Cloud Estimation through Image Analytics: Overview and User Evaluation

This presentation will provide an overview of the Cloud Estimation through Image Analytics (CEIA) algorithm developed at MIT Lincoln Laboratory with sponsorship from the FAA. The algorithm uses images from the FAA WeatherCams website, which has cameras deployed across Alaska, Hawaii, Colorado, Maine and a few other states. CEIA was developed to provide an estimate of the sky condition at locations that do not have human or automated observations available. It is intended for use by the aviation and meteorology communities.

CEIA is an extension of the Visibility Estimation through Image Analytics (VEIA) algorithm. VEIA began development in 2016, has been implemented on the WeatherCams test system since 2020, and is expected to be available on the public website in early 2024. VEIA provides a meteorological estimate of the visibility using the presence and strength of edges in each scene below the identified horizon. CEIA uses the functionality of VEIA to extract the sky portion of the image for analysis. The CEIA algorithm then evaluates two features of the sky portion of the image: the presence and strength of edges as well as the ‘grayness’ of the image. The grayness is defined as the number of pixels in the sky that are in a common range of color associated with clouds. An image with a large grayness is theorized to indicate overcast sky conditions. An image with a large number of strong edges in the sky is theorized to indicate scattered or broken cloud cover. A high edge strength with high grayness may be broken, while lower grayness may be scattered. Additionally, CEIA uses VEIA’s weighting scheme for each camera to ensure the quality of each image used in its analysis.

Along with an overview of the ongoing CEIA work, this presentation will provide a summary of a user evaluation and demonstration that was conducted in summer 2023. Case studies will be presented exhibiting CEIA’s skill at identifying cloud cover when compared to accompanying Automated Surface Observing Systems (ASOS) observations, as well as a discussion of cases for which they differ.

Distribution Statement A. Approved for public release: distribution unlimited.

This material is based upon work supported by the Federal Aviation Administration under Air Force Contract No. FA8702-15-D-0001. Any opinions, findings, conclusions or recommendations expressed in this material are those of the author(s) and do not necessarily reflect the views of the Federal Aviation Administration.