期刊名称：Advance Journal of Food Science and Technology
印刷版ISSN：2042-4868
电子版ISSN：2042-4876
出版年度：2015
卷号：9
期号：1
页码：19-27
DOI：10.19026/ajfst.9.1927
出版社：MAXWELL Science Publication
摘要：Landsat TM and SPOT 5 data are the most popular remote sense data in forest resource monitoring, though, both have advantages and disadvantages. Landsat TM images contain large scale, but with low accuracy; while high resolution Spot 5 images have high accuracy and small scale. We could combine the merits of accuracy and scale by scaling method in the study of forest vegetation cover. Hence, we use Landsat TM and SPOT 5 data to monitor the vegetation cover of three forest types (coniferous forest, broad-leaved forest mixed coniferous and broad-leaved forest) locating around Miyun reservoir in Miyun County in Beijing. These two different resolution images were scaled up by using mathematical statistics and modified the vegetation cover extracted from Landsat TM image by using scale conversion model. Upon testing, SPOT 5 images were used to resolve elements of Landsat TM images. We obtain statistic model from statistic results and information extracted from Landsat TM image. Statistic model may efficiently improve the accuracy of vegetation cover in Landsat TM image. In conclusion, basing on SPOT 5 and Landsat TM data, the scale conversion model has better performance. Combining element resolution and statistic model, we could apply high spatial resolution image to improve information accuracy of low spatial resolution image.
Loading...
