A comparison of Satellite Image Compression methods in the Wavelet Domain

Abstract

Better resolution of remotely sensed satellite images will make images clearer and interpretation easier but will increase the total volume of data that has to be managed. In order to reduce data volume for easier satellite communication transmission and reduce the total volume of data needed to be stores, the images should be compressed. Image compression in wavelet domain can be used for both lossy or lossless compression. Four major compression methods are available using the wavelet domain, i.e. CCSDS, Wavelet, Bandelet, and JPEG 2000. Some optical satellite images, were used as input data in simulation software which analyzed and compared the four compression methods in the wavelet domain. The result showed that the CCSDS method yielded the fastest compression and decompression time, but the Bandelet method retained better image quality when reconstructing original images or approximations of them compared to CCSDS. The JPEG 2000 method delivered better quality images than CCSDS for low bit rate. In summary at a rate of 0.25 bpp, CCSDS is 15 times faster than Bandelet and 3 times faster than JPEG2000. However, CCSDS quality is lower by up to 8.77% compared with Bandelet and up to 13.64% compared with JPEG2000.