i%dZddlZddlZddlZddlZddlZddlZddlZddlZddl Z ddl m Z ddl Z ddl mZddlZddlmZmZddlZddlmZmZdZd Zd Zd Zy) z PassportEye::MRZ: Machine-readable zone extraction and parsing. Command-line scripts Author: Konstantin Tretyakov License: MIT N)Counter)io)TesseractNotFoundErrorTesseractError)read_mrz MRZPipelinectj}|\}}}t|||}tj|z }|||fS)z Processes a file and returns the parsed MRZ (or None if no candidate regions were even found). The input argument is a list (filename, save_roi, extra_params). (Because we need to use this function within imap_unordered) )save_roiextra_cmdline_params)timer)paramsticfilenamer extra_paramsresultwalltimes |/var/www/ITT-ServiceAPI-Services/ITT-Document-Scan/mrzscanner/mrzenv/lib/python3.12/site-packages/passporteye/mrz/scripts.py process_filersF ))+C'-$Hh h UFyy{S H fh ''c 8 tjd}|jdddtd|jdd t j d d d |jdddd |jdddd |jdddd |jdddtd|jddd|j }ttjtjj|jd }|jd!k\r|d!|j}tjtj "tj"d#}t%j$}t'j(|j*}|j-d$t/||j|j-d%|j*g}|j0du}|j2|j4|j0fD]:}|tjj7|r&tj8|<d&fd'} t;} |j<rd(nd)} |j?t@|D cgc]} | || f c} D]} | \}}}|jC| |j-d*tjjE|||| |||jGd+||}|j2e|d!kDr`tIjJ|tjj|j2d,|tjjE|fz|j4e|d!k(r`tIjJ|tjj|j4d,|tjjE|fz|j0|}d-|jLvrod.|tjjE|fz}tOjPtjj|j0||jLd-|d!kDsd/|jLvs| |jLd/xxdz cc<t/|}|Dcgc]\}}}| ||}}}}|Dcgc]\}}}|}}}}|jSd0}|jSd!}tU|}tU|Dcgc]\}}}| c}}}}t%j$|z }|j-d1tWd2|ztWd3|ztWd4|ztWd5|ztWd6|ztWd7t/|Dcgc] }|d8k(s | c}ztWd9t/|Dcgc] }|d:k(s | c}ztWd;|ztWd| j[D]}tWd?|zycc} wcc}}}wcc}}}wcc}}}wcc}wcc}w)@z A script for evaluating the current MRZ recognition pipeline by applying it to a list of files in a directory and reporting how well it went. z]Run the MRZ OCR recognition algorithm on the sample test data, reporting the quality summary. descriptionz-jz--jobsrzNumber of parallel jobs to run)defaulttypehelpz-ddz --data-dirzpassporteye.mrztestdataz@Read files from this directory instead of the package test filesrrz-sdz --success-dirNzOCopy files with successful (nonzero score) extraction results to this directoryz-fdz --fail-dirzNCopy files with unsuccessful (zero score) extraction resutls to this directoryz-rdz --roi-dirzExtract ROIs to this directoryz-lz--limitz6Only process the first files in the directory.--legacy store_truekUse the "legacy" Tesseract OCR engine (--oem 0). Despite the name, it most often results in better results. It is not the default option, because it will only work if your Tesseract installation includes the legacy *.traineddata files. You can download them at https://github.com/tesseract-ocr/tesseract/wiki/Data-Files#data-files-for-version-400-november-29-2016actionrz*.*r)level evaluate_mrzz*Preparing computation for %d files from %szRunning %d workersc"|dS|jS)Nr) valid_score)mrz_s rr(z!evaluate_mrz..valid_scoreNsLq6d&6&66rc |}ttjj|j dd}||z }|dk(rdS|dkDrdSdS#t $rYywxYw)N_r=><?)intospathbasenamesplit Exception)rr) new_score old_scoreschanger(s rscore_change_typez'evaluate_mrz..score_change_typeQss #D)IBGG,,X6< ? @ djj1%&) " @ A   rzz12 s8F??=I?IA IIctjd}|jd|jdddd|jd d d d |jdddtjz|j }t |jd}|jr!tj|jdt|dD]f\}}tjtjj!|jd|z|dzj#t$j&hy)zc Command-line script for extracting all potential MRZ regions from a given image as images zXExtract all potential MRZ regions from a given image and output as individual PNG files.rrz-dz --output-dir.zThe directory where the images for the potential MRZ regions will be output (default is the current directory). The files created will be named 1.png, 2.png, ...rz-cz --create-dirr!zKForce the creation of the directory (and all parents) if it does not exist.r#r~rzPassprtEye extract_mrz_rois v%srroisT)exist_okrz%d.pngrN)r@rArBrrrEr r create_dirr1makedirs output_dir enumeraterr`r2rHrrr)rfrgrnimgs rextract_mrz_roisrs $ $2LMF  # ncLM n\IVW  I?`cnczcz?z{    D t}} %f -D  DOOd3D!$]3 "'',,t1 =c ?Q?QRTRZRZ?[\]r)__doc__r@rGrrKrOr1r]rr collectionsrrCskimagernumpyrpytesseract.pytesseractrrrimagerr rr&rrrrrsT  J( (\!~('V]r