9iSUSSKJrJrJrJrJr SSKrSSKJ r SSK J r SSK J r SSKJr SS KJr SS KJr SS KJr S S KJr S SKJr S SKJrJrJrJrJrJ r SSK!J"r" \RF"SS\55r$\ "SSS9"SS\55r%g))AnyDictListOptionalUnionN)logging)pipeline_requires_extra)ImageBatchSampler) ReadImage) benchmark) HPIConfig)PaddlePredictorOption)(AutoParallelImageSimpleInferencePipeline) BasePipeline) CropByPolys SortPolyBoxes SortQuadBoxescal_ocr_word_boxconvert_points_to_boxes rotate_image) OCRResultc>^\rSrSrSrS S\S\\S\\S\ S\\ \\\ 4\ 4S S4 U4S jjjr S \\R S \\S \\R 4S jrS\S \ 4SjrS\\ S\\ S\\ S \4SjrS!S\\S\\S\\S\\S\\S\\S \4SjjrS"S\ \\\\R \\R 4S\\ S\\ S\\ S\\S\\S\\S\\S\\S\\S\\S\\ S \4SjjrSrU=r$)# _OCRPipeline'z OCR PipelineNconfigdevice pp_optionuse_hpip hpi_configreturnc >[T U]X#XES9 URSS5UlUR(a:URS05RSSS05nUR U5UlURSS5UlUR (a:URS 05RS S S 05nURU5UlURS 05RS S S05nUSUl URS:XaURSS5Ul URSS5Ul URSS5Ul URSS5Ul URSS5UlURSS5UlURSS5Ul[#5Ul['SS 9UlOURS!:XaURSS"5Ul URSS#5Ul URSS5Ul URSS$5Ul URSS5UlURSS%5UlURSS5Ul[+5Ul['S&S 9UlO$[-S'R/UR55eURUURURURURURUR URS(9UlURS 05RS)S S*05n U RS+S,5UlU RS-S.5UlU RSS5UlURXRS/9UlRS0S15S29Ul[=S3S49Ulg)5af Initializes the class with given configurations and options. Args: config (Dict): Configuration dictionary containing various settings. device (str, optional): Device to run the predictions on. Defaults to None. pp_option (PaddlePredictorOption, optional): PaddlePredictor options. Defaults to None. use_hpip (bool, optional): Whether to use the high-performance inference plugin (HPIP) by default. Defaults to False. hpi_config (Optional[Union[Dict[str, Any], HPIConfig]], optional): The default high-performance inference configuration dictionary. Defaults to None. )r r!r"r#use_doc_preprocessorT SubPipelinesDocPreprocessorpipeline_config_errorz+config error for doc_preprocessor_pipeline!use_textline_orientation SubModulesTextLineOrientationmodel_config_errorz,config error for textline_orientation_model! TextDetectionz config error for text_det_model! text_typegenerallimit_side_leni limit_typemaxmax_side_limitithreshg333333? box_threshg333333? input_shapeN unclip_ratiog@quad) det_box_typesealiming?g?polyzUnsupported text type {})r1r2r4r5r6r8r7TextRecognitionz config error for text_rec_model! score_threshrreturn_word_boxF)r7 batch_sizer)rABGR)format) super__init__getr&create_pipelinedoc_preprocessor_pipeliner* create_modeltextline_orientation_modelr/text_det_limit_side_lentext_det_limit_typetext_det_max_side_limittext_det_threshtext_det_box_threshr7text_det_unclip_ratior _sort_boxesr_crop_by_polysr ValueErrorrCtext_det_modeltext_rec_score_threshr@text_rec_modelr batch_samplerr img_reader) selfrr r!r"r#doc_preprocessor_configtextline_orientation_configtext_det_configtext_rec_config __class__s h/var/www/html/land-doc-ocr/venv/lib/python3.13/site-packages/paddlex/inference/pipelines/ocr/pipeline.pyrE_OCRPipeline.__init__+s*   %+JJ/Et$L!  $ $&,jj&D&H&H!+-Z' # .2-A-A'.D *)/ 3Mt(T%  ( (*0**\2*F*J*J%%'UV+ '/3.?.?+/D +!**\26:: 24VW  , >>Y &+:+>+>?OQT+UD ('6':':<'OD $+:+>+>?OQU+VD (#2#6#6x#ED '6':':<'MD $.22=$GD )8)<)<^S)QD &,D "-6"BD  ^^v %+:+>+>?OQT+UD ('6':':<'OD $+:+>+>?OQU+VD (#2#6#6x#ED '6':':<'MD $)8)<)<^S)QD &.22=$GD ,D "-6"BD 7>>t~~NO O"// 77//77''//33((0  !**\26::  !#E F &5%8%8%K".223DeL*..}dC"// )9)90 /&**\ST:UV#51image_array_listrotate_angle_listc[U5[U5:Xd S[U5S[U5S35eUHnUS;aM SU35e /n[X5H&upVUS-n[XW5nURU5 M( U$)aX Rotate the given image arrays by their corresponding angles. 