I need to be able to determine if two images contain the same object. A perfect example would be two photos of a licence plate at different angles.
I've been thinking about OCR (Optical Character Recognition), which would probably get the job done, but I would really like to capitalize on things other than just text (the oil smudge at the top right corner of the licence plate, the dent at the bottom, etc...). This led me to feature matching algorithms like SIFT (Scale Invariant Feature Transform).
I also know that the license plates will always use the same font for the characters printed on it (and have ny states symbol on it) , so maybe some machine learning to train it to that specific character set is in the cards as well? I'm looking for any and all means to reduce mismatching.
In summation, is there a vendor out there that sells an sdk I can incorporate or some opensource code out there that has the following:
Appreciate the help!
There is some great samples that help to you learn about license plate recognition methods like emgu library :
http://www.emgu.com/wiki/index.php/License_Plate_Recognition_in_CSharp
Emgu is open source image processing and machine learning library. Thanks.