全波形激光雷达目标特征多参数反演理论及其方法研究

'Full-waveform' LiDAR is a novel remote sensing technology, which has appeared with the ability to record and analyze the complex waveform of the backscattered signal echo. Thus, it can improve the credibility; measurement precision and resolution of remote sensing by detecting the 'Full-waveform' LiDAR waveforms. By pursuing on the characteristics and variations of terrain, it will become an important way to prevent and monitor the kinds of natural disasters, such as earthquake or landslide. Due to lack of the systemic model and algorithms for 'Full-waveform' LiDAR waveform up to now, the useful information of the detected targets can't be extracted sufficiently.In order to judge the characteristics of the detected targets, such as slope roughness and waviness.The relationship between multi-characteristics of targets and multi-parameters of 'Full-waveform' LiDAR waveform can be accomplished by modeling and simulation. The inversion model can be solved by using the multivariate parametric equation. Based on the traditional waveform processing methods, such as the convolution approach for range determination and target discrimination and the Gaussian decomposition and calibration method, a novel time-frequency analysis method is advanced to decomposed the waveforms multi-parameters and inverse the targets multi-characteristics by the fractional Fourier transform (FrFT).It can provide the better signal-to-noise filter in fractional Fourier, thus the characteristics of the detected targets can be deduced from the response function. For validate the methods mentioned above, a semi-physical simulation experimentation of 'Full-waveform' LiDAR will be established. It will measure the full-waveforms constantly and record the LiDAR data effectively. The data obtained from semi-physical simulation experimentation will verify and modify the multi-parameter inversion method for 'Full-waveform' LiDAR. This work can provide scientific basis for novel 'Full-waveform' LiDAR independently researched and developed by China.

全波形激光雷达是一种新概念遥感雷达手段，可对整个激光回波进行完整的记录和分析，能提高遥感激光雷达波形探测的可靠性、测量精度和分辨率。利用其研究地形地貌特征及其变化，对自然灾害防御具有重要的科学意义。针对目前全波形激光雷达回波参数解算算法不完善，目标信息提取不丰富的问题，需要建立全波形激光雷达回波的多参数与探测目标的多特征之间的映射关系,从而可靠地反演坡度、粗糙度、起伏度等目标特征。基于目标多元变参方程，建立目标特征多参数反演模型。为了更准确地获得回波参数，在传统傅里叶反卷积法和高斯分解法的基础上，引入分数阶傅里叶时频分析方法有效提取回波参数，在分数阶变换域获得更好的信噪分离效果。直接获取被测目标的特征响应函数，判断目标特征。搭建全波形激光雷达半物理仿真实验系统，针对目标模型记录全波形数据，验证目标特征多参数反演模型与方法。该研究成果为我国全波形激光雷达系统的自主研发提供理论依据和验证手段。

全波形激光雷达是一种新型主动式遥感技术，能够实现对目标散射回波波形的完整记录，通过对回波波形进行分析，可以提取丰富的目标特性。本课题从全波形激光雷达系统样机设计，目标后散射探测理论研究，全波形激光雷达数据处理与分类三个方面展开深入地理论研究与实验验证。主要创新性内容包括：.（1）搭建全波形激光雷达实验系统，可以为数据处理方法的验证提供实验数据；验证多种波形处理算法，提取有效回波参数。.（2）针对建筑类型目标建立描述目标特征参数（粗糙度和反射率）的BRDF模型；结合全波形激光雷达实验系统几何结构特征，建立激光雷达强度数据与BRDF之间映射关系，可以实现目标特征参数提取。.（3）面向星载全波形激光雷达数据（ICESat/GLAS）,利用机器学习方法（SVM和GP）实现对地观测目标的二分类和多分类研究，分类精度均达到了90%以上。.本课题搭建的全波形激光雷达实验系统为后续系统优化提供了必要的设计参数支持；提出的模型及数据处理方法可以扩展全波形激光雷达回波数据的有效利用，实现目标特征参数反演和目标分类。.

内容获取失败，请点击重试