Improving Georeferencing Accuracy in Drone Imagery: Combining Drone Camera Angles with High and Variable Fields of View
Georeferencing ascertains the relation of the image or map being used by fixing it onto real-world coordinates estranging the world into sectors which is crucial for purposes such as mapping, surveying, monitoring the environment, and analyzing traffic speed. For accurate analysis and effective dec...
Saved in:
| Main Authors: | , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Qubahan
2025-07-01
|
| Series: | Qubahan Academic Journal |
| Subjects: | |
| Online Access: | https://journal.qubahan.com/index.php/qaj/article/view/1703 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | Georeferencing ascertains the relation of the image or map being used by fixing it onto real-world coordinates estranging the world into sectors which is crucial for purposes such as mapping, surveying, monitoring the environment, and analyzing traffic speed. For accurate analysis and effective decision-making, all drone images and footage necessitate accurate and precise spatial computations vis-a-vis their ground position. In this work, we present a novel mapping process based on drone controlling data including telemetry like coordinates with accompanying GPS, mounted height, drone position, and horizontal and vertical fields of view. The technique applies lens distortion and geographical curvature compensation to the coordinate changing processes. It computes the offsets toward East-West and North-South by summing up slant range with viewing angle of the camera. A point of interest marked as a POI is set in the image where the coordinates that are supposed to be validated are also accepted as real coordinates by GPS. Testing proven that there is a differing benefit in accuracy of mapped images, distance relative to their terrain positions and their devices. The proposed approach further shows a quantitative improvement of 12.50% to 75.0% in the geolocation error reduction claimed. It was achieved by the decrease of MAE from 0.108 km to 0.055 km while RMSE was lowered from 0.111 km to 0.057 km indicating the reliability of the method. The study offers a strong, geometry-driven approach for drone image georeferencing that surpasses conventional techniques. It offers a scalable, accurate, and less labour-intensive substitute for spatial positioning by using FOV parameters and real-time telemetry data. Improved georeferencing accuracy ensures precise spatial data integration, and supports accurate image-frame alignment. Its integration into geospatial workflows enhances situational awareness and decision-making in traffic control, urban development, and environmental observation. This paper stresses the importance of FOV correction for the greater drone geospatial analysis system performance.
|
|---|---|
| ISSN: | 2709-8206 |