Land automatic classification method based on multisource and multi-temporal satellite image data

Abstract

The invention relates to the field of ecological environment, and discloses a land automatic classification method based on multisource and multi-temporal satellite image data. The method comprises the following steps: A, multi-temporal data from the satellite are used for calculating a normalized differential vegetation index (NDVI) value of the land; B, according to the NDVI value, land classification baseline data are predicted to divide and classify the land, and a land initial classification table is acquired; and C, a gravity model and the land initial classification table are used for precisely classifying the land, and a land classification result is acquired. As the multi-temporal data serving as a data source has an exact time point, when the land automatic classification method in the embodiment of the invention is used, land automatic classification complete data in a specific temporal scale are acquired through processing the multi-temporal data at different time points, and evolution analysis of the land is facilitated.

Description

technical field [0001] The invention relates to the field of ecological environment, in particular to an automatic land classification method based on multi-source multi-temporal satellite image data. Background technique [0002] Land use and land cover are not only the way humans use natural space in a certain region, but also involve changes in the natural ecological environment on a global scale. Changes in land use patterns have profound impacts on climate, air, biodiversity, soil and other environmental factors. In the change of land use and land cover, human activities are one of the main influencing factors, especially in today's rapid development of social civilization, human social and economic activities play a leading role around towns and even remote suburbs far away from towns. [0003] The current land use and land cover research mainly includes three aspects: the driving factors of land use and land cover change, the dynamic analysis of land use and land cov...

AI Technical Summary

Problems solved by technology

However, the data in it are only three years of land ecological classification data in 2000, 2005 and 2010, and the longitudinal time is not complete

The complete data of land classification is of great significance for the analysis of land evolution. Due to the limited time of the existing land classification results data, it is impossible to obtain complete data in time, which affects the analysis of land evolution.

[0005] In addition, there are few automatic classification tools for land classification used in the study of ecological environment, especially the spatial relationship of ecosystems, and the accuracy is not high, which hinders the improvement of the accuracy of land evolution analysis.

Images