Closed novioleo closed 7 years ago
I've been tried the following methods: 1.CTPN-----the model is so big and the train process is private and the speed is unsatisfactory 2.text-line ---- the model is so big
IF YOU HAVE OTHER IDEA,PLEASE LEFT IT.
@novioleo,
had text detector implemented yet?
@pribadihcr EAST~ but i havn't push it~~~
i use the algorithm from opencv_contrib to location the text(temporary),to detect the character area.this algorithm is suit for non-chinese text.i'll replace it with other simpler algorithm ,so that the mobile device can detect the text area much more efficient.