Clouds play an important role in weather and climate-related investigations. However, they often influence the quality of images and waste resources of storage and bandwidth in remote sensing. So, it is critical to detect clouds for less cost of payload. In this paper, the design of a real-time cloud detection camera for small satellite platforms is proposed based on field programmable gate array (FPGA). Two MicroBlaze Soft Cores are embedded in the FPGA to accomplish the task without other chips assist. By using this way, the system is highly programmable and integrated, the weight of which also becomes lighter. We implemented the system on a Xilinx Virtex-4 FPGA. The test results show that the signal-to-noise ratio (SNR) is 128.1 at 80% of the saturated exposure. We select Arabian Peninsula-Pakistan-West India area to evaluate the cloud judgment accuracy. Compare with moderate resolution imaging spectroradiometer (MODIS) cloud mask products, the false alarm rate (FAR) is less than 3%. The application of the proposed approach in a simulation and engineering system indicates its effectiveness and practicability.
|