全局收敛LM算法优化的机载激光测深信号提取方法

在机载激光测深系统中,水面与水底回波信号的提取精度是影响系统测深能力的关键因素,然而传统的信号提取方法易受噪声影响,精度低,对测量环境的适应性差。针对上述问题,提出一种全局收敛LM(Levenberg Marquardt)算法优化的机载激光测深信号提取方法。首先通过模糊筛选得到较为可靠的初值;然后利用基于全局收敛LM的波形分解算法对波形进行分析建模,从而获取每个初值点的高斯分解参数;最后通过多条件筛选确定精确的水面、水底回波信号位置。利用实测数据和模拟数据分别进行实验,结果表明提出方法对回波信号具有较高的检测正确率,且稳定性较强。

Signal Extraction for Airborne LiDAR Bathymetry Based on Global Convergent Levenberg Marquardt

In airborne LiDAR bathymetry system, the extraction accuracy of the water surface and bottom are the key factors that affect the bathymetric ability of the system. Since that the traditional signal extraction method is quite sensitive to noise with low accuracy and poor adaptability to different environment, a gradual signal extraction method using waveform decomposition based on global convergent Levenberg Marquardt( LM) is introduced in this paper. In the first step of this method, a loose selection is implemented to calculate the comparatively reliable ini-tial values. For the selected initial values, the corresponding Gaussian decomposition parameters are then computed by waveform decomposition based on global convergent LM. A multi-condition filter is proposed to determine the exact reflectance signal positions of the water surface and bottom. Experimental results show that the proposed method has high detection accuracy and strong robustness.