o 0jpU@sddlZddlmZmZmZmZmZmZddlZ ddl m Z ddl m Z ddlmZddlmZdd lmZmZdd lmZdd lmZd d lmZd dlmZd dlmZd dlm Z d dl!m"Z"ddl#m$Z$m%Z%ddl&m'Z'ddl m(Z(ej)GdddeZ*e dGdddeZ+dS)N)AnyDictListOptionalTupleUnion)logging)pipeline_requires_extra)ImageBatchSampler) ReadImage) HPIConfigPaddlePredictorOption) DetResult) benchmark)(AutoParallelImageSimpleInferencePipeline) BasePipeline) CropByBoxes)DocPreprocessorResult) OCRResult)SingleTableRecognitionResultTableRecognitionResult)get_table_recognition_res)get_neighbor_boxes_idxc!seZdZdZddddddddedeedeedeeeefd eed e d ee eeefe fd dffd dZ dee dee dee dee d e f ddZdededed e fddZdejde d eeejffddZddZd dZ  ! d1dejded"ed#e d$e d%e d efd&d'Z              d2d(e eeeejeejfdee dee dee dee deedeed)eed*eed+eed,eed-eed.eed#e d%ee d ef d/d0ZZ S)3_TableRecognitionPipelinezTable Recognition PipelineNFdeviceengine engine_config pp_optionuse_hpip hpi_configconfigrr r!r"r#r$returnc s(tjd||||||d||dd|_|jr,|didddi} || |_|dd|_|jrH|d id d d i} || |_|d id d di} || |_ |dd|_ |j rw|didddi} || |_ n |didd|_ t |_tdd|_tdd|_dS)aInitializes the table recognition pipeline. Args: config (Dict): Configuration dictionary containing various settings. device (Optional[str], optional): The device to use for prediction. Defaults to `None`. engine (Optional[str], optional): Inference engine. Defaults to `None`. engine_config (Optional[Dict[str, Any]], optional): Engine-specific config. Defaults to `None`. pp_option (Optional[PaddlePredictorOption], optional): Paddle predictor options. Defaults to `None`. use_hpip (bool, optional): Whether to use HPIP. Defaults to `False`. hpi_config (Optional[Union[Dict[str, Any], HPIConfig]], optional): HPIP configuration. Defaults to `None`. ruse_doc_preprocessorTZ SubPipelinesZDocPreprocessorZpipeline_config_errorz+config error for doc_preprocessor_pipeline!use_layout_detectionZ SubModulesZLayoutDetectionZmodel_config_errorz"config error for layout_det_model!ZTableStructureRecognitionz'config error for table_structure_model! use_ocr_modelZ GeneralOCRz&config error for general_ocr_pipeline!Nr)Z batch_sizeZBGR)format)super__init__getr'create_pipelinedoc_preprocessor_pipeliner(Z create_modellayout_det_modeltable_structure_modelr)general_ocr_pipelinegeneral_ocr_config_bakr_crop_by_boxesr batch_samplerr img_reader) selfr%rr r!r"r#r$kwargsZdoc_preprocessor_configZlayout_det_configZtable_structure_configZgeneral_ocr_config __class__r+w/var/www/html/Deteccion_Ine/venv/lib/python3.10/site-packages/paddlex/inference/pipelines/table_recognition/pipeline.pyr-)sZ         z"_TableRecognitionPipeline.__init__use_doc_orientation_classifyuse_doc_unwarpingr(r)cCs\|dur |dur |j}n |dus|durd}nd}|dur |j}|dur'|j}t|||dS)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_layout_detection (Optional[bool]): Whether to use layout detection. use_ocr_model (Optional[bool]): Whether to use OCR model. Returns: dict: A dictionary containing the model settings. NTF)r'r(r))r'r(r)dict)r8r=r>r(r)r'r+r+r<get_model_settingswsz,_TableRecognitionPipeline.get_model_settingsmodel_settingsoverall_ocr_reslayout_det_rescCs|dr|jstddS|dr'|durtddS|js'tddS|drB|dur6td dS|js@td dSd S|durMtd dSd S) a0 Check if the input parameters are valid based on the initialized models. Args: model_settings (Dict): A dictionary containing input parameters. overall_ocr_res (OCRResult): Overall OCR result obtained after running the OCR pipeline. The overall OCR result with convert_points_to_boxes information. layout_det_res (DetResult): The layout detection result. 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(Nz^The layout detection model has already been initialized, please set use_layout_detection=FalsezRSet use_layout_detection, but the models for layout detection are not initialized.r)zLThe OCR models have already been initialized, please set use_ocr_model=Falsez>Set use_ocr_model, but the models for OCR are not initialized.z7Set use_ocr_model=False, but no OCR results were found.T)r'r errorr(r))r8rArBrCr+r+r<check_model_settings_valids@ z4_TableRecognitionPipeline.check_model_settings_valid image_array input_paramscCsP|dr |d}|d}t|j|||dd}|d}||fSi}|}||fS)a Preprocess the document image based on input parameters. Args: image_array (np.ndarray): The input image array. input_params (dict): Dictionary containing preprocessing parameters. Returns: tuple[DocPreprocessorResult, np.ndarray]: A tuple containing the preprocessing result dictionary and the processed image array. r'r=r>r=r>r output_img)listr0)r8rFrGr=r>doc_preprocessor_resdoc_preprocessor_imager+r+r<predict_doc_preprocessor_ress"z6_TableRecognitionPipeline.predict_doc_preprocessor_resc