I am planning to perform ocr on Indian number plates.I used tessract 4.0 beta which uses LSTM engine for ocr. Although recognized characters are not coming out to be correct. i used cv2.Laplacian() while picking up images without blur and performed noise reduction using cv2.fastNlMeansDenoisingColored() on the image . Still results seems not so accurate.
1)Can we please suggest what other preprocessing techniques that need to perform to enhance the image 2)Can we keep a check on tessract such that no special characters are detected.
Test images look like these.
detected characters: *DLBCAUS368 “HR 10¥5803
Thanks in advance