Research on regulatory model and technology of vegetation restoration based on remote sensing

By using Landsat remote sensing images and combining with the self-inspection data records of The Forest Inspection in Eshan County in 2018, the three monitoring methods of vegetation restoration, namely visual interpretation, algebraic operation and time series analysis, were compared to study the role of remote sensing technology in vegetation restoration supervision, with a view to providing effective technical means for the supervision of vegetation restoration by Forestry department. The results show that: The vegetation coverage can be identified by the combination of texture and true or false color bands in medium and low resolution remote sensing images, but it is difficult to accurately identify unformed forest for that the land with bare land spectral characteristics significantly stronger than the vegetations, and non-forest land with vegetation spectral characteristics such as crops, vegetables and tobacco; The median synthesis of annual Tasseled Cap Wetness (TCW, Tasseled Cap Wetness) showed better applicability than the maximum synthesis of annual NBR and NDVI in vegetation restoration monitoring for illegal parcels that were not restored in place after deforestation. Time series analysis can present the changing trend of vegetation spectral characteristics , especially the use of seasonal change characteristics of crops to make up for the shortcomings of the vegetation index extracted by the maximum value synthesis method that identifies too many crops as woodland vegetation, but it is not sensitive to the restoration and monitoring of the vegetation in small areas of fire-involved forests. In conclusion, it is feasible to carry out fine-grained monitoring of vegetation restoration based on the Forest Inspection mapped plots by remote sensing.