[ Tech stack ]
PaddleOCR
The open-source multilingual OCR toolkit, from raw text to structured tables.
PaddleOCR, built by Baidu on PaddlePaddle, chains text detection, orientation classification and recognition in its PP-OCR models, lightweight and accurate, covering more than 80 languages including French. PP-Structure adds layout analysis, table recognition exportable to Excel or HTML, and key information extraction. Installed with one command, usable from Python or the CLI, and deployable on a server, on mobile or through ONNX, it's our choice when text has to be extracted from scans and photos without relying on a cloud API.
[ Why PaddleOCR at Dexon ]
What this technology does well, and why we use it.
Typical usage: Invoice and receipt extraction, ID document reading, tables from scanned PDFs, industrial plates and labels.
- 01
PP-OCR: detection, orientation and recognition in lightweight models for 80+ languages.
- 02
PP-Structure: layout analysis, tables to Excel or HTML, key information extraction.
- 03
Fine-tuning on your business documents: fonts, receipts, plates, specialized scripts.
- 04
Python, CLI, server, mobile or ONNX; open source Apache 2.0, no cloud API.
[ Complementary technologies ]
The building blocks we often mobilise alongside.
A stack rarely exists alone. Here are the technologies Dexon most often pairs with this one, through pipeline habits, usage similarity or internal mastery. Click on a building block to see its scope.
[ Reassurance ]
- 0+
- custom projects delivered
- 30
- engineers, designers, project managers
- 80 %
- from top French schools
- 24 h
- average reply time
[ They trust us ]
More than 100 French and European companies trust us
[ Press ]
They talk about us.
Nationwide coverage by BFM Business, Le Figaro, Challenges, La Tribune and CNews. An outside reading of our work and our innovations.
