Design and Calibration of a Novel Camera-Based Bio-Inspired Polarization Navigation Sensor

Abstract

The remarkable ability of animals using the skylight polarization pattern provides a significant inspiration for the autonomous robotic navigation. The bio-inspired polarization navigation approach with advantages of efficiency and reliability has aroused much interesting for experts to perform further research. The sensor of measuring the skylight polarization pattern plays a key role in bio-inspired polarization navigation. A novel camera-based bionic polarization navigation sensor is designed and implemented. We propose a robust measurement method based on the least squares by utilizing all outputs of measurement units. For improving the precision of measurement, the errors of sensor are analyzed, and an effective calibration algorithm is proposed in detail. The experiment results show that the measurement accuracy is improved greatly after calibration. Finally, to evaluate the performance of our sensor, experiments of measuring the actual skylight polarization pattern are performed and compared with the results from the single-scattering Rayleigh theory. The comparison results indicate that our sensor can achieve an effective and accuracy measurement in practice.