Abstract
This paper presents a new method for fingerprint recognition depending on various sizes of fingerprint images. The proposed algorithm applied on more than 30 fingerprint samples, the results was good.
The proposed algorithm begins with apply enhancement operations on the fingerprint image to eliminate unwanted noise around the fingerprint by using median filter. Then apply thinning operation on the enhanced image and compute co-occurrence matrices for produced image. Next, the properties of the co-occurrence matrices used as inputs of the neural network for recognition process. To speed the recognition process back propagation network used. The ratio of recognition about 100%.
