Research On Vegetation Coverage Based On Multi-source Data In Part Area Of Dalian

This paper mainly discussed the methodology of how vegetation coverage model was established.By planning the experimental area like Shahekou and Jinzhou,and using Landsat8 OLI image data,high-resolution image data from Google Earth,ground sample image data collected by digital camera,an OLI-G-RS Model was formed based on the dimidiate pixel theory.Also image fusion and regression analysis were used to process the data that the model used.Through comparing the results of OLI-G-RS Model with dimidiate pixel model,the OLI-G-RS Model was better.Finally,the OLI-G-RS Model was put into the application of deriving the vegetation coverage information experiment and got the conclusions as follow:(1)The regression analysis was built based on ground vegetation coverage and Estimated vegetation coverage by Remote Sensing,The fitting degree of the model is between 0.7054-0.8808.Four sets of ground vegetation coverage data were used to Validate the model.The RMSE between ground vegetation coverage and model estimation is 0.102042689,0.135219726,0.160497321,The results show the the model results are good(2)Comparative analysis of vegetation coverage in the study area calculated by the two models,the correlation coefficient between the two groups is 0.995046907,the correlation is well.(3)The OLI-G-RS Model and dimidiate pixel model were applied to the Shahekou and Jinzhou districts in Dalian.Extracted the vegetation coverage of the two districts in June and October 2016,and the average vegetation coverage of each street in the whole area is obtained.Analyzing the resultsof different streets can elicit that the results that obtained by the OLI-G-RS model is more realisticing and reflecting the vegetation growth status of the study area.