Time
1,5 months
Location
USA
Sector
Legal
Description
The client’s requirement was to establish a versatile framework capable of processing a wide range of legal documents for the purpose of text block identification and extraction from images.
Solution
We developed a custom image parser to identify and extract blocks of text from legal documents, categorize them, clean up specific categories for OCR, and send those categories to Tesseract. After that, scikit-image and OpenCV were used for image manipulation and retrieval of information of interest.
Results
The implemented solution successfully recognized sections of contracts and their structural elements. This automation effectively replaced routine manual work, resulting in an annual cost savings of $60,000 for the client.
“Michael (CTO of DEX Technologies) is clearly a master in his field.
I recommend him to you! “
William Flynt, CEO of Ferrovia Capital, Phoenix