Aiming at the problem of snowflake noise causing false detection of lidar,this paper proposes a comprehensive method that combines lidar reflection intensity and neighborhood search algorithm to achieve denoising processing of lidar point cloud in snowfall scenes.First,the reflection intensity distribution of lidar point cloud is counted by an adaptive method,and the dynamic boundary threshold of point cloud is calculated.Then,the noise caused by snowflake is filtered out by analyzing the reflection intensity of the point cloud data and the neighborhood search algorithm.This method is verified with real time lidar point cloud data in the park environments under snowy days.The experimental results show that this method has a better performance than the common methods,such as Dynamic Radius Outlier Removal Filter,in filtering effect and retaining the surrounding environment characteristics.It can promote the application in snowy weather of autonomous driving vehicles.