| Development and Performance Analysis of a Near Real-Time Sensor Model Correction System for Frame Motion Imagery |
|
Hyuk Tae Kwon, Jin-Woo Koh, Sanghee Kim, Se Hyoung Park |
1Pixoneer Geomatics, Inc.
2The 3rd Research and Development Institute, Agency for Defense Development |
| 프레임동영상의 근실시간 센서모델 보정시스템 개발 및 성능분석 |
|
권혁태, 고진우, 김상희, 박세형 |
1(주)픽소니어
2국방과학연구소 제3기술연구본부 |
|
|
| Abstract |
|
Due to the increasing demand for more rapid, precise and accurate geolocation of the targets on video frames from UAVs, an efficient and timely method for correcting sensor models of motion imagery is required. In this paper, we propose a method to adjust or correct sensor models of motion imagery frames using space resection via image matching with reference data. The proposed method adopts image matching between the motion imagery frames and the reference frames which are synthesized from reference data. Ground or reference control points are generated or selected through the matching process in near real time, and are used for space resection to get adjusted sensor models. Finally, more precise and accurate geolocation of the targets can possibly be done on the fly, and we have got the promising result on performance analysis in terms of the geolocation quality. |
| Key Words:
Motion Imagery, Image Registration, Sensor Model Correction, Space Resection |
|
PDF Links
Full text via DOI
Download Citation
