Abstract: There are many factors that affect the measurement accuracy of binocular vision system. Currently, the influence of system structure parameters on the measurement accuracy mainly includes the angle between optical axis and baseline, baseline distance, horizontal viewing angle, object distance and lens focal length. Since the aperture size directly affect the imaging resolution, it is the core factor that determines the accuracy of binocular vision measurement. Consequently, according to the incoherent imaging theory, the binocular imaging process is simulated and tested. Moreover, Speeded Up Robust Features algorithm is adopted to extract and match the features of the image pairs to obtain their parallax values. The parallax root mean square error is calculated to represent the systematic errors. The results show that the system error decreases with the increase of lens aperture, and approaches saturation. This research can provide theoretical and experimental basis for the selection of system parameters and aperture size during the design of the binocular system.