The Negative Selection Algorithm (NSA) is a computational technique inspired by the human immune system and widely used in various fields like intrusion detection, network security, data mining, and pattern recognition. However, its effectiveness in human identification has not been thoroughly explored. This study focuses on utilizing NSA for human image classification, specifically in a bi-modal system combining physiological traits (faces and fingerprints) and behavioral traits (signatures and voices), as well as a uni-modal system using all features. The research collected 2400 images from 200 individuals, pre-processed images, and salient features selected for easy classification. NSA was used for image classification in both bi-modal and uni-modal systems. The results demonstrated NSA's effectiveness, particularly in the bi-modal system. The biometric system that fused behavioral traits exhibited high accuracy, with true positive and true negative rates of 141% and 144%, respectively, and an overall accuracy of 95%. The system is based solely on physiological traits and achieved slightly lower accuracy rates at 89%. Furthermore, among the uni-modal systems, the voice-based system stood out with a true positive rate of 131% and an accuracy of 88.33%. These findings emphasize the advantages of combining different biometric traits, showcasing the potential for increased accuracy in identification systems. The study highlights NSA's role in enhancing classification accuracy, suggesting the developed biometric systems could significantly improve the performance and reliability of various integrated identification systems.