0 corresponds to 0 degrees, 1 corresponds to 180 degrees. Args: image_array_list (List[np.ndarray]): A list of input image arrays to be rotated. rotate_angle_list (List[int]): A list of rotation indicators (0 or 1). 0 means rotate by 0 degrees 1 means rotate by 180 degrees Returns: List[np.ndarray]: A list of rotated image arrays. Raises: AssertionError: If any rotate_angle is not 0 or 1. AssertionError: If the lengths of input lists don't match. zLength of image_array_list (z*) must match length of rotate_angle_list ())rrz&rotate_angle must be 0 or 1, now it's )lenziprappend) rYrbrcanglerotated_images image_arrayrotate_indicator rotate_angle rotated_images r_r_OCRPipeline.rotate_images(#$ )   E )#.>*?)@@jknpAlBkCCD E E 'EF? T&LUG$T T?'-01A-U )K+c1L(CM  ! !- 0 .V ramodel_settingscUS(a(UR(d[R"S5 gUS(a(UR(d[R"S5 gg)a Check if the input parameters are valid based on the initialized models. Args: model_info_params(Dict): A dictionary containing input parameters. Returns: bool: True if all required models are initialized according to input parameters, False otherwise. r&zRSet use_doc_preprocessor, but the models for doc preprocessor are not initialized.Fr*z^Set use_textline_orientation, but the models for use_textline_orientation are not initialized.T)r&r errorr*)rYrqs r_check_model_settings_valid'_OCRPipeline.check_model_settings_validsT 0 1$:S:S MMd  5 611 MMp rause_doc_orientation_classifyuse_doc_unwarpingr*cxUcUc URnOUSLdUSLaSnOSnUc URn[UUS9$)a Get the model settings based on the provided parameters or default values. Args: use_doc_orientation_classify (Optional[bool]): Whether to use document orientation classification. use_doc_unwarping (Optional[bool]): Whether to use document unwarping. use_textline_orientation (Optional[bool]): Whether to use textline orientation. Returns: dict: A dictionary containing the model settings. TF)r&r*)r&r*dict)rYrvrwr*r&s r_get_model_settings_OCRPipeline.get_model_settingss]" ( /4E4M#'#<#< +t37HD7P'+$',$ # +'+'D'D $!5%=  rarKrLrMrNrOrPc Uc URnUc URnUc URnUc URnUc URnUc UR n[ UUUUUUS9$)a Get text detection parameters. If a parameter is None, its default value from the instance will be used. Args: text_det_limit_side_len (Optional[int]): The maximum side length of the text box. text_det_limit_type (Optional[str]): The type of limit to apply to the text box. text_det_max_side_limit (Optional[int]): The maximum side length of the text box. text_det_thresh (Optional[float]): The threshold for text detection. text_det_box_thresh (Optional[float]): The threshold for the bounding box. text_det_unclip_ratio (Optional[float]): The ratio for unclipping the text box. Returns: dict: A dictionary containing the text detection parameters. )r1r2r5r4r6r8)rKrLrMrNrOrPry)rYrKrLrMrNrOrPs r_get_text_det_params _OCRPipeline.get_text_det_paramss2 # *&*&B&B #  &"&":":  " *&*&B&B #  ""22O  &"&":":  ($($>$> !2*"2*.   rainputrUr@c # # URX#U5n URU 5(dSS0v URUUUUU U 5nU c URn U c URn [ UR U55GHunnURUR5nU S(a[URUUUS95nOUVs/sHnSU0PM nnUVs/sHnUSPM nn[UR"U40UD65nUVs/sHnUSPM nnUVs/sHnURU5PM nn[URURUU5VVVVs/sH"unnnnUUUUU UUR U U ////S. PM$ nnnnn[[#[%U555nUVs/sHn[%UU5S :dMUPM nnU(Ga/nS /n UHUn[UR'UUUU55n!UR)U!5 U R+U S [%U!5-5 MW U S (aCUR-U5V"s/sHn"[/U"S S 5PM n#n"UR1UU#5nOS /[%U5-n#[ U5Hun$nUUn%U#U U$U U$S -U%S'M [ U5GHun$nUU U$U U$S -n!UUn%UUn[ U!5V&V's/sH1un&n'U&U'R2S [5U'R2S 5- S.PM3 n(n&n'[7U(SS9n)U)V*s/sH n*U!U*SPM n+n*[ UR9U+U S95Hun$n,U)U$Sn-U,U(U-S'M U (a /U%S'/U%S'[#[%U(55Hn.U(U.Sn,U,SU :dMU (ab[;U,SS UU.U,SS 5un/n0U%SR+U/5 U%SR+U05 U%SR+U,SS 5 OU%SR+U,S5 U%SR+U,S5 