Impulse noise & dropout filters for JPEG2000 used in Remote Sensing

The paper presents a novel architecture for implementing onboard pre-processing for a JPEG2000 based compressor system for autonomous remote sensing satellite imagery. The radiometric image preprocessing implementation architecture is modeled and evaluated. As the compression ratio increases, the application of LUT on ground is not applicable as the pixel values become dependent on adjacent pixels due to compression. In order to correct video data errors at the data handling system and thereby also ensure optimum performance, an impulse noise filter and pixel dropout correction is also proposed using indexing & 'weighted interpolation'. The new architecture ensures correction for the radiometric errors prior to compression offering superior quality of images. The system is for bandwise compression, due to band-to-band registration issues, thus restricting the compression to single component. The system offers superior performance as compared to a JPEG2000 based system without corrections, in terms of quality for the given data rate.

[1]  David S. Taubman,et al.  High performance scalable image compression with EBCOT. , 2000, IEEE transactions on image processing : a publication of the IEEE Signal Processing Society.

[2]  W. Sweldens The Lifting Scheme: A Custom - Design Construction of Biorthogonal Wavelets "Industrial Mathematics , 1996 .

[3]  Michael W. Marcellin,et al.  An overview of JPEG-2000 , 2000, Proceedings DCC 2000. Data Compression Conference.

[4]  Philip N. Slater Recent advances in sensors, radiometry, and data processing for remote sensing; Proceedings of the Meeting, Orlando, FL, Apr. 6-8, 1988 , 1988 .

[5]  David S. Taubman,et al.  High performance scalable image compression with EBCOT , 1999, Proceedings 1999 International Conference on Image Processing (Cat. 99CH36348).

[6]  Chaitali Chakrabarti,et al.  A VLSI architecture for lifting-based forward and inverse wavelet transform , 2002, IEEE Trans. Signal Process..