A small number of test points are selected for the traditional HUD display accuracy test.If the test results are not out of tolerance,it is judged to be qualified,and it is impossible to prove whether it meets the overall requirements.Therefore,the scheme needs to be improved.A simple random sampling model is designed by using the inferential statistical theory in mathematical statistics and in combination with the characteristics of HUD field of view.A sampling test is conducted,and the overall requirements are inferred based on the sample test results.Firstly,a random sample model with uniform distribution is designed by using the simple random sampling theory.Then,the hypothesis test theory is adopted to study the decision-making process of display precision,and the relationship among display precision out of tolerance,the minimum sample size and the maximum allowable number out of tolerance is determined.Finally,a verification scheme is designed,and the test results prove the rationality of the improved scheme.The improved scheme can take into account the risks of both the manufacturer and the user.