Facial recognition systems commonly suffer from degraded performance when processing low-quality images due to factors such as blur, occlusion, and extreme head poses. Existing Face Image Quality Assessment (FIQA) techniques typically evaluate these quality attributes in isolation, neglecting their combined effect on facial recognition accuracy. To address this limitation, we propose U-FQA (Unified Face Quality Assessment Score), a novel approach that integrates three critical quality factors–general image quality, facial occlusion, and head rotation–into a unified metric. U-FQA combines these components using empirically determined weights to create a comprehensive quality score aimed at filtering unreliable images, thus improving the detection and rejection of unknown identities in recognition systems. We conducted extensive experiments on facial recognition systems to assess the contribution of each U-FQA component. The results demonstrate that each factor significantly affects recognition confidence, confirming the importance of a unified quality assessment strategy for robust biometric performance.