Fuzzy Rider Optimization Algorithm for vein-based biometric recognition

Biometric recognition system plays a vital role for identifying the person automatically using the characteristic vector from its behavioral or physiological characteristics. In recent days, recognition system based on vein image is a facilitating and promising biometric technology. Vein recognition uses various modalities, like hand, palm and finger images for the identification of person. This article introduces the model for vein-based biometric recognition using the rider- based score level fusion method. The developed method includes four steps, which are pre-processing, vein extraction, feature extraction and biometric recognition. The first step is to input the training images, like finger, hand, and palm into pre-processing step. In the pre-processing level, noises are removed from the input image and the Region of Interest (ROI) extraction is also executed in pre-processing module based on neighbourhood search method. After that, veins are extracted for each modality using the holoentropy, enlightening of image, and the circular averaging filter. After vein extraction, the features are acquired, and then, Entropy- based Euclidean Distance (EED) is developed for score level fusion. At last, the biometric identification is carried out based on the proposed Fuzzy Rider Optimization Algorithm (Fuzzy ROA). The proposed Fuzzy ROA is newly designed by combining the Fuzzy logic and the Rider Optimization Algorithm (ROA). The performance metrics, such as False Rejection Ratio (FRR), False Acceptance Ratio (FAR) and accuracy is analyzed, and then compared with other methods. The evaluation of the developed method attains the maximum accuracy of 90.19%, minimum FAR and FRR of 10.09% and 0.87% respectively.