Overexposure-free automatic exposure algorithm for division-of-aperture polarization camera

This paper proposes a two-stage overexposure-free automatic exposure searching algorithm for division-of-aperture polarization (DoAP) cameras to avoid pixel-level polarization information corruption caused by overexposure. The first search determines the safe exposure upper bound to eliminate yielding overexposed pixels in the image captured by any polarization channel, ensuring the integrity of raw polarimetric data. The second search runs within this bound to maximize the mean gradient of four-channel polarized images, ensuring the imaging qualities for subsequent image registration tasks, thereby accurately acquiring polarization information. Experiments show stable, smooth exposure search, offering a robust and efficient auto-exposure scheme for DoAP imaging.