Frequency modulated continuous wave (FMCW) lidar performs measurements through coherent detection and offers advantages such as high-ranging accuracy, a wide range, and the absence of a need for cooperative targets. The linearity of the laser’s frequency modulation directly influences the ranging accuracy and resolution of the FMCW lidar. Traditional software-based resampling methods require substantial memory space and extensive data processing time to achieve high-precision measurements, which poses a disadvantage for real-time applications. In this paper, we employ the equal optical frequency interval resampling correction technique to resample the target signal using an auxiliary signal, aiming to eliminate the influence of laser frequency modulation nonlinearity on the lidar. To meet the real-time and high-precision measurement requirements, and to enhance the system’s processing speed, a resampling FMCW lidar system based on field programmable gate array (FPGA) technology is designed. This approach is expected to provide a novel concept for achieving high-precision, low-cost three-dimensional (3D) imaging.

The laser’s frequency modulation nonlinearity is corrected using the equal optical frequency interval resampling method, with the entire process implemented based on FPGA. The correction system adopts a double optical path structure, comprising two parallel Mach-Zehnder interference optical paths. One path serves as the measuring optical path, while the other functions as the auxiliary interference optical path. The FPGA uses a 5 kHz triangular wave generated by a digital-to-analog converter (DAC) to modulate the laser. After modulation, the beat frequency signals generated by the measuring and auxiliary optical paths are detected by a balanced detector, collected by an analog-to-digital converter (ADC), and transmitted to the FPGA for synchronous processing. In the subsequent signal processing stage, the FPGA performs targeted digital filtering on the two beat signals. By identifying the characteristic points of the auxiliary signal and converting it into a square wave to serve as the acquisition clock for the measurement signal, the measurement signal undergoes period positioning, resampling, and uniform output processing. Finally, a spectrum free of nonlinearity is obtained through the fast Fourier transform (FFT).

In this paper, we focus on the influence of resampling technology on the spectral characteristics and measurement accuracy of the target signal. Before resampling, the signal measured at a 5 m distance exhibits spectrum broadening, a low signal-to-noise ratio, and a large measurement error. After resampling, the signal-to-noise ratio is greatly enhanced, the spectral width is significantly compressed, and the measurement accuracy improves (Fig. 8). In the static distance measurement experiment within the 1?5 m range, the uncorrected system’s maximum error reaches as high as 257.8 m. After resampling correction, the maximum error is reduced to 3.9 mm, improving the ranging accuracy by a factor of 66 compared to the original system (Fig. 10). In the imaging experiment targeting a scene composed of two water cups at 1.6 m, the FPGA-based signal processing can process a 256 pixel×256 pixel image in approximately 19 s, clearly identifying the target outline (Fig. 11). In the dynamic measurement experiment, the system measures velocities in the ranges of 0.1?0.4 m·s^-1 before correction and 0.05?0.4 m·s^-1 after correction. The results demonstrate that the maximum measurement error for speed before resampling correction is 0.044 m·s^-1, which is reduced to 0.012 m·s^-1 after correction, improving accuracy by a factor of 3.7 (Fig. 13). Overall, the resampling technique effectively enhances the spectral characteristics of the target signal and improves measurement accuracy.

In this paper, the equal optical frequency interval resampling method successfully eliminates the influence of laser frequency modulation nonlinearity in FMCW lidar measurement. Compared to the limitations of software-based resampling methods in real-time, high-precision measurement, the data acquisition and signal processing system based on FPGA demonstrates excellent performance. Within the 1?5 m measurement range, the system’s maximum ranging error is only 3.9 mm, and the maximum relative error for speed measurement is 3.4%. In addition, the system achieves clear reproduction of an object at 1.6 m through target scanning and imaging. The experimental results verify the system’s real-time and effectiveness, providing a feasible solution for high-precision and low-cost 3D imaging. This research outcome holds broad application potential in fields such as autonomous driving and industrial measurement and is expected to drive the development and wide use of lidar technology in related fields. It also lays a solid foundation for future research and technological improvements.