A Real-Time Point Cloud Compensation System for High-Frequency Piezoelectric MEMS Lidar Based on FPGA

Environmental interference lowers point cloud quality. Atmospheric attenuation is a common problem. It severely limits LiDAR performance. Good solutions are still few. This is especially true for high-frame-rate uses. This work introduces a Micro-Electro-Mechanical Systems (MEMS) LiDAR system. A high-frequency large-aperture piezoelectric micromirror is used. A cubic spline interpolation method (SC-ITP) is applied for compensation. This design supports high-speed scanning. Point cloud integrity is kept even with disruptions. The system runs on a heterogeneous FPGA. This setup allows smooth cooperation between hardware and algorithm. Tests show the largest ranging error is under 19 % from 50 to 600 cm. Past 450 cm, the error nearly disappears. The compensation method fully restores lost points. This works for both 1D and 2D scans. The quality of reconstructed data is greatly enhanced. The system is well suited for high-speed lidar tasks. It provides stable and accurate point cloud collection in challenging settings.