Abstract
Establishing the ground control point (GCP) network is a pre-requisite for georeferencing raw image data. Given current typical digital spatial database quality, much interest among users is about the accuracy of the geometric correction model that yields the final product. This paper reports an approach to optimizing GCP assembly using a genetic/evolution algorithm. The paper also suggests an optimal criterion for accuracy assessment through appraisal of global accuracy of the transformation, which is computed at each point of the image space. Experimental results demonstrate that the proposed approach has a great potential for selection of the best GCPs, and considerable improvement to the accuracy of geometric correction models can be expected when it is implemented.
Original language | English |
---|---|
Pages (from-to) | 101-120 |
Number of pages | 20 |
Journal | European Journal of Remote Sensing |
Volume | 48 |
DOIs | |
Publication status | Published - 2015 |
Externally published | Yes |
Keywords
- Genetic algorithm
- Geometric correction
- Georeferencing
- Global accuracy
- Ground control point
- Voronoi diagram