Cloud Removal & Compression Solution

The power of EdgeAI brought to small satellites

Part of the CogniSAT™ family, CogniSAT-CRC is a self-contained solution combining the power of the CogniSAT-XE family of On-Board AI Payload Processers with state of the art lossless image compression and a flight-proven AI-based Cloud Detection and Removal algorithm. This solution guarantees that images are analysed on satellite so that only useful images are transmitted to the ground station and that this transmission is as efficient as possible.

The use of this solution can result in a six-fold increase in the number of useful images which can be transmitted to the ground station each day thereby increasing the revenue-generating potential of the Earth Observation asset.

AI Engine combines image analysis and compression

Maximising Space Asset Utilisation

All optical Earth Observation sensors are subject to their images being obscured by atmospheric conditions, most notably cloud cover. It is estimated that, at any time, over 50% of the earth’s surface is obscured by clouds. This means that at least half of the information captured by a satellite-based imager is of no use for most applications. Transmitting this data, even if compressed, results in communications bandwidth being wasted. The CogniSAT-CRC solves this challenge by removing these cloudy images before compression. This typically results in a six-fold increase in the usable data received by the ground station when compared with the transmission of uncompressed images and a three-fold increase when compared with the use of standard compression approaches.

Additionally, as only useful images are gathered for transmission, the imaging system can capture a greater surface area on each orbit thereby maximising the utilisation of the EO asset and the potential financial return. Studies carried out by Ubotica have shown that the use of the CogniSAT-CRC in conjunction with state of the art imaging systems can result in a 80% reduction in downlink costs while, at the same time, providing a three-fold increase in the surface area of the earth which is successfully imaged.

Modular Implementation

The CogniSAT-CRC solution combines state-of-the-art Neural Network-based cloud detection and compression to efficiently deliver only those images which contain useful information for transmission to the ground station. This approach guarantees the optimum utilisation of the communication network.

The CogniSAT-CRC is a self-contained system solution which incorporates an AI coprocessing module and control application.  Most of the processing load is directed at the AI coprocessor thereby resulting in minimal processing load to the satellite On-Board Computer (OBC) or Payload Processor (PP). The solution includes a lightweight application which runs on the OBC/PP and coordinates the processing of the targeted image set. The satellite operator can set the cloudiness threshold for each tile to optimise the  operation of the CogniSAT-CRC for their desired area of interest. Documentation provided with the CRC solution details the operation of this application and its command structure.

To find out more
contact us and one of our team will be in touch.