FMCW LiDAR Range Imaging and Recognition System Utilizing FPGA‐Based Edge‐Aware Processing

作者：Shangwen Rong, Mengxi Zhou, Wenxuan Ma, Yu Zhang, Yuehua Chen, Zhenxing Sun, S X Li, Jian Xin, Xiangfei Chen · 发表于：Microwave and Optical Technology Letters · 年份：2026 · DOI：10.1002/mop.70701 · 研究领域：Advanced Optical Sensing Technologies、Robotics and Sensor-Based Localization、Indoor and Outdoor Localization Technologies

ABSTRACT We present an FMCW LiDAR range imaging and recognition system utilizing FPGA‐Based edge‐aware processing. To mitigate the frequency sweep nonlinearity of the directly modulated laser, a simple correction method is first applied to improve chirp linearity. Based on the corrected signal, range estimation is performed via FFT‐based beat‐note peak detection, achieving reliable distance retrieval with a maximum absolute error below 25 mm in near‐range measurements. The system further reconstructs range images through point‐by‐point scanning and implements a lightweight FPGA‐based preprocessing pipeline, including median filtering, binarization, morphological operations, and Sobel edge extraction. The proposed processing effectively suppresses background noise, enhances contour continuity, and reduces redundant data, achieving a relative PSNR of 16.65 dB and a CNR of 12.70, while reducing the data volume by 87.5%. To evaluate the usability of the processed range images, a lightweight YOLO‐based detector is employed as a representative downstream task and achieves a mean average precision (mAP@0.5) of 0.995 on a custom data set. These results demonstrate an effective and bandwidth‐efficient FPGA‐based solution for FMCW LiDAR range imaging and feature extraction.