Dark-Image Modeling Analysis of LAPAN-A3 Satellite Imager

To produce a high quality image, raw image produced by LAPAN-A3 satellites imager must be systematically corrected, both in geometric and radiometric aspect. One important variable that is used in radiometric correction process is dark-image, which is an image produced when the camera is closed, thus the camera does not receive any light source. Ideally, the dark-image is acquired directly in orbit just before image acquisition period, but it cannot be done since LAPAN-A3 satellite does not have auto-calibration mechanism or referenced detector. This research aims to model the dark-image of both LAPAN-A3 imagers as a function of several imager intrinsic parameters and period of imager operation, using least-square linear regression method. The resulted model is then used to estimate the dark-image value which is subsequently used in systematic radiometric correction calculation. Experiments and calculation that have been done show that matrix imager has constant dark-image during entire acquisition period, and can be accurately modeled using simple equation as a function of camera parameter only. Meanwhile, multispectral imager has an increasing monotonic dark-image response during acquisition period, which is caused by significant detector temperature increase. Dark-image of multispectral imager can be modeled using linear equation as function of operating-time only, but the model could only produce good estimation for specific camera parameter. All of these results can be used as references in processing raw data of LAPAN-A3 satellites imager, so that the final image produced will have a high radiometric quality