南京市六合区基于0.5m分辨率影像的林地落界及分析

林地落界是为了掌握林地的现有状况,为林地的保护利用规划提供重要的基础数据。高空间分辨率的遥感影像很大幅度地提高了室内林地落界的精度。该文是以高空间分辨率遥感影像为底图,在目视解译的方法上,结合二类调查基础数据,通过建立高清影像的判读解译标志,提取六合区林地,得到南京市六合区林地分布图,与基于中分辨率影像,采用面向对象的方法提取的林地分布图比较,并选取六合区龙袍镇为对象,分析比较了本次林地落界数据和二类调查数据的林地变化率和林地小班的差值变化。结果表明:0.5 m分辨率的影像能够达到将林地落实到田间地块的目的;六合区龙袍镇林地落界由林地到非林地变化率是15%,林地小班差值D的均值为0.24,说明利用高空间分辨率的遥感影像作为目视解译的底图,在林地地类识别和落界精度上都有较大的提高。

Extraction and analysis of the forest land boundary in LuHe by high spatial resolution image

Extracting the boundary of forest land is available to get forest land condition, and it is important for forest protection and utilization planning. Visual interpretation is an effective method of intormation extraction. High spatial resolution remote sensing image was used based on the second type of forest resources survey, visual interpretation method, interpretation marks of DoM images and the extracted boundary to get the distribution of the forest land of Luhe District. Furthermore, the data of the boundary of forest land was compared with the data of the second type of forest resources survey. The results showed that there were 121.57 hm2 forest land in the town of Longpao of Luhe District, the rate was 15% from forest land to no,l-forest land, with the average of D - value of 0.24. By using high spatial resolution remote sensing image, the precision and recognition was obviously improved to fall the boundary of forest land.