A single thermal infrared image can only diagnose conventional temperature anomalies in photovoltaic arrays, such as hotspots and diode issues, and cannot detect surface stain faults on photovoltaic modules. Traditional fault detection methods using thermal infrared images have limited fault type recognition and poor detection performance. This study introduces a fault detection approach for photovoltaic (PV) arrays that integrates thermal infrared and visible light imaging. By leveraging both image types, the proposed method broadens the range of detectable fault types and significantly improves detection accuracy. Experimental results demonstrate that this approach effectively identifies temperature anomalies and surface contamination issues within PV arrays, yielding high overall performance and enhancing fault diagnosis accuracy in photovoltaic power plants.