Automatic Orthorectification of High-Resolution Optical Satellite Images Using Vector Roads

Abstract

This paper presents a completely automatic processing chain for orthorectification of optical pushbroom sensors. The procedure is robust and works without manual intervention from raw satellite image to orthoimage. It is modularly divided in four main steps: metadata extraction, automatic ground control point (GCP) extraction, geometric modeling, and orthorectification. The GCP extraction step uses georeferenced vector roads as a reference and produces a file with a list of points and their accuracy estimation. The physical geometric model is based on collinearity equations and works with sensor-corrected (level 1) optical satellite images. It models the sensor position and attitude with second-order piecewise polynomials depending on the acquisition time. The exterior orientation parameters are estimated in a least squares adjustment, employing random sample consensus and robust estimation algorithms for the removal of erroneous points and fine-tuning of the results. The images are finally orthorectified using a digital elevation model and positioned in a national coordinate system. The usability of the method is presented by testing three RapidEye images of regions with different terrain configurations. Several tests were carried out to verify the efficiency of the procedure and to make it more robust. Using the geometric model, subpixel accuracy on independent check points was achieved, and positional accuracy of orthoimages was around one pixel. The proposed procedure is general and can be easily adapted to various sensors.

Automatic Geometric Processing for Very High Resolution Optical Satellite Data Based on Vector Roads and Orthophotos

“…In order to support these findings 14 RapidEye 1B images were analysed (primary image set, see Section 4.1). The results presented in Table 4 and Figure 11 (left) show that the total RMSE of the non-eliminated GCPs and the RMSE of independent check points (ICPs) are in most cases far below one pixel, which is in line with the results in [24]. ICP residuals (Figure 11, right) reflect the RMSE values.…”

“…The number of GCPs does not drastically affect the results as the same accuracies are achieved with over 400 or as few as 50 points. As it was demonstrated in [24] the physical model is stable and immune to the presence of gross errors. The physical sensor model applied on VHR WorldView-2 images represented a greater challenge.…”

“…Some algorithms and results of the geometric corrections of the HR RapidEye system were partly presented in [24], therefore this paper focuses on the adaptation for the VHR WorldView-2 system. The listed algorithms were developed in a different order.…”

“…The backbone of the developed model [24] is similar to the one described in [34], however numerous important changes have been made. The relationship between the platform and the ground system is represented by collinearity equations.…”

“…The tests with three RapidEye images were extensively presented in [24]. The results showed that the model is capable of orthorectifying optical satellite images with an accuracy of under one pixel even when less than 20 GCPs are used and several gross errors of different magnitudes are present.…”