The temperature induced drift model for FOG is analyzed and established. But data noise would weaken or deteriorate the effect of this compensation model. To solve this problem, digital low pass filter and wavelet threshold value denoising method are used respectively to reduce the data noise. It is found that the wavelet denoising method is more effective than that of the digital low pass filter through the simulation and analysis. Comparison of the compensation results before and after denoising further verified the effectiveness of the wavelet threshold value denoising method for improving the compensation effect of the temperature induced drift model.