Cloud brightness evaluation from ground-based total sky images using the HSV color space parameter

Resumo

The RGB color space and two-dimensional red-to-blue band thresholding methods have been widely used for cloud detection throughout history. However, there is a lack of experimental data to assess the nuances of the effectiveness of pixel segmentation techniques in identifying Solar Radiation Enhancement Events (REE). As a result, a study was conducted to correlate surface solar irradiance and cloud pixels during the occurrence of these events, since there is no systematic description of the use of the HSV code parameter to determine cloud brightness in the literature. Therefore, this paper proposes a method to measure image brightness under different sky conditions by integrating the HSV brightness parameter from histograms. Uneven illumination in full-sky images is the most complicated factor in evaluating cloud brightness, as it varies with camera type, local atmosphere, and solar zenith angle. Experimental results show that the proposed algorithm proved effective for the evaluated image set, as satisfactory visual effects were observed under both cloudless and cloudy sky conditions.

Keywords: Cloud detection, Solar radiation enhancement, Pixel segmentation, color space