OCR with Python and tesseract
Budget: €30 – €250 EUR
Text in images has to be read using Tesseract and Python 3.7. The code must run under Windows 10.
For example, IMG_20210608_132717_hl.jpg_0.jpg , the result should be (the last line is the time on the side.)):
AD09CX
28/02/2023
12:33
Some help:
- there is no letter 'O', only numbers '0'
- the second line must be a date like dd/mm/yyyy
- the line on the side is a time like hh:mm
The whole data is given in Orig.zip, 568 images.
The accuracy has to be 99% of the images - so 562 images must be read correctly!
Documented source code has to be delivered. Only OpenCV, Teseract and standard Python packages are allowed. (If you want to use other packages or something else, please ask.)
Time frame: Absolute maximum is 7 days.
For example, IMG_20210608_132717_hl.jpg_0.jpg , the result should be (the last line is the time on the side.)):
AD09CX
28/02/2023
12:33
Some help:
- there is no letter 'O', only numbers '0'
- the second line must be a date like dd/mm/yyyy
- the line on the side is a time like hh:mm
The whole data is given in Orig.zip, 568 images.
The accuracy has to be 99% of the images - so 562 images must be read correctly!
Documented source code has to be delivered. Only OpenCV, Teseract and standard Python packages are allowed. (If you want to use other packages or something else, please ask.)
Time frame: Absolute maximum is 7 days.