基于卫星IKONOS影像的黄河三角洲人工刺槐林健康状况分类  

Health Situation Classification of Artificial Robinia Pseudoacaci Forests from Yellow River Delta Based on IKONOS

作　　者：张珍珍[1] 王红[1] 

机构地区：[1]河海大学地球科学与工程学院,南京210098

出　　处：《科学技术与工程》2014年第33期73-79,共7页Science Technology and Engineering

摘　　要：对刺槐林生长状况的准确分类制图,是刺槐林生长动态监测和枯梢退化原因分析的基础,对指导防护林建设、更新以及研究区其它树种的植被重建具有重要的意义。研究以黄河三角洲地区刺槐林为对象,采用2013年6月9日IKONOS影像为数据源,利用决策树分类器对归一化植被指数和土壤调节植被指数定义一组规则提取出刺槐林地分布范围,制作刺槐林地掩膜,创建灰度共生矩阵来区分健康树冠和林下禾草。结合实地样方调查信息,选取感兴趣区进行监督分类。刺槐林的三个健康度由原位的5个树冠条件指标决定。分类结果表明,刺槐林健康状况分类结果与实地样方数据具有较好的一致性,利用混合矩阵进行精度评价,总精度达到84.315 9%,Kappa系数为0.765 2。Accurate classified mapping of the growth condition of acacia forest is the growth dynamic monitoring and withered tip degradation cause analysis of acacia forest. It also has vital significance of guiding the construction and renewal of protection forest,and revegetation of other trees in the study area. This article is based on the study in Robinia Pseudoacaci forest of Yellow River delta area. It uses IKONOS image on June 9th,2013 as a data source,and uses decision tree classifier to define a set of rules of Normalized Difference Vegetation Indexand Modified Soil adjusted Vegetation Indexand extracts Robinia Pseudoacaci forest land distribution range,then make Robinia Pseudoacaci forest mask and establish gray level co-occurrence matrix to differentiate healthy crown and forests grasses,and select region of interest to supervise and classify combing the survey information of field samples. The three health degree of Robinia Pseudoacaci forest are determined by Canopy Condition Indicators. Classification results indicate that,the classification results of health condition of Robinia Pseudoacaci forest correspond with the field sample data. Hybrid matrix accuracy assessment makes the overall accuracy reach 84. 315 9% and Kappa coefficient 0. 765 2.

关 键 词：刺槐林 IKONOS影像 灰度共生矩阵 感兴趣区 监督分类 

分 类 号：TP79[自动化与计算机技术—检测技术与自动化装置]