Re: Resize of image improves performance? What other pre-processing can I do?

I am in the same situation. Here is what I have experienced. It helps to remove non-text from the image, such as underlining, graphics, boxes, lines, shading. Grayscale and black-and-white images work better than color, I have heard. If you follow the training document and make a box file from

Re: Help in starting up tesseract in c++

Check the posts from Remi Thomas for tessnet2. If the pictures are like scanned pages, you might not need to pre-process them; otherwise you might need to threshold them, perhaps a local adaptive threshold. On Dec 9, 7:24 pm, Jun Liang wrote: > Hi I am a student who is using tesseract to do some

For people interested in Indic OCR

I found this book lying on a table in the CVIT lab at IIIT Hyderabad < http://cvit.iiit.ac.in/> . I also found the ebook on books.google.com. Go ahead and read the book at http://books.google.com/books?id=WdSR9OJ0kxYC&lpg=PA136&ots=ACmyCZ1n0P&dq=ocr%20cuts%20and%20merges&pg=PR1#v=onepage&q=ocr%20cu

Re: Tesseract-ocr does not have any output

Hi I am using tesseract-ocr in Iphone.I have successfullt compiled it for the IPhone simulator and it is working nice and fine. But the prblm is that i cannot compile it for the device. It will be very nice if somebody can help me out on this issue.. Lots of Thanks in advance... Best Regards Sub

[no subject]

Hello Everyone, Thanks to you and your team for providing such a befitting solution for OCR. I am a bit stuck and was hoping a little help from your side...Posted the same message on tesseract forums but couldn't attach the tiff file... Problem is that Tesseract works very well in case of normal

Problem in character recognition

Thanks RAY and team for providing such a befitting solution for OCR. Tesseract works very well in case of normal tiff files but when the input is a bit scrambled and distorted output is not the same. I have attached the concerned tiff file along for reference. Ghostscript is being used to produce