Hi, I am working on tesseract. My goal is to make a real-time system. So I do want it to run on almost 20+ FPS.
I am getting almost 0.5s - 1.5 s on a single frame via tesseract. I have tried alot of things to improve this speed but of no use. Some of the tried things are * fast integer models given by Tesseract * tessedit_do_invert=0 argument * gray-scaling (dont need further pre-processing as images do not have noise etc. * 4 threads via OMP_THREAD_LIMIT * 4 processes via tessedit_parallelize But none of these gave any improvment except tessedit_do_invert=0 and gray-scaling. I do not want to loose any chracter so can not downsize it more. I have implemented custom multi-processing to run 4 tesseract instances in parallel and it has boosted the speed a bit. I would need suggestions and techniques in-order to improve the latency of the OCR(cant skip the frames). I want to make it to at-max 0.1 seconds per image. If you do have any other OCR engine in mind, let me know. Thanks. -- You received this message because you are subscribed to the Google Groups "tesseract-ocr" group. To unsubscribe from this group and stop receiving emails from it, send an email to tesseract-ocr+unsubscr...@googlegroups.com. To view this discussion on the web visit https://groups.google.com/d/msgid/tesseract-ocr/6d8a19f5-f679-44a4-97e5-a6ecca52f441n%40googlegroups.com.
