As an important component of Autonomous Driving (AD) system,High-Definition Map (HDM) can provide highly accurate prior data of lane lines and road auxiliary facilities for AD system.The reliable evaluation of HDM accuracy is extremely necessary,but has been troubled by the evaluation methods used in mapping.Here,a method based on point set alignment and resampling is proposed to evaluate the relative accuracy of lane lines,and experiments are conducted based on relevant HDM data.First,the points on the verification curve are fitted and sampled,and the aligned point pairs are registered and then resampled,based on which the relative accuracy is calculated.The results showed that the relative limit errors of all the 4 groups of lane lines were verified to be less than 20 cm,meeting the relative accuracy requirements,of which the first group has the relative limit error of 15.9 cm.It can be concluded that the proposed method is more accurate and reliable in accuracy evaluation of HDM than traditional methods.