An Ego-Motion Estimation Method Using Millimeter-Wave Radar in 3D Scene Reconstruction

Ego-motion estimation is important in unknown scenario reconstruction and detections, which is critical to implementing unmanned and intelligent autonomous driving. Compared to optical devices, such as camera and Lidar, millimeter-wave radar is independent of optical conditions. Based on millimeter-wave point cloud imaging technique and target random scattering characteristics, an ego-motion estimation method using millimeter-wave radar is proposed in this paper, which can be applied to 3D scenario reconstruction in the closed environment. To estimate the motion parameters of the platform, linear fitting, key points extraction and registration are introduced. The proposed method is verified by experiments in real scenes using a single-chip millimeter-wave radar and TurtleBot 2 platform.