U%SR+U,S5 U%SR+UU.5 M GM UHvn%UR S:Xa=[=U%S5n1U1U%S'U (a"U%SV2s/sHn2[=U25PM sn2U%S 'O[>R@"/5U%S'[CU%5v Mx GM gs snfs snfs snfs snfs snnnnfs snfs sn"fs sn'n&fs sn*fs sn2f7f)!a Predict OCR results based on input images or arrays with optional preprocessing steps. Args: input (Union[str, list[str], np.ndarray, list[np.ndarray]]): Input image of pdf path(s) or numpy array(s). use_doc_orientation_classify (Optional[bool]): Whether to use document orientation classification. use_doc_unwarping (Optional[bool]): Whether to use document unwarping. use_textline_orientation (Optional[bool]): Whether to use textline orientation prediction. text_det_limit_side_len (Optional[int]): Maximum side length for text detection. text_det_limit_type (Optional[str]): Type of limit to apply for text detection. text_det_max_side_limit (Optional[int]): Maximum side length for text detection. text_det_thresh (Optional[float]): Threshold for text detection. text_det_box_thresh (Optional[float]): Threshold for text detection boxes. text_det_unclip_ratio (Optional[float]): Ratio for unclipping text detection boxes. text_rec_score_thresh (Optional[float]): Score threshold for text recognition. return_word_box (Optional[bool]): Whether to return word boxes along with recognized texts. Returns: OCRResult: Generator yielding OCR results for each input image. rsz0the input params for model settings are invalid!Nr&)rvrw output_imgdt_polys) input_path page_indexdoc_preprocessor_resrrqtext_det_paramsr/rUr@ rec_texts rec_scores rec_polys vis_fontsrr* class_idsrtextline_orientation_angles) sub_img_id sub_img_ratioc US$)Nr)xs r_&_OCRPipeline.predict..s ?9Kra)keyr)r@rec_res text_wordtext_word_region rec_scorerec_textrrrvis_fontrr0 rec_boxestext_word_boxes)"rzrtr}rUr@ enumeraterWrX instanceslistrHrTrQrh input_paths page_indexesr/rangergrRextendrirJintrshapefloatsortedrVrrnparrayr)3rYrrvrwr*rKrLrMrNrOrPrUr@rqr_ batch_data image_arraysdoc_preprocessor_resultsarritemdoc_preprocessor_images det_results dt_polys_listrrrrresultsindicesidxall_subs_of_imgs chunk_indicesall_subs_of_imgtextline_angle_infoanglesiresimg_idsub_imgsub_img_info_listsorted_subs_inforsorted_subs_of_imgrrsnoword_box_content_list word_box_listrlines3 r_predict_OCRPipeline.predictsmF00 (=U ..~>>NO O22 #  #   !   ! ($($>$> !  ""22O&t'9'9%'@AMAz??:+?+?@L45+/22$5Q*;3,(LX+X?G&-Mgs]35G1H11LsgGM#% !" "C&*++3C8-:L'O %++O<!((r):S=Q)QR#""<=483R3R,44//  >} M #K 0 7 7 8KA8N O #K 0 7 7 8K L -44W[5IJ ,33GJ4GH ,33HSMB! =;1\>>Y. 7K8H II'0C $&),,>(?2(?4D9(?2-. (*xx|C $n$CB,Y'GN2N ,)*F2sCU TUT!-"UT&U$T+?)U()T0 &U7T8T8BUT=9BU8U 8U UA,U DUU %A-U)rRrQrWrHrXr7r@rOrKrLrMrTrNrPrVrUr/rJr&r*)NNFN)NNNNNN) NNNNNNNNNNN)__name__ __module__ __qualname____firstlineno____doc__rrstrrboolrrrrErrndarrayrrrtryrzrr}rr__static_attributes__ __classcell__)r^s@r_rr's~ !%59AE `2`2 `212 `2  `2 U4S>9#<=> `2 `2`2D" $RZZ 0"EI#Y" bjj "H$8 &.tn $D> #+4.    D26-115+//315, !)#, &c], "*# , "% , &e_ ,  (,  , b8<,03715-115+//31515*.E%S$s)RZZbjj1AABE%'/tnE%$D> E% #+4. E% "*# E%&c]E%"*#E%"%E%&e_E% (E% (E%"$E% E%E%rarocrzocr-core)altc.\rSrSrSr\S5rSrSrg) OCRPipelineiOCRc[$)N)r)rYs r_ _pipeline_clsOCRPipeline._pipeline_clssrac&URSS5$)NrAr)rF)rYrs r__get_batch_sizeOCRPipeline._get_batch_sizeszz,**rarN) rrrrentitiespropertyrrrrrar_rrsH +rar)&typingrrrrrnumpyrutilsr utils.depsr common.batch_samplerr common.readerr utils.benchmarkr utils.hpirutils.pp_optionr _parallelrbaser componentsrrrrrrresultr time_methodsrrrrar_rs4325&("4@ x%<x%x%v J/+:+0+ra