Holistic Information Theory of Spatial Remote Sensing Imaging

TL;DR

Holistic Information Theory for Spatial Remote Sensing treats the imaging chain as a unified information-transfer system, linking optical, electronic, and computational subsystems through the MTF-SNR product. The core insight is that the information capacity $I_{info}$ scales with $\sum_{(u,v)} \log_2(1+\mathrm{MTF}(u,v)\cdot\mathrm{SNR}(u,v))$, and that restoration performance is governed by the noise-power magnification $\mathrm{NPM} = \frac{1}{Z}\sum 1/\mathrm{MTF}^2$. The authors show that increasing integration time can compensate for smaller apertures, enabling low-cost, high-resolution remote sensing exemplified by the Jilin-1 constellation. They also present a design methodology that jointly optimizes optics, detectors, and algorithms under MSP constraints, validated by simulations and physical experiments that demonstrate consistent reconstruction quality across high- and low-MTF, high- and low-SNR configurations. This framework offers a practical pathway to cost-effective, high-resolution satellite imaging by trading optical complexity for improved signal-to-noise and integration strategies.

Abstract

To address the non-optimal global design caused by the independent optimization of optical lenses, photodetectors, and computational processing subsystems in traditional remote sensing imaging system design, this paper proposes a holistic information theory for spatial remote sensing imaging. This theory integrates the optoelectronic imaging hardware front end and computational reconstruction back end into a unified framework. It establishes a complete spatial imaging chain information transfer model with the objective of obtaining the required effective information. The paper innovatively proposes a quantifiable Modulation Transfer Function (MTF)-Signal-to-Noise Ratio (SNR) product criterion. It demonstrates that the system information transmission ability is determined by the product of MTF and SNR, and that these parameters can compensate for each other to achieve equivalent information transfer. Validation through a high-resolution Earth observation system case shows that under consistent reconstruction mean square error conditions, increasing time delay integration stages reduces optical aperture size and significantly lowers primary mirror mass. Simulations and physical experiments further indicate that by increasing integration time, low-resolution optical systems achieve reconstructed fidelity comparable to high-resolution systems. This verifies that small-aperture optical systems can achieve equivalent imaging performance by enhancing SNR. This theory has been successfully applied in the design of the Jilin-1 satellite constellation, providing a new paradigm for low-cost high-resolution remote sensing systems.

Figures (3)

• Figure 1: The required SNR and integration stages vary with optical aperture to maintain an equal level of MSE in the reconstructed image.
• Figure 2: Simulated imaging and computational reconstruction results of the same target object under a high-MTF, low-SNR system and a low-MTF, high-SNR system. The target object used in the simulation experiment and the region of interest labeled by red box is shown in (a). The images of the target region with the two systems is shown in (b) and (c). The reconstructed images based on system parameters and the images is presented in (d) and (e). The grayscale values of periodic textures labeled by the blue line within the target regions is illustrated in (f) and (g). The system with a higher F-number produces images with lower contrast of the target texture, but after reconstruction, both images maintain consistent contrast and fidelity.
• Figure 3: Experimental setup and results. Figure (a) shows the experimental setup, where the grayscale-to-radiance mapping relationship for the LED and the non-uniform response of the camera are calibrated as much as possible. Figure (b) shows the experimental results, examining the PSNR of reconstructed images for three targets, The PSNR values are 49.91, 50.50, and 50.37 in the F/# 2.8 system, respectively, while there are the values of 50.42, 50.67, and 50.92 in the F/# 16 system. The reconstruction quality of both systems was basically maintained at an equivalent level. Please zoom in to observe pixel level details.