中文摘要：

机载小光斑LiDAR是近年来遥感领域的新兴技术，它对回波脉冲进行全数字化的记录，通过分解返回波形能够得到更加丰富的信息。针对传统LM波形分解算法容易陷入局部最优的问题，引入改进LM模型，采用信赖域技巧使得参数解全局收敛。实验证明，该方法相比传统LM能够得到更加可靠的波形分解结果，并且对城区、郊区和山地区域适应性好，具有较强实用价值。

英文摘要：

Airborne small foot-print LiDAR is a rising technique in remote sensing field. It can digitize all the back-scattering pulses. More information of targets can be obtained by waveform decomposition, which is the most impor-tant step of small foot-print LiDAR data processing. Aiming to solve local convergence problem, modified LM is in-troduced and trust region skill is used to make the parameter solutions global convergent. Experimental results prove that modified LM is more reliable and practical, and adaptive to urban, suburban and mountainous regions in comparison with traditional LM.