B 5dy@sfddlZddlZddlZejeZddlZejej edddl Z ddl Z ddl Z ddlmZddlZddlmZddlmZeddZeddZedd Zdd lmZdd lmZeZdd lmZm Z dd l!m"Z"m#Z#m$Z$m%Z%ddl&m'Z'm(Z(m)Z)ddl*m+Z+m,Z,ddl-m.Z.m/Z/m0Z0dddddddgZ1dgZ2dZ3ddgZ4ej5dZ6dZ7dddgZ8d Z9d!d gZ:d"d#id"d$id"d%id&d'd(d)d*d+d)d,d-d)d.d/d)d0d1d)d2d3d)d4d5d)d6d7d)d8d9d)d:d;d)dd?d)d@ dAd"dBiidCdAd"dDiidAdEd(d)idAd"dBiidCd"dFid"dGid"dHidIdJd(d)dKd+d)dLdMd)dNdOd)dPd-d)dQd/d)dRd1d)dSd3d)dTd5d)dUd7d)dVd9d)dWd;d)dXd=d)dYd?d)dZd[d)d\dAd"dBiidCd]d^d_d`dad)iidbdad)dcddd)dedfdgd)dhdid)dedjdkdlZ;d}dndoZe?dtdudvZ@dwdxZAGdyddejBZCGdzdde.ZDd{d|ZEdS)~N)Path)BytesIO)Image.toolsppocr ppstructure)predict_system) get_logger)check_and_readget_image_file_list)maybe_downloaddownload_with_progressbaris_linkconfirm_model_dir_url)draw_ocrstr2bool check_gpu) init_argsdraw_structure_result)StructureSystemsave_structure_resto_excel PaddleOCR PPStructurerrrrrZDBz2.6.1.3ZCRNNZ SVTR_LCNetz ~/.paddleocr/zPP-OCRv3zPP-OCRzPP-OCRv2zPP-StructureV2z PP-StructureurlzJhttps://paddleocr.bj.bcebos.com/PP-OCRv3/chinese/ch_PP-OCRv3_det_infer.tarzJhttps://paddleocr.bj.bcebos.com/PP-OCRv3/english/en_PP-OCRv3_det_infer.tarzYhttps://paddleocr.bj.bcebos.com/PP-OCRv3/multilingual/Multilingual_PP-OCRv3_det_infer.tar)chenmlzJhttps://paddleocr.bj.bcebos.com/PP-OCRv3/chinese/ch_PP-OCRv3_rec_infer.tarz./ppocr/utils/ppocr_keys_v1.txt)r dict_pathzJhttps://paddleocr.bj.bcebos.com/PP-OCRv3/english/en_PP-OCRv3_rec_infer.tarz./ppocr/utils/en_dict.txtzShttps://paddleocr.bj.bcebos.com/PP-OCRv3/multilingual/korean_PP-OCRv3_rec_infer.tarz"./ppocr/utils/dict/korean_dict.txtzRhttps://paddleocr.bj.bcebos.com/PP-OCRv3/multilingual/japan_PP-OCRv3_rec_infer.tarz!./ppocr/utils/dict/japan_dict.txtzXhttps://paddleocr.bj.bcebos.com/PP-OCRv3/multilingual/chinese_cht_PP-OCRv3_rec_infer.tarz'./ppocr/utils/dict/chinese_cht_dict.txtzOhttps://paddleocr.bj.bcebos.com/PP-OCRv3/multilingual/ta_PP-OCRv3_rec_infer.tarz./ppocr/utils/dict/ta_dict.txtzOhttps://paddleocr.bj.bcebos.com/PP-OCRv3/multilingual/te_PP-OCRv3_rec_infer.tarz./ppocr/utils/dict/te_dict.txtzOhttps://paddleocr.bj.bcebos.com/PP-OCRv3/multilingual/ka_PP-OCRv3_rec_infer.tarz./ppocr/utils/dict/ka_dict.txtzRhttps://paddleocr.bj.bcebos.com/PP-OCRv3/multilingual/latin_PP-OCRv3_rec_infer.tarz!./ppocr/utils/dict/latin_dict.txtzShttps://paddleocr.bj.bcebos.com/PP-OCRv3/multilingual/arabic_PP-OCRv3_rec_infer.tarz"./ppocr/utils/dict/arabic_dict.txtzUhttps://paddleocr.bj.bcebos.com/PP-OCRv3/multilingual/cyrillic_PP-OCRv3_rec_infer.tarz$./ppocr/utils/dict/cyrillic_dict.txtzWhttps://paddleocr.bj.bcebos.com/PP-OCRv3/multilingual/devanagari_PP-OCRv3_rec_infer.tarz&./ppocr/utils/dict/devanagari_dict.txt) rrkoreanjapan chinese_chttatekalatinarabiccyrillic devanagarirzRhttps://paddleocr.bj.bcebos.com/dygraph_v2.0/ch/ch_ppocr_mobile_v2.0_cls_infer.tar)detrecclszJhttps://paddleocr.bj.bcebos.com/PP-OCRv2/chinese/ch_PP-OCRv2_det_infer.tarzJhttps://paddleocr.bj.bcebos.com/PP-OCRv2/chinese/ch_PP-OCRv2_rec_infer.tarzRhttps://paddleocr.bj.bcebos.com/dygraph_v2.0/ch/ch_ppocr_mobile_v2.0_det_infer.tarz\https://paddleocr.bj.bcebos.com/dygraph_v2.0/multilingual/en_ppocr_mobile_v2.0_det_infer.tarz[https://paddleocr.bj.bcebos.com/dygraph_v2.0/table/en_ppocr_mobile_v2.0_table_det_infer.tar)rr