Geometric Accuracy Assesment of Orthorectification Method Based on Sensor Model Refinement in Open Source System Environment (Case Study : Sangir Subdistrict, South Solok District, West Sumatra Province)

Abstract

Orthorectification is one of the methods of correcting geometric condition of satellite imagery. Commonly used in high resolution satellite imagery such as IKONOS, Quickbird, and Pleiades, to reduces the effect of relief displacement. The Goal of this study is to implement the sensor model refinement procedures in orthorectification method. In this case we use orfeo toolbox (OTB) version 6.6.0 Installed in linux Mint 19. A true color Pleiades imagery of the area study in 2013, ACE2 DEM, and ground control points obtained by were used to obtain refined sensor model. This Study aimed to obtain the geometric accuracy of the output imager. The sensor model refinement process was done by access the imagery sensor model and changes its parameters using ground control points. The process of systematic orthorectification then carried out to transform the satellite imagery using modified sensor models. The geometric accuracy presented as horizontal accuracy were obtained by plotting several check points in the orthorectification output measured and calculated to obtain the horizontal accuracy. In this study we obtained that the sensor model refinement has an overal process accuracy of 1.664 meters. The model refined the imagery horizontal accuracy to 0.800 meters.