Abstract We propose a new lossy compression approach for hyperspectral images which comes from hadamard transform (HT) spectral imager for space-based surveillance that uses the Digital Micro-mirror Device (DMD) as the programmable mask other than the conventional mechanical mask. This paper firstly introduces that the DMD Hadamard spectrometer described in this paper is good at obtaining multi-channel high-energy high-SNR spectral data, which is superior to traditional ones. And then,based on the hardware platform composed of Xilinx FPGA and ADV212 (a codec chip for special JPEG2000), the sampled image data was then compressed with high SNR,high-resolution compression ratio. Last, we show the experimental results which demonstrate that the PSNR is up to 47.3dB and make no significant distortion for vision. Obviously, this approach takes distinguished advantages on real-time ability, image distortion and the practical value. The design can be utilized in national defense, earth resource survey aerospace, environmental disaster monitoring and other fields.
[1]
Andreas Uhl,et al.
Hardware-based JPEG 2000 video coding system
,
2007,
Electronic Imaging.
[2]
P. Treado,et al.
Multichannel Hadamard Transform Raman Microscopy
,
1990
.
[3]
L Bolinger,et al.
High-resolution imaging using Hadamard encoding.
,
1999,
Magnetic resonance imaging.
[4]
Zhongyuan Zhao,et al.
A Real-Time High Resolution Image Compression System Based on ADV212
,
2009,
2009 2nd International Congress on Image and Signal Processing.
[5]
Nicolas H. Younan,et al.
JPEG2000 coding strategies for hyperspectral data
,
2005,
Proceedings. 2005 IEEE International Geoscience and Remote Sensing Symposium, 2005. IGARSS '05..
[6]
N. Sloane,et al.
Hadamard transform optics
,
1979
.
[7]
A. Marshall.
Fourier, Hadamard, and Hilbert Transforms in Chemistry
,
1982
.