In recent years, with continuous technological advancements, polarized light navigation technology has garnered widespread attention due to its high accuracy and anti-interference capabilities. However, practical applications face the challenge of heading angle ambiguity. This paper proposes a heading angle disambiguation technology based on polarization gradient changes. First, we detail the physical basis of polarized light and its applications in nature, particularly in the navigation abilities of certain insects. We analyze atmospheric polarization models to obtain polarization angle and polarization degree information. Next, we designed a set of algorithms capable of extracting directional information from polarized light scattered in the sky, and then calculating the absolute heading of the device to achieve full observability of the carrier’s heading, ensuring the feasibility of practical navigation applications. Finally, we conducted multiple sets of experiments in an outdoor clear weather environment, using a multi-scale indexing platform as a heading reference to test the performance and accuracy of the navigation system. This research not only effectively addresses the issues in polarized light navigation technology but also explores new possibilities for future applications.