All-sky imagers have been widely used for cloud cover monitoring, while advances in image recognition and deep learning are further expanding their observational capabilities. This report will introduce more robust cloud cover estimation under complex illumination conditions, together with richer information on cloud genera and weather conditions. It will also discuss how continuous sky images can be used to track cloud movement and evolution over time. These capabilities can support short-term solar irradiance forecasting and provide valuable input for photovoltaic power forecasting, further extending the application of all-sky imaging in meteorological observation and solar energy.