OCR scanning of directory pages with small font and processing of scanned text into n
- or -
Post a project like this- Posted:
- Proposals: 6
- Remote
- #2149101
- OPPORTUNITY
- Expired
Description
1) Our own attempt at OCR didn't produce accurate results because the font is so small - perhaps the scans need to be enhanced before OCR?
2) After scanning the data needs to be normalised - for example the last name is in block capitals, if the next row has the same last name only the initial is shown - so the last name needs to be added.
We need the data as: name, address, phone number
We can scan the data into jpg one page per file, or PDFs with multiple pages
See example PDF file scanned in high dpi greyscale https://www.dropbox.com/s/80v0rv8mwd0bxpw/2018_09_21_18_22_42.pdf?dl=0
3) I would prefer that you provide software and advice to enable us to scan and process the files here.
Can you show me a sample of the output?
Graham L.
97% (49)New Proposal
Login to your account and send a proposal now to get this project.
Log inClarification Board Ask a Question
-
I think this is the correct format, Right?
Prefix First name MId name Last name Suffix Address Contract
Mr John Olker David Jr 21 Garderton Ct, Pershore 01886-821-788 -
Dear Graham,
Do you need the output data in an Excel Spreadsheet?
Warm regards.
Shafeeu -
I've tried the sample and found that 1. The scan needs straightening 2. importing to Photoshop and levels adjusted 3. Opened and exported from Acrobat. This makes a decent job of OCR but, not perfect, Excel file then needs editing. So, labour intensive and expensive. You may be better trying to access digital directory?