Energy compaction based novel Iris recognition techniques using partial energies of transformed iris images with Cosine, Walsh, Haar, Kekre, Hartley Transforms and their Wavelet Transforms

The theme of work presented in this paper is a novel Iris recognition technique using partial energies of transformed iris image. To generate transformed iris images, various transforms like Cosine, Walsh, Haar, Kekre, Hartley transforms and their wavelet transforms are applied on the iris images. Feature vectors are then generated from these transformed Iris images using the concept of energy compaction of transforms in higher coefficients. 5 different ways are used to generate the feature vectors from the transformed iris images. First way considers all the higher energy coefficients of the transformed iris image while the rest considers 99%, 98%, 97%, and 96% of the higher energy coefficients for generating the feature vector. Considering partial energies reduces the feature vector size thus lowering the number of computations and results shows that this gives better performance. To test the performance of the proposed techniques, Genuine Acceptance Rate (GAR) is used as a metric. Better Performance in terms of Speed and Accuracy is obtained by considering Partial Energies. Among all the Transforms and Wavelet Transforms, Walsh Transform and Walsh Wavelet Transform gives highest GAR value. Results show that most wavelet transforms outperforms other transforms. Also, using Partial Energy gives better performance as compared to using 100% energies. The proposed technique is tested on Palacky University Dataset.