Compressive sensing based acquisition and reconstruction technique for remote sensing image

In this paper, the acquisition and reconstruction technique for remote sensing image (RSI) is proposed based on the new compressive sensing theory. By the sparsity feature of RSI, a less but efficient sparse expressions of the RSI are obtained, using the measurement matrix of the partial Fourier, permute fast Fourier transform and structurally random matrices respectively in RSI acquisition. And the images are reconstructed based on gradient projection of sparse reconstruction. Using 4-meter resolution panchromatic RSI and 32-meter resolution multi-spectral RSI from Beijing-1 micro-satellite, the experimental results indicate that the new technology provides excellent objective and subjective reconstructed quality. The peak signal-to-noise ratio (PSNR) results of the reconstructed images reach to 23dB. So the work has a promising application potential for remote sensing imaging.