ __module__ __qualname____doc__model_input_namesrBILINEARboolrstrr"floatrrrr`npndarrayrr rFIRSTrrPILImagery __classcell__)rds@r&rOrOgs$L((#'9'B'B&!4837"&*,1138n1% 1  1  11%%011/0113-1#3-1 1H(:'B'B>BDH . zz. 38n. % . eC)9$9:; . $E#/?*?$@A .  . `%&# $!4837"&*.;?(8(>(>DH!UUU38n U  UUU%%01U/0UU3-U#3-U!sJ!78U&U $E#/?*?$@A!U" #U'Ur(rO)N)-r}typingrrrrnumpyrimage_processing_utilsrr r image_transformsr r r image_utilsrrrrrrrrrrrutilsrrrrrrrr6 get_loggerrzrsr'rrrMrOr^r(r&rs,22UUQQ       H %AE ## ::## 3-##sm## &6&; <= ##LO1Or(