Processing Method Research On Unmanned Aerial Vehicle(UAV) Near Infrared Data

With the development of science and technology of unmanned aerial vehicle(UAV), the demand for unmanned aerial vehicle(UAV) data in every walk of life is increasing. UAV as a new kind of remote sensing technology, it gets more recognition and more opportunity of application. Lower operating costs, highly efficient and flexible task arrangement, and automatic and intelligent operation application make it more and more popular in the field of application.But in comparison with the space remote sensing, the sensor can be loaded on UAV is still single, mainly visible light camera.it lack of new sensors which develop rapidly in recently years such as IR,hyperspectral and Li DAR. So how to broaden the UAV sensor data band range, and how to improve the application ability of UAV, it is in urgent need of the research question.Based on the predecessors’ research results and reference to foreign research data, this paper studies for the feasibility of low altitude UAV remote sensing can load near infrared sensor, and designs data acquisition scheme.Based on the acquisition method, selecting the study area(research area must have river, the vegetation and building such as artificial features information), acquiring the UAV data of study area, treating the data as an example, and integrate comparatively analyzing visible light, near infrared and satellite data(for LANDSAT TM5 data).selected integrated contrast factor is NDVI(normalized difference vegetation index) and SAVI(soil adjusted vegetation index), access to near infrared data characteristic information of different types of terrain drones.First of all, from the aspects of near infrared sensor calibration, by selecting a large number of features of objects, by using near infrared sensor to image, by referencing the characteristics of different objects have different reflective properties, and obtaining the reflection value to calculate the gain and offset value, this paper uses these related values to do radiation calibration for near infrared sensors.Due to the data obtained by near infrared sensors of UAV is single photo, in order to form a clear contrast in picture with the satellite data, it needs to do Mosaic treatment and geographical correct treatment for all single pictures. In respect of geographical correction, this paper chose three distinctive geographical correction method, they are collinear equation method, polynomial geometric correction method, the RFM model geometric correction method, through the comparison of three methods and the results of the geometric correction of comprehensive considered in this paper, choose the best way to correct as geometric correction method in this paper.In order to compare the unmanned aerial vehicle(UAV) data and satellite data to obtain unmanned aerial vehicle(UAV) data availability in near infrared wave band, and verify its stability. Data can be divided into, artificial vegetation, ground construction, water and other selected categories, and then select AOI area(point, line, plane), for obtaining NDVI and SAVI values of each type of characteristic feature area of AOI. By comparing NDVI value and SAVI acquired by the unmanned aerial vehicle(UAV) data and LANDSAT TM5 data, to explore the connection between the two.Through comparison and analysis to verify the stability and reliability of the data that obtained by the near infrared sensors loaded by UAV, And compared with the traditional satellite platform, unmanned aerial vehicle(UAV) to obtain the data with higher spatial resolution, stronger real-time, accuracy, cheap cost and high spectral resolution ability, etc. Through this research, it can provide a wide range of new sensors used to provide a more detailed data acquisition method, as more can carry on the development of the light small sensors on the UAV, the development of data processing and application technology, unmanned aerial vehicle(UAV) applications also will be further expanded.