Dear all,
I successfully created my 'eng.traineddata' by referring the training
guidelines on tesseract site. I am getting pretty good results with the
fonts I added to my training set.
However I was not getting good results for some images so I searched for
that particular font type on net & add
VietOCR (Java version) does not feed the original image to Tesseract,
but rather it reads and then writes back out an uncompressed TIFF
file, rescaled to 300 DPI if instructed so, which is then sent to the
engine. I found this regurgitated image somehow has been more amenable
to Tesseract.
The pro
Hi everyone,
I am using Tesseract-OCR 3.01 to do license plate recognition. I
trained the database with 200 license plate images without
preprocessing.(License plate have been located from pictures by my
algorithm) Every char has at least 15 samples. However, the results I
got are not good enough.
3 matches
Mail list logo
