论文部分内容阅读
为了充分应用航天遥感资料解决森林自动分类成图,正确估算森林面积和蓄积,提供森林经营管理和森林生态环境研究的基本资料,五年来南京林业大学对课题进行了全面的探索研究,解决技术关键问题,已获得令人满意的成果。近年来随着计算机科学和航天遥感技术的不断发展,对卫星数据的计算机分析和自动识别有了一定的开发深度和广度;国内外利用卫星数据进行森林图像自动分类的精度,五年来徘徊于75～80％之间(指分类结果与实际测量结果的对比精度,不是指计算机统计计算的精度)。这样的精度尚不能满足林业生产上的要求,但是计算机图像分析处理是未来的森林调查和资源动态监测的方向。为了提高图像分类精度,在理论上进行一系列探讨,获得以下几方面的研究成果: 1.利用陆地卫星图像的光谱信息和空间结构信
In order to fully apply the aerospace remote sensing data to solve the problem of forest automatic classification into maps, correct estimation of forest area and accumulation, and provide the basic information of forest management and forest ecological environment research, Nanjing Forestry University conducted a comprehensive exploration and research on the issues over the past five years The problem has yielded satisfactory results. In recent years, with the continuous development of computer science and space remote sensing technology, the depth and breadth of computer analysis and automatic identification of satellite data have been developed. The accuracy of automatic classification of forest images by satellite data at home and abroad has hovered 75 ~ 80% (referring to the accuracy of the classification results and the actual measurement results, not referring to the accuracy of computer statistics). Such accuracy still can not meet the requirements of forestry production, but computer image analysis and processing is the future direction of forest investigation and resource dynamic monitoring. In order to improve the accuracy of image classification, a series of theoretical discussions are made to obtain the following findings: 1. Using the spectral information and the spatial structure of terrestrial satellite images