机载激光雷达和高光谱技术的遥感监测数据分类

Remote sensing data classification of airborne lidar and hyperspectral technology

针对随机森林算法、支持向量机以及线性判别分析3种分类方法分类准确率较低、性能不佳的问题,基于机载激光雷达和高光谱技术提出一种新的遥感监测数据分类方法。分析机载激光雷达和高光谱遥感的工作原理,依据二者的工作原理提取遥感监测数据。为了提升数据分类的准确性,需要进行数据预处理,具体包括噪声点剔除和平滑校正处理,并进行数据配准。在此基础上,采用K最近邻搜索算法提取遥感监测数据特征,最终运用决策树算法中的C4.5算法构建分类器,实现遥感监测数据的分类。实验结果表明:通过混淆矩阵得到所设计方法的分类准确性达到了95以上,分类结果优于传统分类方法,能准确识别目标,证明了方法有效性和可行性。

Aiming at the problems of low classification accuracy and poor performance of three classification methods,random forest algorithm,support vector machine and linear discriminant analysis,a new classification method of remote sensing monitoring data is proposed based on airborne lidar and hyperspectral technology.The working principle of airborne lidar and hyperspectral remote sensing is analyzed,and the remote sensing monitoring data is extracted according to the working principle of both.In order to improve the accuracy of data classification,the data preprocessing is needed,including noise point removal and smoothing correction,and data registration.On these basis,K nearest neighbor search algorithm is used to extract the features of remote sensing monitoring data.Finally,C4.5 algorithm of decision tree algorithm is used to build a classifier to realize the classification of remote sensing monitoring data.The experimental results show that the accuracy of designed method is above 95%,and the classification result is better than that of traditional classification method.It can accurately identify the target,which proves the validity and feasibility of the method.