Mid-wave infrared super-resolution imaging based on compressive calibration and sampling

TL;DR

This paper introduces a compressive sensing-based super-resolution imaging method for mid-wave infrared cameras, improving calibration and achieving 4x4 super-resolution for long-distance objects.

Contribution

It presents an improved calibration method and demonstrates 4x4 super-resolution imaging in MWIR using compressive sensing, reaching the system's design limits.

Findings

Enhanced calibration matrix measurement improves image contrast and SNR.

Achieved 4x4 super-resolution reconstruction of long-distance objects.

System reaches the super-resolution limit based on design constraints.

Abstract

Mid-wave infrared (MWIR) cameras for large number pixels are extremely expensive compared with their counterparts in visible light, thus, super-resolution imaging (SRI) for MWIR by increasing imaging pixels has always been a research hotspot in recent years. Over the last decade, with the extensively investigation of the compressed sensing (CS) method, focal plane array (FPA) based compressive imaging in MWIR developed rapidly for SRI. This paper presents a long-distance super-resolution FPA compressive imaging in MWIR with improved calibration method and imaging effect. By the use of CS, we measure and calculate the calibration matrix of optical system efficiently and precisely, which improves the imaging contrast and signal-to-noise ratio(SNR) compared with previous work. We also achieved the 4x4 times super-resolution reconstruction of the long-distance objects which reaches the limit of the system design in our experiment.