Intro
Barcodes are being used in many documents or forms to enable machine reading capabilities and reduce manual processing effort. Simple 1D barcodes for example can mark a specific page on a form, or indicate a relevant document identification number. More complex 2D barcodes allow to encode even the full data of the document in a more structured form.
Docbarcodes extracts 1D and 2D barcodes from scanned PDF documents or images. It can be used to automate extraction and processing of all kind of documents.
Some working documents with barcodes are:
- Swiss tax statements (Zurich, other cantons can be added as well)
- Swiss salary statement
- US drivers licenses with MRZ (machine readable zone)
- Swiss QR Code Invoices, introduced by Six-group
- Swiss Covid Certificates
The approach works as follows:
- Detect barcode regions on the document using opencv and image transformation heuristics
- Extract the raw barcode data using zxing
- Combine multiple barcodes and decode the data.
Quick start
Required:
- Java 8
Install package
pip install docbarcodes
Download example pdf document
wget https://github.com/ArlindNocaj/document-barcodes/raw/master/data/salary_swissdec/SalarySinglePage.pdf
CLI usage
Extract barcodes from salary statement document and format result json using jq
.
docbarcodes ./SalarySinglePage.pdf | jq .
{
"BarcodesRaw": [
{
"page": 0,
"num_candidate": 2,
"raw": "´cý¸z\u0000\u0002V\u0001\u0003\u0000\u0001\u0000\u0001PK\u0003\u0004\u0014\u0000\b\b\b\u0000év\u0003Q\u0000\u0000\u0000\u0000\u0000\u0000\u0000\u0000\u0000\u0000\u0000\u0000\u0004\u0000\u0000\u0000txabM�Ýnâ0\u0010�_Åò}b;¡!¬\u001cWá¯EbYÔÐjÕ;\u0013Ü\u00105±+Û,ð>û&ûb;\tÐö\"Îñ�\u0019Ïç�ùý©mÐ\u001fe]mt�YH1Rº4»ZW\u0019~ÞÌ�\u0014ß\u000b¾A�¦]�÷Þ\u007fü äx<�îX;·SeXî�+÷ª�ÄíHD#J£��B6Ò�§ª��ôpúæ��1*\u0016Ó\fS\nm�³{é$Ã�OLû!õ\u0019=/¦Áx^dxò8\u000bX\u0014�×\u000f£5 v�kyh$Z)ç\u001be1z|\n�&ÁJ¶\nj~ý\\/g¿Ñ¬QïÞ\u001a]¿£ü\u0001£×Å:ÃqÒµ�,3�o½ª��®PáÕAYPJ\u0003�·Jù\fÏêJÙ£ªP\u0002=�ó[s�±D\u0003(®ý9Ã�ý\u0001²×{£¡#�c\u0014Å)\u001a�Ð�a\"ø�r1EKé¼î©rµ1�×öÓÐZÞ¨è�ê\u000b`ùﯮ*é�ó\u0016V�¢ï$É\b\n�¦�MOs£\u001a+«Ñ\bvæ ½í\u0002�\u0017/A^\u0004++º\u0019\u000eî�a��(\u000b£��Ï\u0010'7bÁsÁ§¦\u0004Å8¹�î:µÙ\tþfM+\"Ê��Ñ�B¼78ôª�«Ï��rrqúSûÂ\u0005<¥V�AÊX\u001a\u000e!ájð\u0007k�»n�\u0014Æ\u0010R�~wùJù«¼KâAÜ'|y�\u0000/Ù�ÿPK\u0007\bK¦nJÎ\u0001\u0000\u0000Á\u0002\u0000\u0000PK\u0001\u0002\u0014\u0000\u0014\u0000\b\b\b\u0000év\u0003QK¦nJÎ\u0001\u0000\u0000Á\u0002\u0000\u0000\u0004\u0000\u0000\u0000\u0000\u0000\u0000\u0000\u0000\u0000\u0000\u0000\u0000\u0000\u0000\u0000\u0000\u0000txabPK\u0005\u0006\u0000\u0000\u0000\u0000\u0001\u0000\u0001\u00002\u0000\u0000\u0000\u0000\u0002\u0000\u0000\u0000\u0000",
"format": "PDF_417",
"points": [
[
0.08509771986970684,
0.2564102564102564
],
[
0.32166123778501626,
0.2564102564102564
],
[
0.32166123778501626,
0.35522904062229904
],
[
0.08509771986970684,
0.35522904062229904
]
],
"resultMetadata": {
"ERROR_CORRECTION_LEVEL": "2",
"PDF417_EXTRA_METADATA": {
"addressee": "None",
"checksum": -1,
"fileId": "None",
"fileName": "None",
"fileSize": -1,
"optionalData": "None",
"segmentCount": -1,
"segmentIndex": 0,
"sender": "None",
"timestamp": -1
}
}
}
],
"BarcodesCombined": [
{
"content": "\"1.0\" encoding=\"UTF-8\"?>
\"http://www.swissdec.ch/schema/sd/20200220/SalaryDeclarationTxAB
\" SID=
\"000
\" SysV=
\"001
\">
\"CHE-123.123.123
\" Person=
\"Paula Nestler
\" HR-RC-Name=
\"COMPLEX Elektronik AG
\" ZIP=
\"3600
\" CL=
\"Abteilung Steuerungen
\" Street=
\"Eigerweg 6
\" Postbox=
\"124
\" City=
\"Thun
\" Phone=
\"033 238 49 71
\"/>
\"Aebi
\" Firstname=
\"Anna
\" ZIP=
\"3000
\" CL=
\"\" Street=
\"Länggassstrasse 26
\" Postbox=
\"690
\" Locality=
\"\" City=
\"Bern 9
\" Country=
\"\">
123.4567.8901.28
1
2016-10-01
2016-11-30
48118.70
68000.00
56343.00
",
"format": "PDF_417",
"sources": [
0
]
}
]
}
Code Usage
from docbarcodes.extract import process_document
barcodes_raw, barcodes_combined = process_document("./SalarySinglePage.pdf")
print(barcodes_raw)
print(barcodes_combined)
FAQ
On Windows only: If you have problems with the installation of package dependencies, I recommend using conda
to install java and poppler
conda install -y -c conda-forge jpype1=1.3.0
conda install -c conda-forge poppler=21
Show package licenses
pip-licenses --with-urls --with-system --format=markdown
Improvements to be made:
- implement multithreading class for zxing in java which returns proper objects for python consumption
- extension mechanisms for other 2D barcode aggregations