Extraction of preview elevation of road based on 3D sensor

With preview road elevation information, we could prepare the vehicle for preview road inputs and allow the vehicle body to remain as smooth as possible to improve ride comfort. But few people gave a detailed description of it. In this paper, we proposed an approach using 3D sensor, IMU and GPS to get accurate preview three dimensional points cloud data of the road, attitude angle and position information for extracting the elevation information of road in advance. GPS/INS loosely-coupled integrated navigation and Kalman Filter were used to get accurate attitude angle and position information, which could ensure the precision of data fusion. Then the points cloud data was processed by VoxelGrid filtering algorithm and PassThrough filtering algorithm to get the useful data, which could reduce the calculation burden. In the end, these road elevation information was interpolated and fitted by cubic spline interpolation, which is a computationally fast and accurate method.