A Nonideal Measurement Error Model of Polarized Skylight for Model-Based Orientation Determination.

Zhongyuan Zhang,Jinkui Chu,Hao Yu,Ran Zhang,Suiyu Liu
DOI: https://doi.org/10.1109/tim.2023.3315399
IF: 5.6
2023-01-01
IEEE Transactions on Instrumentation and Measurement
Abstract:The orientation determination based on skylight polarization field model, inspired by biological compound eye mechanism, is a novel autonomous navigation technique for large-scene environments without cumulative errors. However, due to the nonideal characteristics of transmission mediums, the skylight polarization information measured on the ground is disturbed by such nonlinear factors, which affects the orientation accuracy of navigation sensors. This article establishes a novel measurement error model of polarized skylight for model-based orientation determination, which fully considers the forward transmission process of polarized skylight through near-surface atmospheric cloud medium and defective depolarized optical lens. A linearized simplified algorithm is proposed for the nonideal skylight polarization transmission model (NSPTM), and the effects of the key extrinsics on the orientation error of two classical polarization-opponent unit (POU) detection models are analyzed in detail. With indoor and outdoor experiments, we explain the NSPTM algorithm reasonableness and verify the orientation error consistency of actual POU detection devices with simulations. This work quantifies the complete forward transmission process of polarized skylight for the real-time orientation computing and provides guidance for the design and practical applications of polarized skylight navigation sensors.
What problem does this paper attempt to address?