Replies: 1 comment
-
You should try finetuning the recognition model. You may need around 5000 images (recommended by the documentation). Optionally, iirc it could help if you try reducing the char dictionary if you're only working with specific characters |
Beta Was this translation helpful? Give feedback.
0 replies
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
-
i am using paddleocr, sometime it confuses between 0 and O , L to 1 , how can I improve that
eg 2000 --> 2O00, GL001 ---> G1001, D003 ---> DO03
Beta Was this translation helpful? Give feedback.
All reactions