Kamirul1, R Madina2 and P R Hakim2
1 Satellite Control, Space, Atmospheric, and Remote Sensing Observation Center, Indonesian National Institute of Aeronautics and Space, Biak, Indonesia
2 Satellite Technology Center, Indonesian National Institute of Aeronautics and Space Bogor, Indonesia
kamirul.apr@gmail.com
Abstract. Due to data transmission, several schemes of unavoidable error may be experienced. Those conditions inflict byte shifting on transferred data and lead to error of processed data i.e. processed remote sensing image. One of the common problems is the appearance of line-drop pattern covering the image and is possible to lead misleading interpretation. This work is aimed to develop line-drop correction algorithm applied on three different data level of LAPAN-A2 microsatellite imagery. Developed algorithm can be used as a solution to avoid data re-transmission in order to recover a clearer image data.
Introduction
Transmitting satellite data becomes a huge challenge since unavoidable errors may occur i.e. attitude error of satellite, frequency interference, weather condition, and time difference between satellite and antenna system. Those problems account for byte-shifting and undesirable arrangement of transmitted coded satellite data. As one of experimental satellite developed in 2016, same problem was faced by LAPAN-A2/ORARI microsatellite. This satellite was built to support earth observation mission by capturing RGB image using matrix camera. Processing byte-shifted LAPAN-A2 data results in noise disturbed image. This common forms of noise include systematic striping or banding called line-drop effect [1].
The presence of line-drop effect affects error interpretation on such image while used to recognize existing object. In specific case of LAPAN-A2 images, the image above was used to validate Automatic Identification System (AIS) data that also recorded by this satellite. For this purpose, existed ship might be covered by line-drop effect so that this image cannot be used to validate AIS data. Therefore, correcting line-drop effect become an important work to be conducted since recovering fine data must be done by re-downloading same satellite data which is not an effective solution to do.
In this work, we have performed line-drop correction on three level of matrix camera data of LAPAN-A2. The first correction data was the raw data. Since line-drop effect caused by byte-shifting effect, performing correction on this level was focused on detecting and correcting byte-shifting pattern on every single line of raw data. This can be done since byte-shifted raw data having a unique saturated anomaly value. The anomaly data having slightly different value compared to normal data so that anomaly data location can be detected by applying peak finding method on raw data array. Detected anomaly data is then replaced by interpolating its nearest neighbor data.
Result Overview
