Fingerprint Identification with fusion of Gabor features using CNN classifier

This study presents a novel approach to fingerprint identification by integrating Gabor features using a Convolutional Neural Network (CNN) classifier. The methodology begins with preprocessing steps, including input image resizing, morphological operations, and extraction of Gabor phase and texture features. These preprocessing steps aim to enhance the quality of feature extraction from the fingerprint images. Principal Component Analysis (PCA) is then applied to reduce the dimensionality of the feature space, optimizing feature selection. Subsequently, the CNN classifier, leveraging deep learning techniques, is employed for classification. The CNN model is trained using labelled data to recognize distinct fingerprint patterns, including Arch, Left Loop, Right Loop, Tented, and Whorl. Through the training process, the CNN learns to discriminate between different fingerprint patterns based on the extracted Gabor features. The accuracy of the classification system is evaluated to assess its effectiveness in correctly categorizing fingerprints into their respective patterns. By fusing Gabor features with CNN classification, this methodology aims to achieve improved accuracy in fingerprint identification, thereby advancing the capabilities of biometric security systems.

Keywords: Finger print images dataset, CNN, Deep learning techniques, PCA, classification and pre-processing.