News for hackers. Stuff that matters.

A comment on Advanced Image Manipulation and Data Extraction.

Re:Advanced Image Manipulation and Data Extraction by peterburkimsher on Friday December 22, 2017 @01:43AM (#15985690)
SmartTextExtraction looks like exactly what I need! However, I have to be able to run offline.

How can I download it and run the algorithm locally?

Tesseract gave me pretty poor results, probably because I'm using photos of printed pages instead of flatbed scans.

[ Parent | on HN ]
Re:Advanced Image Manipulation and Data Extraction by mikeyanderson on Friday December 22, 2017 @01:13PM (#15989568)
We have the source code for that algo: https://algorithmia.com/algorithms/character_recognition/tes...
[ Parent | on HN ]
Re:Advanced Image Manipulation and Data Extraction by dan_quixote on Friday December 22, 2017 @02:50AM (#15985906)
I have never achieved reliable results with Tesseract regardless of image quality.
[ Parent | on HN ]
Re:Advanced Image Manipulation and Data Extraction by newusertoday on Sunday December 24, 2017 @03:24AM (#15998606)
did you tried pre processing the images before giving it to tesseract? i had very good results after pre processing the images.
[ Parent | on HN ]