Bionic polarized light positioning draws inspiration from nature's organisms, which employ the dynamic variations in atmospheric polarization patterns for navigation. This innovative approach offers the advantages of autonomous interference resistance and error non-accumulation. It serves as a crucial method for global positioning in satellite-denied scenarios, particularly for long-range unmanned systems. In this paper, in order to solve the problem that the traditional atmospheric polarization mode measurement method cannot effectively solve the error problems such as polarization image distortion and noise, a four-channel atmospheric polarization mode measurement method based on double linear interpolation is proposed according to the imaging characteristics of the pixel-level polarization camera. Enhancements have been made to the solar position solution model within the linear camera projection framework by incorporating the vector cross-product of the incident light's electric field. This improvement is instrumental in resolving bionic polarized light localization even in cloudy conditions. To address this, we introduce a cloud segmentation algorithm that leverages polarization information. Finally, experiments are designed to validate the theoretical analysis results only under clear sky with few clouds and clear sky with many clouds conditions. The experimental results show that the proposed bionic polarized light localization algorithm achieves an average localization accuracy of 27.5 km under clear and slightly cloudy conditions, and it can reach 13.46 km after removing the clouds under clear and cloudy conditions. The results of the study show that the improved bionic polarized light localization algorithm has a very good effect of enhancement, which is valuable for the realization of the development of high-precision localization of bionic polarized light.