structurezRhttps://paddleocr.bj.bcebos.com/dygraph_v2.0/ch/ch_ppocr_mobile_v2.0_rec_infer.tarz]https://paddleocr.bj.bcebos.com/dygraph_v2.0/multilingual/en_number_mobile_v2.0_rec_infer.tarzZhttps://paddleocr.bj.bcebos.com/dygraph_v2.0/multilingual/french_mobile_v2.0_rec_infer.tarz"./ppocr/utils/dict/french_dict.txtzZhttps://paddleocr.bj.bcebos.com/dygraph_v2.0/multilingual/german_mobile_v2.0_rec_infer.tarz"./ppocr/utils/dict/german_dict.txtzZhttps://paddleocr.bj.bcebos.com/dygraph_v2.0/multilingual/korean_mobile_v2.0_rec_infer.tarzYhttps://paddleocr.bj.bcebos.com/dygraph_v2.0/multilingual/japan_mobile_v2.0_rec_infer.tarz_https://paddleocr.bj.bcebos.com/dygraph_v2.0/multilingual/chinese_cht_mobile_v2.0_rec_infer.tarzVhttps://paddleocr.bj.bcebos.com/dygraph_v2.0/multilingual/ta_mobile_v2.0_rec_infer.tarzVhttps://paddleocr.bj.bcebos.com/dygraph_v2.0/multilingual/te_mobile_v2.0_rec_infer.tarzVhttps://paddleocr.bj.bcebos.com/dygraph_v2.0/multilingual/ka_mobile_v2.0_rec_infer.tarz_https://paddleocr.bj.bcebos.com/dygraph_v2.0/multilingual/latin_ppocr_mobile_v2.0_rec_infer.tarz`https://paddleocr.bj.bcebos.com/dygraph_v2.0/multilingual/arabic_ppocr_mobile_v2.0_rec_infer.tarzbhttps://paddleocr.bj.bcebos.com/dygraph_v2.0/multilingual/cyrillic_ppocr_mobile_v2.0_rec_infer.tarzdhttps://paddleocr.bj.bcebos.com/dygraph_v2.0/multilingual/devanagari_ppocr_mobile_v2.0_rec_infer.tarz[https://paddleocr.bj.bcebos.com/dygraph_v2.0/table/en_ppocr_mobile_v2.0_table_rec_infer.tarzppocr/utils/dict/table_dict.txt)rrfrenchgermanr!r"r#r$r%r&r'r(r)r*r.)zPP-OCRv3zPP-OCRv2zPP-OCRtablerzahttps://paddleocr.bj.bcebos.com/dygraph_v2.0/table/en_ppocr_mobile_v2.0_table_structure_infer.tarz)ppocr/utils/dict/table_structure_dict.txtzehttps://paddleocr.bj.bcebos.com/ppstructure/models/slanet/en_ppstructure_mobile_v2.0_SLANet_infer.tarzehttps://paddleocr.bj.bcebos.com/ppstructure/models/slanet/ch_ppstructure_mobile_v2.0_SLANet_infer.tarz,ppocr/utils/dict/table_structure_dict_ch.txt)rrzahttps://paddleocr.bj.bcebos.com/ppstructure/models/layout/picodet_lcnet_x1_0_fgd_layout_infer.tarz6ppocr/utils/dict/layout_dict/layout_publaynet_dict.txtzfhttps://paddleocr.bj.bcebos.com/ppstructure/models/layout/picodet_lcnet_x1_0_fgd_layout_cdla_infer.tarz1ppocr/utils/dict/layout_dict/layout_cdla_dict.txt)r1layout)z PP-StructurezPP-StructureV2)OCR STRUCTURETcCsddl}t}||_|jdtdd|jdtdd|jdtdd|jdtd d|jd ttd d d |jdttddd x|jD]}|j dkrd|_ qW|r| Si}x|jD]}|j ||j <qW|j f|SdS)Nrz--langr)typedefaultz--detTz--recz--typeocrz --ocr_versionzPP-OCRv3aROCR Model version, the current model support list is as follows: 1. PP-OCRv3 Support Chinese and English detection and recognition model, and direction classifier model2. PP-OCRv2 Support Chinese detection and recognition model. 3. PP-OCR support Chinese detection, recognition and direction classifier and multilingual recognition model.)r5choicesr6helpz--structure_versionzPP-StructureV2zModel version, the current model support list is as follows: 1. PP-Structure Support en table structure model. 