基于3S技术的棉花面积估测方法研究

基于3S(遥感、地理信息系统和全球定位系统)技术的支持,以SPOT影像为信息源,结合GPS地面样方调查、地面定位数据分析以及其他资料进行试验区作物种植面积估测的综合研究。简述了样方的采集,影像的选取、合成和预处理,以及用GPS数据进行影像几何纠正的基本原理及其优点,重点研究了遥感分类中的非监督分类、监督分类中的最小距离法、最大似然法和马氏距离法4种方法及其在作物种植面积估测上的应用和精度。在样点调查与分类的基础上,通过对不同样区、不同时相遥感数据的解译、野外地面调查样方数据的统计,计算样区内作物种植面积,并以泽普县为例,利用抽样内推外延的方法,计算该县的棉花种植面积,作为研究成果推广的实践基础。

Study on 3S-Based Estimation of Cotton-growing Area

In this paper,an integrated study on the estimation of cotton-growing area in the experimental area,Zepu County,Xinjiang,is carried out based on 3S(remote sensing,geographical information system and global position system) technology,SPOT images,GPS plot investigation,ground position data,etc.Firstly,the acquisition of plot data,selection,composition and preprocessing of the images, and principles and advantages of geometric correction of the images are lucubrated using the GPS data,the unsupervised classification and four supervised classifications,such as the minimum distance,maximum likelihood and Mahalanobis distance,are emphatically researched,and the principles and precisions of the four classifications are analyzed.Finally,the cotton-growing area in Zepu County is estimated based on the plot investigation and classification results,interpreted results of the RS data in different plots and different phases,and statistical results of plot data by applying the inside-deducing and outside-extension.It is considered that the optimal phenological phases of deriving the data of cotton-growing area from the images are mid-June or late-June and mid-September or late-September.The study results reveal that the classification precision of the minimum distance is the highest.