The Multi-source Image NDVI Interactive Corrective Method for Long-term Remote Sensing Monitoring of Rare Earth Mining Area

Abstract

ABSTRACT In order to solve the inconsistency of the normalized difference vegetation index (NDVI) caused by the differences in the use of sensors, this essay is aimed to provide a practicable method to make the long-term remote sensing monitoring of rare earth mining area possible. The quantitative relationship between the two NDVI was studied by regression analysis using four pairs of Landsat TM/OLI and HJ-1B CCD image pairs that were transited on the same day. Two pairs of images were selected as experimental images to construct the NDVI conversion equation, and the other two pairs of images were used to verify the accuracy of the conversion equation. The results show that there is an obvious linear positive correlation between the NDVI of Landsat TM/OLI and HJ-1B CCD images. The conversion equations constructed have higher precision, and the root mean square error of the conversion equation is within 0.05. In the scope, the conversion equation can be used to realize the high precision phase conversion of the NDVI of Landsat TM/OLI and HJ-1B CCD data, which is beneficial to complement each other’s vegetation monitoring results and provide support for monitoring and analysis of multi-source time series images.