摘要绿地对我们极为重要，从古至今，绿地在我们的生活中均起到无法取代的作 用。社会发展到今天，人们越来越意识到绿地的重要性。2016 年 1 月，徐州被 评为“国家生态园林城市”，且徐州市区有 177 个 5000 平米以上的公园，全部 公园均不收取任何费用。公园绿地是城市中向公众开放的、以游憩为主要功能， 同时兼有健全生态、美化景观、防灾减灾等综合作用[1]。

本文以 2016 年 4 月的 GF-1 影像作为主要信息源，以徐州市公园为研究区域， 利用 ENVI、Arcgis 等相关软件，对原始数据 GF-1 影像进行了预处理，然后采用 了监督分类和非监督分类的方法对徐州市区公园的绿地信息进行了提取，通过对 比分析，最终采用了基于最大似然法的遥感影像监督分类，并参考天地图徐州网 站上的航片和百度地图进行目视解译，得到了最终的分类结果图。 本文主要结论：69972

(1) 本论文使用的分类方法是监督分类和非监督分类，对徐州市区 2016 年 4 月 的高分辨率 GF-1 影像进行了绿地信息提取，通过比对分析，得出基于最大 似然法的遥感影像监督分类是最有效的分类方法，总体分类精度达到 96.1%。

(2) 以绿地公园边界为线，分别设置了不同的缓冲半径（500m、1000m），并判 断出不在缓冲区域范围内的地方，得出了缓冲半径为 1000m 的缓冲区占徐州 市区面积的 39.4%，缓冲半径为 500m 的缓冲区占徐州市区面积 19.7%。 该论文有图 12 幅，表 5 个，参考文献 21 篇。

毕业论文关键词：GF-1 城市绿地 分类 缓冲区

Extraction and analysis of park green space in Xuzhou city based on high spatial resolution remote sensing image

Abstrcat Green is very important to us, since ancient times, green space plays a irreplaceable role in our life. Society continues to develop, people become more and more aware of the importance of green space.In January 2016, Xuzhou was named "National Ecological Garden City", and Xuzhou city has 5000 more than 177 square meters Park, all the parks are free to the public, the real do is also green in the people.Park green space is open to the public in the city, the main function of the recreation, not only has a certain recreational facilities and service facilities, but also has a sound ecological, landscaping, disaster prevention and mitigation, and other integrated functions.

This paper takes the GF-1 image of April 2016 as the main information source, taking Xuzhou City Park as the research area,using Arcgis, ENVI and other related software to deal with the original data GF-1 image ,then it uses the method of supervised classification and unsupervised classification to extract the green space information of Xuzhou urban park, and finally uses the method of remote sensing image classification based on the maximum likelihood method.And refer to the map

of Xuzhou on the day of the aerial film and Baidu map for visual interpretationand the final classification result is obtained.

The main conclusions of this paper:

(1) Using supervised classification, unsupervised classification ,high resolution GF-1 image of Xuzhou city in April 2016 was extracted from the green space information.Through comparison and analysis, it is concluded that remote

sensing image supervised classification based on maximum likelihood method is

the most effective classification method, and the overall classification accuracy is 99.3610%.

(2) Taking the edge of the Green Park as the line, set the buffer radius of 500m,

1000m buffer, and determine the area not in the Forest Park area.Finally, it is concluded that the buffer radius of 1000m is 39.4% of the area of Xuzhou, and the buffer radius is 500m, which occupies 19.7% of Xuzhou urban area.