Tesseract ocr performance is slow

Question

I use wrapper tess4j for tesseract 3 to make OCR on cyrillic characters. Before I execute image pre-processing: skewed image, remove background and noise, adjust image quality and finaly get ROI for OCR. Image pre-processing takes 0.5-1 seconds. But tesseract OCR process takes a long time.

How can I enhance tesseract OCR speed perfomance?

Answer 1

There are two ways to accelerate tessercat. It refers to c++ version of tesseract.

1. Tesseract can enable parallelization of computations. See "tessedit_parallelize" variable.

2. You can use OpenCL version of tesseract.