This study introduces a novel procedure combining image analysis techniques to examine the temporal changes in natural light, a key aspect in daylighting and built environment research. Our approach utilizes both Low Dynamic Range (LDR) and High Dynamic Range (HDR) camera outputs, leveraging the complementary strengths of both to capture an extensive range of sky conditions, identifying overall light distribution patterns and detailed luminous fluctuations. A key aspect of this study is the simultaneous use of both LDR and HDR imaging to capture intricate light variations, without requiring specialized equipment, and to rely on the potential offered by image processing algorithms to effectively detect subtle luminance shifts. Additionally, our process utilizes deep learning to distinguish between sky and cloud regions, and conducts a detailed comparison with empirical values derived from HDR captures to ensure the robustness of our computational analysis. This offers a practical and economical alternative to conventional methods that depend on dedicated instrumentation like hyperspectral or photosensor-based cameras, thereby broadening its applicability in future daylight studies.