Hand Geometry based Person Identification System using Support Vector Machine

Abstract

In recent times, it is immensely important to identify a person to access the secure data in a system. In this case, verification and identification techniques play a vital role to secure the process. Password system has failed to provide such kind of security. We highly emphasize on the biometric characteristics to design such secured systems. In this paper, we provide a way to identify a unique person by his hand geometry feature. Shape based features of human hand have been extracted using different image processing techniques. Then the feature vector has been classified by multi-classifier support vector machine technique. The output of this person identification technique has provided a better performance which is about 85 percent using low quality non planer, noisy gray images. The system is expected to bring great comfort to users as there is no restriction of the palm position.
Keywords: Biometric, hand geometry, recognition, shape, appearance, support vector machine
Cite this Article: Rizoan Toufiq, Rawshon Raha, Md. Rabiul Islam. Hand Geometry based Person Identification System using Support Vector Machine. Current Trends in Signal Processing. 2015; 5(1): 26-33p.