Onboard real-time cloud detection using reconfigurable FPGAs for remote sensing

The remote sensing images often encounter the problem of cloud cover, and this may obstruct the use of certain image data. So a cloud detecting process is required to mask out the cloudy pixels. With the trend of satellite missions towards lower cost and wide coverage, the demand of on-board processing becomes more insistent. Reconfigurable logic devices combining the flexibility of software-based solutions with the performance of dedicated hardware are very suitable for building an ideal platform. This paper focuses on the design and implementation of a real-time cloud detection system with high accuracy. The determination whether a pixel indicates cloud is made by the spectrum threshold comparison and the texture analysis. A branching tree diagram is proposed to improve the precision and accelerate the process. A platform has been designed and manufactured to test and verify the cloud detection system. The details of the design and implementation are presented. The experimental results illustrate very promising performance in terms of detection accuracy and execution efficiency. Further improvement is now under way.