去噪是激光回波处理的重要一环,好的去噪算法可以有效地去除噪声带来的影响,保留有效信号,减少失真。经验模式分解以信号的局部特征时间尺度为依据,将时域信号分解为若干个不同频带尺度的时域信号分量,传统的去噪实现一般是直接舍弃高频分量。然而,由于在某些高频分量中,存在着有效信号和噪声的混叠,如果直接舍弃,会导致有效信号的损失。由此考虑并提出了一种基于经验模式分解及分数阶傅里叶变换的激光雷达回波去噪算法,用分数阶傅里叶变换处理特定的本征模态函数,以此来提高去噪的有效性及准确性。 Denoising is an important part of laser echo processing. A good denoising algorithm can effectively remove the influence of noise, retain effective signal and reduce distortion. Empirical mode decomposition decomposes the time-domain signal into several time-domain signal components of different frequency bands based on the local feature time scale of the signal. The traditional denoising implementation generally abandons the high-frequency components directly. However, because of the aliasing of effective signal and noise in some high-frequency components, the direct loss will lead to the loss of effective signal. Therefore, a de-noising algorithm for lidar based on empirical mode decomposition and fractional Fourier transform is proposed and proposed. Fractional-order Fourier transform is used to process specific intrinsic mode functions to improve de-noising Effectiveness and accuracy