A Study of Forestland Change Detection Based on Dimidiate Pixel Model

This paper took Shapotou District of Zhongwei city in Ningxia Hui Autonomous Region as study area, using GF-1 remote sensing images of the year 2015 and 2016, from which the NDVI value were extracted to generate difference images of vegetation index generation two pixel inversion model, detect the reduced index information of vegetation. At the same time, overlay all kinds of recent forestry thematic data, and identify changes of forestland through visual interpretation, screening, classification. The results showed that the model reflected the change in the research area more quickly and accurately, which provided technical support for the management of forest resources, forest management, supervision and law enforcement.