Comparing analysis of multispectral and polarimetric imaging for mid-infrared detection blindness condition.

Abstract

When detection blindness occurs due to thermal crossover for conventional broadband mid-infrared (IR) imaging, multispectral and polarimetric imaging present their advantages in solving the problem to some extent. However, neither of them is satisfactory for all application environments. In this paper, we demonstrate two types of detection blindness situations, analyze their corresponding mechanism, and identify the respective optimum application condition for multispectral and polarimetric imaging. First, a brief classification about a detection blindness situation that arises from self-radiation or reflected radiation of an object is presented. Second, based on the blindness model and detection mechanism, multispectral imaging is analyzed as superior in the emitted radiation dominant scene, while polarimetric imaging is more suitable in the field of the reflection dominant scene. Thirdly, a multispectral prototype and a polarimetric prototype for mid-IR thermal imaging are made and used to conduct a test experiment in two different environmental scenarios. The experimental results show that the multispectral and polarimetric mid-IR imaging systems can enhance the contrast of the detected images when traditional thermal imagery fails. In particular, the multispectral image has a high contrast in the scene that the emitted radiation is dominant over the ambient IR loading, such as the target and background of 295K below a relatively less sunny sky, while the polarimetric imaging is more suitable in the scene where the reflection is dominant when the ambient IR loading is adequate, which is when the direct solar radiation is relatively strong.