Hello friends,
I have a peculiar problem. Though it looks simple (might be too) but I wasn't able to find the solution to it. I hope your advice will help me solve it. Ok so now without wasting much time, I'll come directly to the problem.
I have an image of a currency note and I want to read the serial no. written on that image. The image is colored.
Sample Images
Rs.1000 http://www.banknotes.com/in94.htm
Rs 500 http://www.banknotes.com/in93.htm
Rs 100 http://www.banknotes.com/in91.htm
Rs 10 http://www.banknotes.com/IN89.JPG
I can guess the type(amount) of note by reading the pixel color value at some specific locations and by matching it with standard values. This will help me find the location of the serial no.
Please help me find a solution to this problem.
Thanking You in advance.
How to read Serial No. of currency note ?
Moderator: General Moderators
Re: How to read Serial No. of currency note ?
Requires OCR. Can become somewhat to extremely difficult, depending on how flexible you want your Serial Nr reader to be.
Are the images of the banknotes that it needs to process always reasonably clean and visible, and with the same orientation (i.e. not somewhat rotated) as in your examples? Or do you intend to process images from random scans or photographs, i.e. the displayed banknotes may be rotated, wrinkled, partially covered by shadow, blurred, noisy, etc?
Are the images of the banknotes that it needs to process always reasonably clean and visible, and with the same orientation (i.e. not somewhat rotated) as in your examples? Or do you intend to process images from random scans or photographs, i.e. the displayed banknotes may be rotated, wrinkled, partially covered by shadow, blurred, noisy, etc?
Re: How to read Serial No. of currency note ?
The images will be as shown in the links above. All images will be clean, visible and with same orientation.Apollo wrote:Requires OCR. Can become somewhat to extremely difficult, depending on how flexible you want your Serial Nr reader to be.
Are the images of the banknotes that it needs to process always reasonably clean and visible, and with the same orientation (i.e. not somewhat rotated) as in your examples? Or do you intend to process images from random scans or photographs, i.e. the displayed banknotes may be rotated, wrinkled, partially covered by shadow, blurred, noisy, etc?