The spatial continuity study of NDVI based on Kriging and BPNN algorithm
Under the framework of the soil-wheat system, the sampling area was selected in the demonstration site of Lingxian country of Shandong province, Normalized Difference Vegetation Index (NDVI) of 195 sites were collected by GreenSeeker optical sensor based on the GPS localization data. The main objective of the paper developed the comparison study by Kriging and BPNN algorithm for NDVI continuity surface during wheat growth stage. The results showed that the strong variability existed in the spatial pattern of NDVI values, the structural factors variability is 88.6％, indicated the physiological parameters of wheat growth stage mainly affected NDVI measure values on Kriging algorithm, for BPNN algorithm, its simulation results showed that the compact and continuity changes of NDVI values in case area. The spatial distribution trend of NDVI values, there is in accordance with the simulation results on the large scope for two algorithms, but BPNN algorithm has higher estimated value than Kriging algorithm, and the prediction value is higher in the west of the whole study area than the measured value corresponding algorithm structure and the approximation ability of BPNN algorithm. In a summary, whether algorithm structure characteristics, or the interpolation description, results confirmed that BPNN algorithm has the better accuracy and more advantage than Kriging algorithm in the study.
NDVI Back Propagation Neural Network Kriging Spatial continuity surface
Yujian Yang Jianhua Zhu Chunjiang Zhao Shuyun Liu Xueqin Tong
S&T Information and Engineering Research Center of Shandong Academy of Agricultural Science. Jinan 2 National Engineering Research Center for Information Technology in Agriculture. Beijing, 100097. PR
国际会议
南昌
英文
1138-1144
2010-10-22（万方平台首次上网日期，不代表论文的发表时间）