Ct|ts |}g}tt|D]-}dd||D\}}}}t||||||ddfd} |d| dq|S)t Splits OCR bounding boxes by table cells and retrieves text. Args: ori_img (ndarray): The original image from which text regions will be extracted. cells_bboxes (list or ndarray): Detected cell bounding boxes to extract text from. Returns: list: A list containing the recognized texts from each cell. cSg|]}t|qSr+mathceil.0kr+r+r< M_TableRecognitionPipeline.split_ocr_bboxes_by_table_cells..Nr rec_texts isinstancerJtolistrangelenr3appendjoin r8Zori_imgZ cells_bboxesZ texts_listix1y1Zx2y2Zrec_ter+r+r<split_ocr_bboxes_by_table_cells (z9_TableRecognitionPipeline.split_ocr_bboxes_by_table_cellsc CrN)rOcSrPr+rQrTr+r+r<rW!rXrYNrrZr[r\rcr+r+r<rhriT table_box use_ocr_results_with_table_cellsflag_find_nei_textcell_sort_by_y_projectioncCst||d}|dkr|d}dd|D}|||} ng} t|||| ||d} d} |rKt|d|} t| dkrK| D] } | |d | d 7} q>| | d <| S) a+ Predict table recognition results from an image array, layout detection results, and OCR results. Args: image_array (np.ndarray): The input image represented as a numpy array. overall_ocr_res (OCRResult): Overall OCR result obtained after running the OCR pipeline. The overall OCR results containing text recognition information. table_box (list): The table box coordinates. use_ocr_results_with_table_cells (bool): whether to use OCR results with cells. flag_find_nei_text (bool): Whether to find neighboring text. cell_sort_by_y_projection (bool): Whether to sort the matched OCR boxes by y-projection. Returns: SingleTableRecognitionResult: single table recognition result. rTZbboxcSs(g|]}|d|d|d|dgqS)rrrr+)rUrectr+r+r<rWCszR_TableRecognitionPipeline.predict_single_table_recognition_res..rmrZZ rec_boxesr[z; Zneighbor_texts)rJr2rhrrr`)r8rFrBrjrkrlrmZtable_structure_predZtable_cells_resultZcells_texts_listZsingle_table_recognition_resZ neighbor_textZmatch_idx_listidxr+r+r<$predict_single_table_recognition_res)s8 z>_TableRecognitionPipeline.predict_single_table_recognition_resinputtext_det_limit_side_lentext_det_limit_typetext_det_threshtext_det_box_threshtext_det_unclip_ratiotext_rec_score_threshc ks|||||}|durd}||||sddiVt||D]\}}||jd}|dr?t|j|||dd}nd|i}|d}|d r\t|j||| | | | | d d}n|d krn|j dksgJ| |j |_g}d }|d s|duri}|j dd\}}dd|d |d g}|j ||||d|d}||d<| ||d 7}nD|d rt||d}|dD]2}|ddvr|||g}|d}|d}|j |d||||d}||d<| ||d 7}q|jd|jd|||||d}t|Vq"dS)az This function predicts the layout parsing result for the given input. Args: input (Union[str, list[str], np.ndarray, list[np.ndarray]]): The input image(s) of pdf(s) to be processed. use_layout_detection (bool): Whether to use layout detection. use_doc_orientation_classify (bool): Whether to use document orientation classification. use_doc_unwarping (bool): Whether to use document unwarping. overall_ocr_res (OCRResult): The overall OCR result with convert_points_to_boxes information. It will be used if it is not None and use_ocr_model is False. layout_det_res (DetResult): The layout detection result. It will be used if it is not None and use_layout_detection is False. use_ocr_results_with_table_cells (bool): whether to use OCR results with cells. cell_sort_by_y_projection (bool): Whether to sort the matched OCR boxes by y-projection. **kwargs: Additional keyword arguments. Returns: TableRecognitionResult: The predicted table recognition result. NFrDz0the input params for model settings are invalid!rr'rHrIr))rtrurvrwrxryTrr(r)rlrmtable_region_idZboxeslabel)tableboxZimgrp)Z input_pathZ page_indexrKrCrBtable_res_listrA)r@rE enumerater6r7Z instancesrJr0r3r4r/shaperrrar1lowerr5Z input_pathsZ page_indexesr) r8rsr=r>r(r)rBrCrtrurvrwrxryrkrmr9rAZimg_idZ batch_datarFrKrLr~rzZ img_heightZ img_widthrjZsingle_table_rec_resZbox_infoZ crop_img_infoZsingle_img_resr+r+r<predict^s'        z!_TableRecognitionPipeline.predict)FTF)NNNNNNNNNNNNFN)!__name__ __module__ __qualname____doc__rrstrrrboolrrr-r?r@rrrEnpZndarrayrrrMrhrJrrrrintfloatrr __classcell__r+r+r:r<r%s  N ' 7    8     rZocrc@s&eZdZdgZeddZddZdS)TableRecognitionPipelineZtable_recognitioncCstS)N)r)r8r+r+r< _pipeline_clssz&TableRecognitionPipeline._pipeline_clscCsdS)Nrr+)r8r%r+r+r<_get_batch_sizesz(TableRecognitionPipeline._get_batch_sizeN)rrrentitiespropertyrrr+r+r+r<rs   r),rRtypingrrrrrrnumpyrutilsr Z utils.depsr Zcommon.batch_samplerr Z common.readerr modelsrrZmodels.object_detection.resultrZutils.benchmarkrZ _parallelrbaser componentsrZdoc_preprocessor.resultrZ ocr.resultrresultrrZ!table_recognition_post_processingrrZ time_methodsrrr+r+r+r<s2              G