Open urlan opened 1 year ago
Have you tried enabling OCR (although I think tesseract is not trained on license plates) and looking at Regex:BR_CAR_PLATE metadata property?
PS: The BR_CAR_PLATE regex should be updated to detect Mercosur plate syntax.
I don't know if you remember, but you said to me to ask this improvement here, due the requeriment of training. I'm quite sure this will help a lot.
BTW, I never tried that. How can I change IPED to do that? Thanks.
I don't know if you remember, but you said to me to ask this improvement here, due the requeriment of training. I'm quite sure this will help a lot.
Now I remember. Actually I suggested asking the feature here because someone else from the community may volunteer to implement it, since this project has very few human resources dedicated, no funding and the desired feature queue keeps growing...
BTW, I never tried that. How can I change IPED to do that? Thanks.
Go to conf/OCRConfig.txt->enableOCR
and set it to true
. Then process your images. After that, open the case and try to locate the possible license plates found in the bottom-left corner:
This is the REGEX to detect Mercosur plate pattern.
MERCOSUL_CAR_PLATE = \b[A-Z]{3}[-\s]?[0-9][A-Z][0-9]{2}|[A-Z]{3}[0-9][A-Z][0-9]{2}\b
I'm testing this and the results are positive. It would be nice if more people could try it before implementation.
Make OCR of license/number plates, search them and show in a tag.