Automated bias-compensation of rational polynomial coefficients of high resolution satellite imagery based on topographic maps

As the need for efficient methods to accurately update and refine geospatial satellite image databases is increasing, we have proposed the use of 3-dimensional digital maps for the fully-automated RPCs bias compensation of high resolution satellite imagery. The basic idea is that the map features ar...

Full description

Saved in:
Bibliographic Details
Published in:ISPRS journal of photogrammetry and remote sensing 2015-02, Vol.100, p.14-22
Main Authors: Oh, Jaehong, Lee, Changno
Format: Article
Language:English
Subjects:
Accuracy
Bias
Bias-compensation
Compensation
Digital topographic map
Georeferencing
High resolution
High-resolution satellite imagery
Matching
Mathematical models
RECC
RPCs
Satellite imagery
Topographic maps
Citations: Items that this one cites
Items that cite this one
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:As the need for efficient methods to accurately update and refine geospatial satellite image databases is increasing, we have proposed the use of 3-dimensional digital maps for the fully-automated RPCs bias compensation of high resolution satellite imagery. The basic idea is that the map features are scaled and aligned to the image features, except for the local shift, through the RPCs-based image projection, and then the shifts are automatically determined over the entire image space by template-based edge matching of the heterogeneous data set. This enables modeling of RPCs bias compensation parameters for accurate georeferencing. The map features are selected based on four suggested rules. Experiments were carried out for three Kompsat-2 images and stereo IKONOS images with 1:5000 scale Korean national topographic maps. Image matching performance is discussed with justification of the parameter selection, and the georeferencing accuracy is analyzed. The experimental results showed the automated approach can achieve one-pixel level of georeferencing accuracy, enabling economical hybrid map creation as well as large scale map updates.
ISSN:0924-2716
1872-8235
DOI:10.1016/j.isprsjprs.2014.02.009