In the evaluation of high-precision flat interferometer systems, the Instrument Transfer Function (ITF) has become an essential index for assessing interferometer performance. It accurately characterizes the resolution capability of optical systems in measuring profiles at different spatial frequencies. However, strict experimental environment requirements and interference system noise can significantly distort ITF test results. To accurately characterize interferometer performance, this paper combines simulation analysis with practical experiments, introducing simulated interferograms into varying levels of Gaussian random noise. We explore the impact of noise on ITF test results, propose the ITF_RMS value as an evaluation index for measurement error, and analyze the noise tolerance of the interferometer system. Several ITF test optimization methods are presented based on this evaluation index. An experimental system is built for analysis and validation. The final results indicate that in the unoptimized ITF testing process, the ITF_RMS value of the interferometer system is 0.111 2. After optimization, with the same noise level, ITF_RMS values reduce to 0.069 3, 0.036 7, and 0.057 9, respectively. These results demonstrate the feasibility of the proposed optimization scheme, effectively suppressing noise interference during ITF testing of the interferometer system.