^j/ JdZddlZddlZddlmZddlmZddlm Z m Z m Z ddl m Z mZmZmZddlmZmZdd lmZmZmZmZmZerddlZdd lmZej<eZ d Z! dd d de"dzde"dzde"e zdzfdZ#GddedZ$eGddeZ%dgZ&y)z%Image processor class for LayoutLMv2.N)TorchvisionBackend) BatchFeature)group_images_by_shapereorder_images to_pil_image)ChannelDimension ImageInputPILImageResamplingSizeDict) ImagesKwargsUnpack) TensorTypeauto_docstringis_pytesseract_availableloggingrequires_backends) functionalctd|d|z ztd|d|z ztd|d|z ztd|d|z zgS)Nirr)int)boxwidthheights /var/www/ramen.bs-engineer-server.com/venv/lib/python3.12/site-packages/transformers/models/layoutlmv2/image_processing_layoutlmv2.py normalize_boxr*s` DCFUN #$ DCFVO $% DCFUN #$ DCFVO $%  imageznp.ndarray | torch.Tensorlangtesseract_configinput_data_formatczttdgt|dr|jj }n/t |t jst j|}||nd}t||}|j\}}tj||d|}|d|d|d |d |d f\}} } } } t|D cgc]\} }|jr| }} }t|D cgc] \} }| |vs |}} }t| D cgc] \} }| |vs |} } }t| D cgc] \} }| |vs |} } }t| D cgc] \} }| |vs |} } }t| D cgc] \} }| |vs |} } }g}t| | | | D]$\}}}}||||z||zg}|j!|&g}|D]}|j!t#||| 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. pytesseractcpur"dict)r output_typeconfigtextlefttoprrz-Not as many words as there are bounding boxes)rapply_tesseracthasattrr%numpy isinstancenpndarrayarrayrsizer$ image_to_data enumeratestripzipappendrlen)rr r!r" pil_image image_width image_heightdatawordsr,r-rridxwordirrelevant_indicescoord actual_boxesxywh actual_boxnormalized_boxesrs rr.r.3sWo 7ue !!# rzz *+;+G'RU6GHI )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 YsHH*H? H H! H%.H% H+H+% H12H1 H7H7c<eZdZUdZeed<edzed<edzed<y)LayoutLMv2ImageProcessorKwargsa apply_ocr (`bool`, *optional*, defaults to `self.apply_ocr`): 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. apply_ocrNocr_langr!)__name__ __module__ __qualname____doc__bool__annotations__strrrrMrMbs" ODjDj rrMF)totalceZdZeZej ZdddZdZ dZ dZ dZ dZ deeffd Zededeed effd Z dded d ed edddedzdeezdzdededzdedzd efdZxZS)LayoutLMv2ImageProcessor)rrNTr&kwargsc $t|di|y)NrW)super__init__)selfr\ __class__s rr_z!LayoutLMv2ImageProcessor.__init__s "6"rimagesreturnc $t||fi|S)N)r^ preprocess)r`rbr\ras rrez#LayoutLMv2ImageProcessor.preprocesssw!&3F33rz torch.Tensor do_resizer5resamplez7PILImageResampling | tvF.InterpolationMode | int | Nonedisable_groupingreturn_tensorsrNrOr!c  |rt|dg} g} |D]s} | jrtjdt | j || t j\}}| j|| j|ut||\}}i}|jD] \}}|r|j|||}|||<"t||}t||\}}i}|jD]\}}|jd}|||<t||}td|i|}|r  |d < |d <|S) Nr$z]apply_ocr can only be performed on cpu. Tensors will be transferred to cpu before processing.r')rh)rr5rgr pixel_values)r? tensor_typer@boxes)ris_cudalogger warning_oncer.r%r FIRSTr:ritemsresizerflipr)r`rbrfr5rgrhrirNrOr!r\ words_batch boxes_batchrr@rmgrouped_imagesgrouped_images_indexresized_images_groupedshapestacked_imagesresized_imagesprocessed_images_groupedprocessed_imagesr?s r _preprocessz$LayoutLMv2ImageProcessor._preprocesss  dM 2KK *==''w /IIK+;O_OeOe  u""5)""5) *0EV^n/o,,!#%3%9%9%; ; !E>!%>W_!`,: "5 ) ;((>@TU0E^fv/w,,#% %3%9%9%; = !E>+003N.< $U + = **BDXY.2B!CQ_` 'DM'DM r)TNN)rPrQrRrM valid_kwargsr BILINEARrgr5rescale_factorrfrNrOr!rr_rr rrelistrTr rVrr __classcell__)ras@rrZrZus1L!**HC (DNIIH#(F!G#44v>\7]4bn44#'+7^$77 7 L 7 + 7j(4/77*7*7 7rrZ)NN)'rSr0r2torchimage_processing_backendsrimage_processing_utilsrimage_transformsrrr image_utilsr r r r processing_utilsr rutilsrrrrrr$torchvision.transforms.v2rtvF get_loggerrProrrVr.rMrZ__all__rWrrrs, ;2SSUU47   H %$(7; ,# &,# *,#Dj,#--4 ,#^!\!&H1HHV & &r