Polarization sensitive mechanism of the POL-neuron model of desert ant (Cataglyphis) offers the basis for the study on bionic polarization navigation. Referring to the angle ambiguity error of the existing signal processing method of POL-neuron model, a new angle method is proposed based on high-order polynomial fitting after analyzing the working principle of the model. According to the relationship between navigation direction and the rotation angle of reference axis, so the effective angle range can be increased from [-45°, 45°] to [0°, 180°] linearly. Finally, the corresponding experiment is also done to test the performance of this new angle method using a 4-channel polarized light navigation sensor in real skies. The result demonstrates that this method can solve the angle ambiguity error of POL-neuron model effectively, and improve the angle range dramatically.