Research On Techniques For Attitude Correction Of Airborne Scanning Images Based On Non Rigid Connection Position And Attitude Data

With the development of hyperspectral remote sensing technology.it has been widely used in the fields of meteorological and environmental monitoring,agricultural crop growth and soil erosion analysis,mineral resource detection,military investigation and service,etc.Aerial remote sensing,also known as airborne remote sensing,is remote sensing technology that uses aircraft,airships and other vehicles.The quality of aerial remote sensing image is an important factor that hinders the development of hyperspectral remote sensing technology.The main factors that affect image quality are listed as follows:(1)Flight height and velocity of the carrier;(2)Imager's shooting angle;(3)The change of aircraft attitude;(4)Environmental factors,including the change of light and environmental disturbances.These interference factors may lead to the undesirable characteristics of the image,such as large geometric variation,small overlap area and many similar features.In order to improve the quality of the remote sensing image,the system depicted in this thesis combines low precision image stabilization platform that is used to ensure the image data distortion is not too large with high precision pose measurement system that is used to correct the distortion of the image data afterwards.Usually,the attitude correction of the airborne scanned image is based on a rigid connection between the position and attitude system and the imaging system.While the system depicted in this thesis is different,the imaging system is installed on a small image stabilization platform,while the position and attitude system is installed on the vehicle.There are non-rigid connection between them.The stabilization platform is an active attitude adjustment system,and the stability range is limited.Once overrun,it will automatically reset.so the relative movement is more complex.The characteristics of the system make the image attitude correction process much more difficulties,for it brings in more interference characteristics.From discussions above,we can see that the image attitude correction technology is be widely used in many aspects in the science fields,and it faces much more challenges and deserves more research.This thesis studies the geometric calibration,POS data preprocessing and attitude correction for airborne hyperspectral imager with the stabilization platform,and completes the image attitude correction software system.The main work of the paper thesis are listed as follows:(1)Briefly introduce the main part of the image acquisition system,including the image data acquisition system,the image stabilization platform system,the position and attitude data acquisition system.(2)In view of the structural characteristics of multi-field hyperspectral imager and the requirements of high-precision geometric scaling,we use experiment instruments such as full-color light source,mirror,mirror rotation platform,parallel light pipe and computer to create a new method of geometric calibration method.The results of geometric calibration show that the pixel error of the imaging instrument reaches sub-pixel accuracy and meets application requirements.(3)The attitude correction is the geometric attitude for image data based on pose data correction,the correct attitude correction requires pose data and image data strictly synchronized,but the system will bring the synchronization deviation,this paper proposes a variety of ways to eliminate the deviation.It will determine the best effect of the hyperspectral image data by screening that,and save the parameters for later related data processing.(4)This paper discusses the geometric correction algorithm based on the registration of the position and orientation data,including the calculation of exterior orientation elements,the transformation of spatial coordinates,the resampling that uses a new local distance reciprocal algorithm,and the multi-threaded tool OpenMP that speed up the image processing speed.The result of the experiment meets the goal.(5)The thesis describe the framework and realization of airborne scanning image data attitude correcton software system.