Retinal scan recognition using wavelet energy entropy

Retina blood vessels have high degree of uniqueness making retina based biometric systems as an emerging security and authentication mechanism. With increased advancement in digitization of authentication and medical image processing, there is demand of accurate and faster digital image processing. Wavelets are excellent in extracting localized texture information in digital images. Blood vessels have vessels with different thickness and width; they can be analyzed using multi-resolution analysis method. A retina feature, named wavelet energy feature (WEF) is defined in this paper, employing wavelet, which is a powerful tool of multi-resolution analysis. WEF can reflect the wavelet energy distribution of vessels with different thickness and width in several directions at different wavelet decomposition levels, so its ability to discriminate retinas is very strong. This paper also presents new and faster type of wavelets called Kekre's wavelets for creating WEF and extracting retinal feature vector. Wavelet energy entropies based on Kekre wavelets are calculated for retina features and used for match using Euclidian distance. The paper finds retinal match accuracy using Kekre wavelet better than Haar wavelets. Use of wavelet for segmenting the blood vessels and use of Energy Entropy for feature extraction holds promise of simplicity and computationally less expensive.