Inicio  /  Drones  /  Vol: 7 Par: 1 (2023)  /  Artículo
ARTÍCULO
TITULO

A Spatio-Temporal Fusion Framework of UAV and Satellite Imagery for Winter Wheat Growth Monitoring

Yan Li    
Wen Yan    
Sai An    
Wanlin Gao    
Jingdun Jia    
Sha Tao and Wei Wang    

Resumen

Accurate and continuous monitoring of crop growth is vital for the development of precision agriculture. Unmanned aerial vehicle (UAV) and satellite platforms have considerable complementarity in high spatial resolution (centimeter-scale) and fixed revisit cycle. It is meaningful to optimize the cross-platform synergy for agricultural applications. Considering the characteristics of UAV and satellite platforms, a spatio-temporal fusion (STF) framework of UAV and satellite imagery is developed. It includes registration, radiometric normalization, preliminary fusion, and reflectance reconstruction. The proposed STF framework significantly improves the fusion accuracy with both better quantitative metrics and visualized results compared with four existing STF methods with different fusion strategies. Especially for the prediction of object boundary and spatial texture, the absolute values of Robert?s edge (EDGE) and local binary pattern (LBP) decreased by a maximum of more than 0.25 and 0.10, respectively, compared with the spatial and temporal adaptive reflectance fusion model (STARFM). Moreover, the STF framework enhances the temporal resolution to daily, although the satellite imagery is discontinuous. Further, its application potential for winter wheat growth monitoring is explored. The daily synthetic imagery with UAV spatial resolution describes the seasonal dynamics of winter wheat well. The synthetic Normalized Difference Vegetation Index (NDVI) and Enhanced Vegetation Index 2 (EVI2) are consistent with the observations. However, the error in NDVI and EVI2 at boundary changes is relatively large, which needs further exploration. This research provides an STF framework to generate very dense and high-spatial-resolution remote sensing data at a low cost. It not only contributes to precision agriculture applications, but also is valuable for land-surface dynamic monitoring.

 Artículos similares

       
 
Sheeraz Arif, Jing Wang, Tehseen Ul Hassan and Zesong Fei    
Human activity recognition is an active field of research in computer vision with numerous applications. Recently, deep convolutional networks and recurrent neural networks (RNN) have received increasing attention in multimedia studies, and have yielded ... ver más
Revista: Future Internet

 
Yongxing Wu, Shaofu Lin, Fei Peng and Qi Li    
Grand sites are important witnesses of human civilization. The archeology of grand sites has the characteristics of a long period, interdisciplinary study, irreversibility and uncertainties. Because of the lack of effective methods and valid tools, large... ver más

 
Luliang Tang, Qianqian Zou, Xia Zhang, Chang Ren and Qingquan Li    

 
Chu He, Zhi Zhang, Dehui Xiong, Juan Du and Mingsheng Liao    

 
Tang-Hsien Chang, Albert Y. Chen, Yu-Ting Hsu, Chien-Li Yang     Pág. 452 - 459
Travel time prediction has been the fundamental brick for developing both Advanced Traffic Management Systems (ATMS) and Advanced Traveler Information Systems (ATIS). The problem of travel time prediction is primarily shaped by the data available for mod... ver más