Research on Precise Monitoring of Forest Land Based on Multi-source Remote Sensing Data:A Case Study of Licheng District,Jinan City
BEI Guimin
DING Li
Abstract：Accurate and efficient monitoring of forest landl by remote sensing can provide important data reference for forest protection.In cloudy and foggy areas,optical remote sensing images are prone to cloud and fog interference,which makes forest monitoring difficult.To solve this problem,this paper constructs a forest land monitoring method that integrates optical images,radar images and digital elevation images.The reliability and advantages of the constructed method are verified by analyzing the accuracy of different feature stacks extracted by random forest and maximum likelihood classifier.With Licheng District of Jinan City as the study area and Landsat 8 and Sentinel-1 images in July 2018 as the data sources,the study results showed that the overall accuracy of the constructed method was improved by 1.56％and 2.55％,respectively,compared with solely using optical images under random forest and maximum likelihood classifier.The producer accuracy of forest extraction was improved by 1.13％and 1.32％,and the user accuracy was improved by 5.72％and 2.68％,respectively.
Keywords：forest monitoringremote sensingrandom forestmaximum likelihoodfeature stack
Publication Date：2024-10-28
Online Publishing Date：2026-05-22(First online date of this platform, not the publication date of the document)
Pages：6( 94-99 )