Aquatic vegetation remote-sensing classification threshold calculating method based on spectrum index ordering method

Abstract

The invention provides an aquatic vegetation remote-sensing classification threshold calculating method based on a spectrum index ordering method. Pixel spectrum indexes of the same natural object incertain area in two images are ordered, a spectrum index related linear model is obtained by fitting, classification thresholds of the image spectrum indexes with synchronous actual measurement samplepoints are substituted into the linear model, and the classification threshold of the spectrum index without synchronous actual measurement sample points can be calculated. The method can be used todetermine the classification threshold of the spectrum index during remote-sensing classification of aquatic vegetation without a lot of images, especially historical images, of synchronous actual measurement point, support is provided for reconstruction of spatial distribution data of the aquatic vegetation of shallow-water lakes, especially shallow-water lakes with eutrophication, and ecologicalrestoration of the lakes and salvage of aquatic plants can be guided.

Description

technical field [0001] The invention belongs to the field of remote sensing applications, and relates to a threshold value calculation method for aquatic vegetation remote sensing classification based on a spectral index sorting method. Background technique [0002] Aquatic plants are an important part of the lake ecosystem, as well as an important regulator and indicator of the lake ecosystem. The real-time and rapid monitoring of aquatic plants has guiding significance for the evaluation of the ecological function of lake ecosystems, and at the same time, it has important practical significance for the restoration of turbid lake ecosystems and the salvage of aquatic plants. [0003] Satellite remote sensing technology has the characteristics of fast, real-time, large-scale and periodicity, and is an effective means for monitoring the temporal and spatial distribution of aquatic vegetation in shallow lakes, and some related researches have been carried out. Most of the cur...

AI Technical Summary

Problems solved by technology

Most of the current aquatic vegetation remote sensing monitoring methods are decision tree classification methods. In the decision tree, the classification spectral index used to extract the tree nodes is the same, but the classification threshold will vary with the image. A large number of real-time sample points for training and acquisition of image synchronization, but for those images without a large number of synchronized real-time sample points, especially historical images, how to determine the classification threshold of the spectral index in the decision tree, there is no relevant report in the prior art

Images