Combined Block Adjustment for Optical Satellite Stereo Imagery Assisted by Spaceborne SAR and Laser Altimetry Data

Guo Zhang,Yuanxin Ye,Xin Li,Boyang Jiang,Taoyang Wang

doi:10.3390/rs13163062

Abstract

To ensure the accuracy of large-scale optical stereo image bundle block adjustment, it is necessary to provide well-distributed ground control points (GCPs) with high accuracy. However, it is difficult to acquire control points through field measurements outside the country. Considering the high planimetric accuracy of spaceborne synthetic aperture radar (SAR) images and the high elevation accuracy of satellite-based laser altimetry data, this paper proposes an adjustment method that combines both as control sources, which can be independent from GCPs. Firstly, the SAR digital orthophoto map (DOM)-based planar control points (PCPs) acquisition is realized by multimodal matching, then the laser altimetry data are filtered to obtain laser altimetry points (LAPs), and finally the optical stereo images’ combined adjustment is conducted. The experimental results of Ziyuan-3 (ZY-3) images prove that this method can achieve an accuracy of 7 m in plane and 3 m in elevation after adjustment without relying on GCPs, which lays the technical foundation for a global-scale satellite image process.

Highlights

Satellite image mapping plays an important role in many fields such as social economy and national defense
Li et al used both a rational function model (RFM) and rigorous model for joint adjustment with laser points as elevation constraints to verify the feasibility of laser altimetry points (LAP) in block adjustment, and the results showed that both methods could achieve a positioning accuracy better than 3 m after adjustment [20]
The observation equation of image point coordinates is established separately, the constraint of plane direction and elevation direction is realized by combining synthetic aperture radar (SAR) DOMbased planar control points (PCPs) and LAPs, respectively, and the image orientation correction parameters are obtained by solving the equation to improve the plane and elevation geometric accuracy of the stereo image

Summary

Introduction

Satellite image mapping plays an important role in many fields such as social economy and national defense. To eliminate the influence of the system error and achieve high-accuracy geometric positioning of satellite images, the image positioning error can usually be eliminated by using a large number of ground control points (GCPs) through bundle block adjustment [1,2]. With the development of bundle block adjustment technology under sparse control conditions, the GCP-dependent adjustment method no longer requires control points for each image in the area, reducing the need for the number of control points [3]. This adjustment method does not completely eliminate the reliance on GCPs, and the quality and quantity of ground control points are still crucial to Remote Sens.

Methods

Results

Discussion

Conclusion