<p>Biometric recognition systems are essential for secure authentication, leveraging unique physiological and behavioral characteristics to identify individuals. The Biometric recognition systems risk of unauthorized access is heightened. This study presents the development of a multimodal biometric recognition system that incorporates vein recognition, specifically palm and knuckle veins, as a reliable alternative to traditional methods. The proposed system employs a contrast-enhancement technique for pre-processing to improve image quality, while Gray-Level Co-Occurrence Matrix (GLCM) and Discrete Wavelet Transform (DWT) are utilized for feature extraction. Feature selection is performed using the Chimp Optimization Algorithm (ChOA) to reduce computational complexity and enhance classification accuracy, with final classification executed by a Deep Neural Network (DNN). The DNN-ChOA model achieves an accuracy of 99.85%, sensitivity of 98.25%, and specificity of 97%, significantly outperforming traditional and other modern methods. These results underscore the effectiveness of the multimodal biometric system in delivering robust security while meeting the demands of identity digitalization and virtualization efficiently.</p>