Hyperspectrcal Imagery Remote Sensing Technology Based On Lightweight Unmanned Aerial Vehicle

Hyperspectral imaging technology combines both spectroscopy and imaging technology and it can obtain the spectral information and spatial information of the target at the same time. It has been widely used in mineral exploration, forestry resource survey, environment monitoring, city planning, etc. But most of the remote-sensing platforms for hyperspectral imaging system are mainly satellites or manned aircrafts. It is difficult for those small and medium- sized users to afford the high price, and simultaneously they can hardly to stand the harsh conditions of use. Comparing with those platforms, the light-weight UAV has many advantages like low-cost, real-time, and adapting to a more complex environment, ect. With the continuous development of light-weight UAV technology in recent years, it becomes possible for UAV to be the remote-sensing platform. This paper developed a set of hyperspectral imaging system on the lightweight UAV based on the project- "Development of satellite, aviation, ground spectral detection integration key technology research and equipment", which started by C hina Geological Survey Bureau in 2011. In order to adapt to the characteristics of lightweight UAV platform, the main consideration in this system is lightweight, miniaturization, modularization and low power consumption. This system consists of five subsystems: power management system, hyperspectral imaging system, data acquisition and control system, POS system and monitoring control system. The paper’s research content and achievements include: After the entire system completed, we made a serials of experiments to calibrate the parameters and test the performance. In order to do quantitative analysis of the hyperspectral data, firstly, the spectral calibration and radiometric calibration of the hyperspectral imaging system were made.Then analyzed the precision of the calibration based on the principle of the spectral calibration and radiometric calibration.According to the principle of collimation monochromatic calibration,we designed a set of automated spectral calibration software to reduce the workload of calibration. Found the phenomenon that the dark background of SWIR hyperspectral imager drift with the temperature. So made the theoretical analysis and simulated through Tracepro in this paper. At the same time, a serial of image experiment have been performed in the temperature control box. The results show that dark background temperature drift is caused by thermal radiation instrument itself. Finally this paper the puts forward a series of measures to reduce the effect of thermal radiation instrument. According to the method of geometric correction of push-broom imaging data, this paper analyzed the effect that on the accuracy of the attitude and position data has been for the accuracy of geometric correction and put forward the methods to reduce the effect. Put the entire hyperspectral imaging system on the lightweight diesel powered UAV, and obtained the hyperspectral data successfully. After make the geometric correction of the hyperspectral data, we found that the spatial resolution of the image is lower than the spatial resolution of the hyperspectral imager due to the shock of the UAV. According to principle of the staring hyperspectral imaging system, this paper analyzed the practicability of it used on the lightweight electric power UAV and researched the effect of the UAV platform to the remote-sensing. This paper proposes the method that correct spectral information of the target through image registration by extracting feature points of the image. Simulated this method by the real hyperspectral data. Then we simplified the entire system and replaced the Offner hyperspectral imager by AOTF hyperspectral imager so that the system can been taken on the lightweight electric power UAV. Finally we carried the remote imaging experiment and obtained the hyperspectral data successfully.