| In recent years,with the rapid development of UAV technology and the advent of various kinds of sensors,UAV remote sensing has been widely used in agriculture,forestry and grass industry,power,surveying and mapping,disaster emergency response and other fields,and the data of high spatial and temporal resolution and high spectral resolution are increasing in geometric order.How quickly and effectively to manage these multi-source,multi-scale and multi-temporal UAV remote sensing data has important practical significance for the application research of UAV remote sensing.This thesis focuses on the organization,management and sharing of UAV multi-source remote sensing data and carries out management research based on UAV visible light,multi-spectral and hyperspectral data,with a view to achieving efficient storage,fast search,online browsing,convenient sharing of UAV remote sensing data.The main research contents and conclusions are as follows:(1)Organization and management of UAV multi-source remote sensing data.From the perspective of the organization and management of image data and the formulation of metadata standards,aiming at the problems of the diversity of image data storage structure and the nonstandard data storage naming,a data organization directory structure and corresponding data storage naming specification for the UAV multisource remote sensing data management are designed;In combination with relevant industry standards,the detailed content of metadata is designed,and the automatic extraction of multi-source heterogeneous remote sensing data is completed through algorithm design and with the help of Python’s geographic location information database.The extracted metadata information is inserted into the database in batches for comprehensive management,realizing the precise expression of metadata and the rapid retrieval of metadata content.(2)Production and release of public data sets.According to the actual planting situation of crops in Xinjiang,cotton is selected as the data acquisition object.Through the Rikola hyperspectral sensor carried by the DJI M600,a total of 7 hyperspectral data of cotton seedling,bud,flowering,full flowering,boll,full boll and boll opening periods are obtained.Through the data preprocessing operation,the hyperspectral orthophoto images of cotton in 7 growth periods are obtained.The spectral accuracy of airborne cotton hyperspectral data is verified by using ground ASD data.The results show that the wavelengths of Rikola imaging spectrometer and ASD ground object spectrometer are in the range of 503～850 nm,and the trend of reflectivity curve has good consistency.The three typical spectral characteristics of "green peak feature","red valley feature" and "red edge feature" are also basically consistent.The data set can not only reflect the spectral characteristics of cotton in different growth periods,but also provide sample data for the fine monitoring of cotton in low altitude remote sensing.(3)UAV multi-source remote sensing data management and visualization system implementation.In view of the difficulties in the organization and management of UAV multi-source remote sensing data,the low efficiency of data search,and the dispersion of data,a set of data management and visualization system based on UAV visible light,multi-spectral and hyperspectral types was developed.The relational database My SQL and file management system are used for centralized storage of metadata and image data,and the B/S model architecture is adopted to build the fuzzy retrieval and spatial index based on metadata content according to the metadata information extracted from the original data.Through Web GIS technology,the release and call of image services are realized,and the online browsing and display of UAV remote sensing images are realized. |
PDF Full Text Request