Face-Based Biometric Recognition System Using LBP, DWT, and SVM Techniques

In biometric security, facial recognition has become a widely adopted contactless alternative to traditional authentication methods such as ‎passwords and fingerprint scanning. This research introduces a novel machine learning-based face recognition system that combines Local ‎Binary Patterns (LBP), Discrete Wavelet Transform (DWT) with Cohen-Daubechies-Feauveau 9/7 filters, and Support Vector Machine ‎‎(SVM) classifiers. The proposed system operates through three primary stages: image preprocessing, feature extraction, and classification. ‎LBP effectively captures detailed local texture features of the face, while the CDF 9/7 DWT reduces data dimensionality by transforming ‎spatial features into the frequency domain. Finally, SVM is employed to classify the extracted features with high precision. The system is ‎validated using the ORL face dataset comprising 400 images from 40 subjects, achieving a total success rate of 98.22%, outperforming ‎several existing face recognition approaches. Additionally, the model’s robustness is evaluated using False Acceptance Rate (FAR) and ‎False Rejection Rate (FRR) metrics across various thresholds, confirming its reliability. Overall, this integrated approach presents a ‎practical, efficient, and accurate solution suitable for real-time biometric face recognition applications‎.