SkyCam: A Dataset of Sky Images and their Irradiance values
This dataset addresses the need for accurate local solar radiation prediction, which is incremental as it provides new data for existing methods.
The authors tackled the problem of predicting solar radiation locally by creating SkyCam, a dataset of sky images paired with precise irradiance values from three locations in Switzerland over a year, enabling image-based deep learning solutions for short-term solar radiation prediction.
Recent advances in Computer Vision and Deep Learning have enabled astonishing results in a variety of fields and applications. Motivated by this success, the SkyCam Dataset aims to enable image-based Deep Learning solutions for short-term, precise prediction of solar radiation on a local level. For the span of a year, three different cameras in three topographically different locations in Switzerland are acquiring images of the sky every 10 seconds. Thirteen high resolution images with different exposure times are captured and used to create an additional HDR image. The images are paired with highly precise irradiance values gathered from a high-accuracy pyranometer.