Coastal wetland remote sensing classification method based on a layering strategy

Abstract

The invention relates to a coastal wetland remote sensing classification method based on a layering strategy. The method comprises the following steps: step 1, preprocessing remote sensing data; 2, carrying out sample selection on the vegetation index and the water body index obtained based on the remote sensing index and the brightness auxiliary data obtained based on tasseled cap transformation;Step 3, performing coarse classification on the multispectral image and the auxiliary data; 4, the cultivated land, the forest land and the town area are covered with a mask; Step 5, describing localtexture features of the image by using an LBP local binary pattern operator; Step 6, performing image segmentation on the water body and the wetland area by combining the spectral information and thespatial texture information; 7, obtaining vector data of the coastal wetland; And step 8, making a coastal wetland thematic map. The method has the advantages that time, manpower, material resourcesand financial resources can be saved, coastal wetland classification can be rapidly and accurately obtained, the classification result is high in precision, and obvious geographic significance is achieved.

Description

technical field [0001] The invention belongs to a coastal wetland remote sensing classification method, in particular to a coastal wetland remote sensing classification method based on a stratification strategy. Background technique [0002] Coastal wetland refers to the staggered transition zone between terrestrial ecosystems and marine ecosystems. According to the definition of the International Convention on Wetlands, the lower limit of coastal wetlands is 6 meters below sea level (it is customary to limit the lower limit to the outer edge of the growth area of ​​large algae), and the upper limit is freshwater or brackish water lakes connected to inland river basins above the tide line The swamp and the section of the river that cannot be reached by seawater upstream have high productivity and serve as an important habitat for various animals and plants, and have the function of maintaining ecological balance and environmental stability. Such as providing coastal protect...

AI Technical Summary

Problems solved by technology

[0005] With the gradual refinement of coastal wetland monitoring, the current research on coastal wetland classification has achieved fruitful results, but it also faces some problems: 1) Traditional classification methods based on remote sensing images, such as based on maximum likelihood method and decision tree Although the image classification of coastal wetlands can obtain the dynamic distribution information of coastal wetlands, "homogeneous and different spectrum" and "same spectrum and heterogeneous" are likely to cause misclassification of coastal wetland features and reduce classification accuracy; 2) due to the need for object-oriented methods Extract a variety of feature information from images for attribute calculation in the classification process. Too many features will easily lead to a decrease in classification accuracy

Images