2. PP-StructureV2 Support ch and en table structure model.)rec_char_dict_pathtable_char_dict_pathlayout_dict_path) argparseradd_help add_argumentstrrSUPPORT_OCR_MODEL_VERSIONSUPPORT_STRUCTURE_MODEL_VERSION_actionsdestr6 parse_args Namespace)mMainr=parseractionZinference_args_dictrJ7/tmp/pip-unpacked-wheel-ndi_cy3p/paddleocr/paddleocr.pyrE6s8    rEc*CsNddddddddd d d d d ddddddddddddddddddd d!d"d#d$d%d&d'd(d)d*g*}d+d,d-d.g}d/d0d1d2d3d4d5d6d7d8d9d:d;dg}d?d@dAdBdCdDdEdFdGdHdIdJdKg }||krdL}n(||krdM}n||krdN}n ||krdO}|tdPtdQkstdRtdPtdQ||dSkr"dS}n$|dTkr2dT}n|dUkrBdV}ndW}||fS)XNafazbscscydadeesetfrgahrhuidisitZkuZlaltlvmimsmtnlnoocpiplptroZrs_latinskslsqsvswtltruzvir/r0arfaZugurruZ rs_cyrillicbebgukmnZabqZadyZkbdZavaZdarZinhZcheZlbeZleztabhimrneZbhmaiangZbhoZmahZscknewZgomsaZbgcr'r(r)r*r3r,z!param lang must in {}, but got {}rr.)rr'rr) MODEL_URLSDEFAULT_OCR_MODEL_VERSIONAssertionErrorformatkeys)langZ latin_langZ arabic_langZ cyrillic_langZdevanagari_langdet_langrJrJrK parse_lang_s8     rcCs|dkrt}n|dkrt}ntt|}||kr4|}|||krv|||krR|}n$td|||t d||||kr||||kr|}n*td|||||t d||||S)Nr3r4z,{} models is not support, we only support {}z8lang {} is not support, we only support {} for {} models) rDEFAULT_STRUCTURE_MODEL_VERSIONNotImplementedErrorrloggererrorrrsysexit)r5versionZ model_typerZDEFAULT_MODEL_VERSIONZ model_urlsrJrJrKget_model_configs.    r)contentcCstj|tjd}t|tjS)N)Zdtype)np frombufferuint8cv2imdecode IMREAD_COLOR)rZnp_arrrJrJrK img_decodesrc Csht|trt|}t|tr8t|r4t|dd}|}t|\}}}|s|st|d}|}t|}WdQRX|dkry|t }t |}t |}| d} | |d| d|} tt| dd} t| } t| tj} t| tj}Wntd|dS|dkr8td|dSt|tjrdt|jd krdt|tj}|S) Nztmp.jpgrbZRGBZjpegrzutf-8)encodingzerror in loading image:{}) isinstancebytesrr@rrr openreadrrconvertsaveseekbase64 b64encode b64decoderrrrrrrrrndarraylenshapeZcvtColorZCOLOR_GRAY2BGR)imgZ image_fileflag_gifflag_pdffZimg_strbufimageZimZrgbZ image_bytesZ data_base64Z image_decodeZ img_arrayrJrJrK check_imgsF             rcs&eZdZfddZdddZZS)rc  stdd}|jjf||jtks4tdt|jt|j|_|j sRt t j |j|_t|j\}}td|jd|}t|jtjtdd||d\|_}td|jd|}t|jtjtdd||d\|_}td|jd d } t|jtjtdd | d\|_} |jd krd |_nd |_|jsNt|j|t|j|t|j| |jtkrtt dtt !d|j"t#krt dt#t !d|j$dkrt%t&t'j(|d|_$t )|t*+||j,|_,dS)zm paddleocr package args: **kwargs: other params show in paddleocr --help F)rGz"ocr_version must in {}, but get {}r3r+whlrr,r-rzPP-OCRv3z 3, 48, 320z 3, 32, 320zdet_algorithm must in {}rzrec_algorithm must in {}Nr )-rE__dict__update ocr_versionrArrruse_gpushow_logrsetLevelloggingINFO use_angle_clsrrrr det_model_dirospathjoinBASE_DIR rec_model_dirZ cls_model_dirZrec_image_shapeZuse_onnxrZ det_algorithmSUPPORT_DET_MODELrrrZ rec_algorithmSUPPORT_REC_MODELr:r@r__file__parentdebugsuper__init__page_num) selfkwargsparamsrrdet_model_configdet_urlrec_model_configrec_urlZcls_model_configZcls_url) __class__rJrKrsX                   zPaddleOCR.__init__TcCst|tjtttfstt|tr<|dkr,sz!PaddleOCR.ocr..cSsg|] }|qSrJ)r)rrrJrJrKr4s)rrrlistr@rrrrrrwarningrrr enumerate__call__zipappendZ text_detectorZtext_classifierZtext_recognizer)rrr+r,r-ZimgsZocr_residxZdt_boxesZrec_res_Ztmp_resZelapseZcls_resZ cls_res_tmprJrJrKr7sT       z PaddleOCR.ocr)TTT)__name__ __module__ __qualname__rr7 __classcell__rJrJ)rrKrs s                    )&(wB