Improving Japanese Text Detection
106 views
Skip to first unread message
Mayank Agrawal
Aug 9, 2022, 12:11:16 PM8/9/22
to tesseract-ocr
I currently use Google cloud vision OCR, But it is not able to detect Ichi kanji(A Japanese Character) and The Accuracy is quite poor for Vertical Fonts(tategaki).
The same is the case with tesseract 4.0.
I want to ask if Traning on Tesseract for Ichi Kanji in general and Vertical fonts will significantly improve the accuracy of both of them. If Yes, How Many datasets is required to achieve a significant Improvement in accuracy. Which Tresseract version should I use 4.0 vs 5.0
The same is the case with tesseract 4.0.
I want to ask if Traning on Tesseract for Ichi Kanji in general and Vertical fonts will significantly improve the accuracy of both of them. If Yes, How Many datasets is required to achieve a significant Improvement in accuracy. Which Tresseract version should I use 4.0 vs 5.0
Reply all
Reply to author
Forward
0 new messages