Fingerprint verification applying invariant moments

Traditional security systems use passwords or ID cards have been used to moderate access to restricted systems, but these kind of systems have a poor performance because the security can be easily breached. Based in this disadvantage, the biometrics systems have a great popularity, the mains biometrics systems are: face recognition, iris recognition, voice recognition, fingerprint recognition and sinning recognition. The fingerprint recognition is the oldest method used to recognition or verification of person. Our proposed a people recognition system with verification by invariant moments using two methodologies for the fingerprint enhancement. The goal in this work is to get a robust system in security issues. In this work a method for fingerprint verification is considered using a combination of Fast Fourier Transform (FFT) and Gabor Filters by image enhancement, both methods are first applied separately and later on an algebraic sum is done to obtain a single output. After that, a thinning algorithm is applied to get an image with the minimum thickness of 1 pixel. After this thinning algorithm, we apply an algorithm to look minutiae using a window of 3 by 3 pixels to scan the image. In this work we extract two types of minutiae, bifurcation and ending. Then, the feature vector is generated with the distance between minutiae, angle between minutiae and coordinates. In the recognition stage using the coordinates from the minutiae position on the image a comparison is done. After that, apply a verification stage using the invariant moments. With the invariant moments values a comparison is done. The comparison was done using the values obtained for the images into database and the test image for to get the output. The results obtained in this research are better when we used FFT and Gabor filters algorithms to image enhancement than we used separately.