n efficient compressive imaging method with modulation transfer function compensation

Chen, Xinyi

doi:10.1016/j.optcom.2025.132632

Return

n efficient compressive imaging method with modulation transfer function compensation

2026-10-01

PRE

Chen, Xinyi

Thou, Han

Sanz, Angel S.

Sanchez-Soto, Luis L. *

DOI：10.1016/j.optcom.2025.132632

Original

Original request for help

Save

Abstract

En 中文

Compressive imaging is a valuable remote sensing imaging method, because it can directly acquire a compressed image with a few pixels at the orbit imaging stage. At the ground, the reconstructed image can obtain a high resolution version with several million pixels, an approximated original image captured by conventional remote sensing optical cameras. However, a compressive sensing based imaging system often suffers from the same drawbacks as traditional Nyquist sampling theorem based imaging systems, where the captured images are unavoidably degenerated by point spread functions and noises of the imaging link. Here, we propose an efficient compressive imaging method with modulation transfer function compensation for remote sensing cameras. This method relies on the alternating iterative inversion of two Fourier spectrum functions (the modulation transfer function and the object image spectrum) by a sequence of unconstrained optimization subproblems. Experimentally, we have confirmed that the method can recover the degenerated image from incomplete measurements through the proposed compressive sensing recovery algorithm. This method thus provides a more promising way of blind deconvolution from compressed measurements for compressive sensing based imaging influenced by blurring and noise disturbances.

Keywords:

Compressive sensing

Deconvolution

Modulation transfer function

Cameras

Measurements