Impact of imaging geometry on 3D geopositioning accuracy of stereo ikonos imagery

Rongxing Li, Xutong Niu, Chun Liu, Bo Wu, Sagar Deshpande

Research output: Contribution to journalArticlepeer-review

37 Scopus citations

Abstract

Special Ikonos data acquisition and investigation were conducted to study the relationship between three-dimensional (3D) geopositioning accuracy and stereo imaging geometry, in particular, convergence angles. Six Ikonos images (four on one track and two on another track) were collected for a test site at Tampa Bay, Florida, in 2004 and 2007, respectively. Different combinations of Ikonos stereo image pairs, both along-track and cross-track, were formed. Using the highresolution satellite image processing system developed at The Ohio State University, DGPS(Differential Global Positioning System) controlled ground control points, and a number of check points, we demonstrated: (a) The convergence angle plays an important role in along-track or cross-track stereo mapping, especially in improvement of the accuracy in the vertical direction; (b) Regardless of stereo configuration (alongtrack or cross-track), the accuracy in the X (cross-track) direction is better than that in the Y (along-track) direction; and (c) Although there is a slight correlation between the convergence angle and the accuracy in the Y (along-track) direction in the case of along-track stereo configuration, no distinct relationship is found in the X (cross-track) direction. Similarly, improvement of the horizontal accuracies is found with increased convergence angles when dealing with crosstrack stereo pairs.

Original languageEnglish (US)
Pages (from-to)1119-1125
Number of pages7
JournalPhotogrammetric Engineering and Remote Sensing
Volume75
Issue number9
DOIs
StatePublished - 2009

All Science Journal Classification (ASJC) codes

  • Computers in Earth Sciences

Fingerprint

Dive into the research topics of 'Impact of imaging geometry on 3D geopositioning accuracy of stereo ikonos imagery'. Together they form a unique fingerprint.

Cite this