Water body index variable coefficient-based rice sub-pixel recognition method

Abstract

The invention discloses a water body index variable coefficient-based rice sub-pixel recognition method. The method comprises the following steps of: S1, constructing a water body index, vegetation index and surface temperature time series data set; S2, getting rid of non-cultivated land pixels in a research area; S3, determining a period of a crop growth season according to a surface temperature; S4, calculating variable coefficients of water body indexes in each pixel growth season; S5, determining a relationship between water body index variable coefficients and rice planting area proportions; and S6, calculating a rice planting area proportion according to the variable coefficients of the water body indexes. According to the method, by utilizing the water requiring property of rice, the water body index variable coefficients are reduced along with the increase of the rice planting area proportions, and the water body index variable coefficients and the rice planting area proportions have prominent linear dependence. Through designing the variable coefficients of the water body indexes to recognize rice sub-pixels and calculating the planting area proportion, the method has the characteristics of being independent of prior knowledge, good in robustness, high in classification precision and strong in knowledge ability.

Description

technical field [0001] The invention belongs to the technical field of agricultural remote sensing. More specifically, it relates to a rice sub-pixel identification method based on the coefficient of variation of water body index. Background technique [0002] Changes in the spatial distribution of rice planting are crucial to my country's agricultural industrial structure adjustment, water resource monitoring, and greenhouse gas emissions. As the largest rice producing country in the world, rapid and accurate monitoring of the spatial distribution of rice is of great significance for ensuring food security and climate change research. Due to the limitations of data availability and computing power, high-precision optical remote sensing and synthetic aperture radar are difficult to apply to regional rice area mapping. At present, most of the regional or national-scale rice area mapping uses medium-resolution remote sensing data, which have high time resolution and are rela...

AI Technical Summary

Problems solved by technology

However, none of the above methods consider the sub-pixel problem of rice, and the recognition degree of rice in mountainous areas and intercropping between rice and other crops is relatively low.

Images