Professional Documents
Culture Documents
Team Members
M.AKSHAY (2013506060)
For example, the staffs who work for the money exchanging (e.g. Forex Bank)
have to distinguish different types of currencies and that is not an easy job.
This may cause some problems (e.g. wrong recognition), so they need an efficient
and exact system to help their work.
EXISTING SOLUTION:
There are already some papers related to Currency Recognition System using
image processing. The papers are:
http://www.diva-portal.org/smash/get/diva2:324361/ FULLTEXT02
PROPOSED SOLUTION:
The solution is we use image processing software to find the face present in each
currency and then their eyes are extracted from the face. We use image processing
softwares like MATLAB which helps us to segment, enhance and make the image
help us get the details and makes our evaluation easy.
Here we train the software by adding many currencies so that the system exactly
identifies the face and their eyes present in the currencies. Then the serial numbers
present in the currency can be extracted from the image using Optical Character
Recognition.
MATLAB CODE:
USING CASCADE OBJECT DETECTOR:
1. Face Detection:
INPUT OUTPUT
2. Eye Detection:
EyeDetect = vision.CascadeObjectDetector('EyePairBig','MergeThreshold',1,
'MinSize',[11 45],'MaxSize',[20 50]);
BB=step(EyeDetect,I);
Eyes=imcrop(I,BB);
figure,imshow(Eyes);
title('Eye');
INPUT OUTPUT
REFERENCES:
http://in.mathworks.com/help/matlab/ref/rgb2gray.html
http://in.mathworks.com/help/vision/ref/vision.cascadeobjectdetector-
class.html
http://in.mathworks.com/help/matlab/creating_plots/the-image-object-and-
its-properties.html