土地利用的遥感信息变化提取研究

利用卫星遥感影像进行土地利用类型分类和动态变化监测是遥感应用中的一个重要课题。本文选择不同时相的ETM 和SPOT-5卫星遥感影像数据，对两期影像进行监督分类，快速提取不同时期的土地利用数据；然后进行动态变化监测，获得土地利用情况的变化特征和信息；最后对其分类精度进行评价分析。研究表明，两期影像中耕地、居民用地和未利用地这三个类别的变化面积较大；ETM 影像进行监督分类的精度为90.1692%，Kappa系数值为0.8268；SPOT-5影像进行监督分类的精度为95.1477%，Kappa系数值为0.9361。由于SPOT-5影像的分辨率较高，分类效果更优于ETM 影像，更能准确的反映土地类型的信息和特征。

Research on the land use of remote sensing information change extraction

By making use of satellite remote sensing image of types of land use classification and dynamic changes in the applications of remote sensing monitoring is an important issue. At present, with the tense situation of land is getting worse and land use pattern changes constantly, so using remote sensing technology to land in the use of resources and planning has important social value and practical significance. This paper, based on different periods of ETM and SPOT-5 remote sensing satellite image data, carried out supervised classification, land use data of two periods were extracted, and dynamic monitoring of the study area land use were completed, then summarized the information and characteristics of the land use changes. Finally, the classification precision evaluation was analysed. Establish by establish Interpretion marks and analyzing the study area we can get the data of land use type change and extract rapidly the land use type change information.The study area Jinxi City located in the western Liaoning Province. The altitude is generally 20 ~ 500m. The mountain toward the north-east, the terrain is generally high in the northwest to southeast. the area has many types of vegetation including the the type of the forest, bush, farmland, alkaline land.According to the actual situation, the area was divided into forest land, cultivated land, residents land and unused land (the area to bare land primarily), water (including rivers and reservoirs, sea and beach), and other categories. Research showed that the three categories of cultivated land, residents land and unused land change largely. The supervised classification precision of ETM image is 90.1692%, and Kappa value is 0.8268; for SPOT-5, the image classification precision is 95.1477%, and Kappa value is 0.9361. Due to the resolution of the image SPOT-5 is higher, so the classification effect is better than ETM image, which can more accurately reflects the land types of information and characteristics.