论文部分内容阅读
选取1974～2007年期间的5期多源遥感影像数据作为数据源,在解决多源遥感影像(MSS、TM、ETM+和CBERS影像)和多时相(1月、6～9月)影像数据的遥感分类问题上,以模糊物元理论来解释多数据源和多时相数据之间的关系,设计关联函数概率转化的多源遥感数据分类物元模型,利用该模型得到了若尔盖县湿地区1974～2007年期间的地物类型分布,并进行了变化检测和概率修正算法处理,揭示了1974～2007年若尔盖县湿地变化特征。2007年,若尔盖县各类型湿地的面积占湿地总面积的比例分别为河流和湖泊1.37%、泥炭沼泽4.23%、沼泽化草甸9.91%和湿草甸6.04%。1974～2007年期间,若尔盖县的河流和湖泊、泥炭沼泽、沼泽化草甸和湿草甸的面积在减少,以旱生植物为主的中、低覆盖度草地、居民点、建筑用地和沙地面积在增加。泥炭沼泽主要向沼泽化草甸、湿草甸演替。人为因素则是在较短时间尺度上影响湿地变化的主要驱动力。不确定性问题仍是遥感分类与变化检测存在的主要问题。
In this study, we selected 5 multi-source remote sensing image data from 1974 to 2007 as the data source, and solved the remote sensing data of multi-source remote sensing images (MSS, TM, ETM + and CBERS images) and multi-temporal (January to June ~ September) On the classification problem, the relationship between multiple data sources and multi-temporal data is explained by the fuzzy matter-element theory. The multi-source remote sensing data classification matter-element model of probability transformation of correlation function is designed. By using this model, Year type distribution of ground objects, and change detection and probability correction algorithm processing, revealing the characteristics of wetlands in Zoige County from 1974 to 2007. In 2007, the proportions of all types of wetlands in Zoige County accounted for 1.37% of the total area of wetlands, peat bogs 4.23%, swampy meadow 9.91% and wet meadow 6.04% respectively. Between 1974 and 2007, the area of rivers and lakes, peat swamps, swampy meadows and wet meadows in Ruoergai County decreased. The middle and low coverage grassland, settlements, construction land and sandy land The area is increasing. Peat swamps mainly to the swamp meadow, wet meadow succession. Human factors are the main driving forces that affect wetland change on a shorter time scale. Uncertainty is still the main problem of remote sensing classification and change detection.