高光谱成像
遥感
计算机科学
计算机视觉
全球定位系统
校准
惯性测量装置
万向节
人工智能
全光谱成像
传感器融合
地理
物理
量子力学
电信
作者
Adduru U G Sankararao,Neeraj Kumar,P. Rajalakshmi
出处
期刊:2021 IEEE 4th International Conference on Computing, Power and Communication Technologies (GUCON)
[Institute of Electrical and Electronics Engineers]
日期:2020-10-02
卷期号:: 757-762
被引量:14
标识
DOI:10.1109/gucon48875.2020.9231236
摘要
Hyperspectral sensors are used to photograph an object in several well defined narrow optical bands in visible and near-infrared regions of the electromagnetic spectrum. The hyperspectral images have rich spectral information of the objects in the wide spectral range which allows us to identify the unique features of the objects. Airborne hyperspectral imaging in which a hyperspectral sensor is mounted on to a UAV and the aerial images are taken. The system setup and data acquisition are quite challenging as it involves the integration of and coordinating with many units. The GPS/IMU unit has to be properly installed and calibrated, whose data is used to geo-rectify the data cubes. The sensor has to be properly mounted on to the UAV using a gimbal which keeps the imager stable against any fluctuations due to motion of UAV or any external factors like wind etc. The data cubes need to be geo-corrected, which will map the image information with the GPS information that is acquired using the GPS/IMU unit and also removes any distortions present due to the motion of UAV. The speed and altitude of flight of the UAV and the frame rate of the imager have to be carefully chosen to acquire good quality data. The data acquired by the imager not only consists of the reflectance data from the object but also the instrument response and solar illumination, which needs to be eliminated to get absolute reflectance of the object. The data which is in absolute reflectance form is useful for analysis purpose and then orthomosaic is created to get the aerial view of the scene. In this paper, we present the system setup, proper data acquisition procedure of an airborne hyperspectral imaging system and also we discuss different calibrations involved at various stages of data acquisition.
科研通智能强力驱动
Strongly Powered by AbleSci AI