No abstract

Abstract

Image compression is an important requirement of imaging payloads on board Earth Observation satellites. This paper presents a new on-board real-time compression system, capable of lossless and lossy image compression. A cost-effective lossless image compression scheme, based on the CCSDS recommendation, is proposed and tested with multi/hyperspectral images. An efficient hardware implementation is achieved using FPGA-based acceleration. The hardware accelerator design is optimized by employing novel techniques at algorithmic and logic levels

This paper has proposed an automatic onboard multispectral image compression system, for LEO Earth Observation satellites. This system has combined tiling, sub-pixel band registration, radiometric calibration, and multispectral image compression, to achieve automatic on-board image processing requirement. Two radiometric calibration methods are compared, and Brightness Difference Compensation (BDC) is preferred. A new robust band registration technique, which is gradient image based phase correlation (GradPC), is proposed. Experimental work based on natural satellite test images is given and results validate the system design.

No abstract