A New Iris Recognition Method Based on Gabor Wavelet Neural Network

After a thorough analysis and summarization, a new method based on Gabor wavelet neural network and 2-dimensional principal components analysis (2DPCA) is proposed for iris recognition. A Gabor-wavelet neural network model is developed in this study. The extraction algorithm layer of GWNN is used for selecting the feature extraction method and obtaining the optimum wavelet basal function parameter values. In this process, Gabor parameters are adjusted adaptively through Gabor wavelet atomic transform function, once defined, Gabor filtering and wavelet methods are used to extract the iris texture features. This will result in a compact and efficient feature vector. In the next verification stage, the 2D principal component analysis (2DPCA) technique and the classification layer structure perceptron of GWNN, which the followed parts layers of network are employed for dimensionality reduction and classification respectively. In the end network simulation experiments can be completed using Gabor wavelet neural networks to classify. Simulation results showed that the proposed iris recognition method based on the Gabor wavelet neural network is a better recognition performance.