IRIS-BASED HUMAN IDENTITY RECOGNITION WITH MACHINE LEARNING METHODS AND DISCRETE FAST FOURIER TRANSFORM
Abstract:
Human identity recognition plays a crucial role in various applications, such as security systems, access control, and biometric authentication. Among the different biometric modalities, the iris has gained significant attention due to its uniqueness and stability over time. This abstract presents a novel approach for iris-based human identity recognition that combines machine learning methods with the discrete fast Fourier transform (FFT).
The proposed method consists of multiple stages. Initially, the iris region is segmented from the input image using advanced image processing techniques. The segmented iris region is then preprocessed to enhance its quality and reduce noise. In the next stage, the discrete fast Fourier transform is applied to the preprocessed iris image to extract discriminative features from the frequency domain. The use of FFT allows capturing the unique frequency patterns present in the iris texture, which are essential for identification purposes.
Once the iris features are extracted, a machine learning algorithm is employed for classification. Different machine learning techniques, such as support vector machines (SVM), random forests (RF), or deep neural networks (DNN), can be utilized for this purpose. The classifier is trained on a large dataset of iris images, where each image is associated with a specific individual’s identity. During the testing phase, the trained model predicts the identity of an unknown iris image based on its extracted features.
The proposed approach offers several advantages. Firstly, the use of the discrete fast Fourier transform allows efficient extraction of frequency-based features from the iris texture, which enhances the recognition accuracy. Secondly, the integration of machine learning algorithms enables robust and scalable identification capabilities. Furthermore, the iris modality provides a high level of security due to its uniqueness and resistance to forgery.
Experimental evaluations on benchmark iris datasets demonstrate the effectiveness of the proposed method. The results show superior recognition performance compared to existing iris recognition approaches. The combination of the discrete fast Fourier transform and machine learning algorithms exhibits promising potential for iris-based human identity recognition in real-world applications.
Keywords: Iris recognition, biometrics, discrete fast Fourier transform, machine learning, feature extraction, classification.
IRIS-BASED HUMAN IDENTITY RECOGNITION WITH MACHINE LEARNING METHODS AND DISCRETE FAST FOURIER TRANSFORM , GET MORE MASTERS COMPUTER SCIENCE