Band co-registration modeling of LAPAN-A3/IPB multispectral imager based on satellite attitude

4th International Symposium on LAPAN-IPB Satellite for Food Security and Environmental Monitoring

农业科学;生态环境科学

Hakim, P.R.^1 ; Syafrudin, A.H.^1 ; Utama, S.^1 ; Jayani, A.P.S.^1

Satellite Technology Center, National Institute of Aeronautics and Space (LAPAN), Rancabungur, Bogor

16310, Indonesia^1

关键词: Distorted images;    Geometric distortion;    Image matching algorithm;    Multispectral images;    Onboard image processing;    Pitch and roll angles;    Satellite attitude;    Sensor calibration;   

Others  :  https://iopscience.iop.org/article/10.1088/1755-1315/149/1/012060/pdf
DOI  :  10.1088/1755-1315/149/1/012060

来源: IOP

【 摘 要 】

One of significant geometric distortion on images of LAPAN-A3/IPB multispectral imager is co-registration error between each color channel detector. Band co-registration distortion usually can be corrected by using several approaches, which are manual method, image matching algorithm, or sensor modeling and calibration approach. This paper develops another approach to minimize band co-registration distortion on LAPAN-A3/IPB multispectral image by using supervised modeling of image matching with respect to satellite attitude. Modeling results show that band co-registration error in across-track axis is strongly influenced by yaw angle, while error in along-track axis is fairly influenced by both pitch and roll angle. Accuracy of the models obtained is pretty good, which lies between 1-3 pixels error for each axis of each pair of band co-registration. This mean that the model can be used to correct the distorted images without the need of slower image matching algorithm, nor the laborious effort needed in manual approach and sensor calibration. Since the calculation can be executed in order of seconds, this approach can be used in real time quick-look image processing in ground station or even in satellite on-board image processing.