Vision-Based Terrain-Referenced Navigation with Iterative Closest Point Algorithm

Taeyun Kim, Seongho Nam, Changky Sung, and Junpyo Park

Abstract:	Terrain-referenced navigation (TRN) is an alternative navigation technique for use in environments with limited GPS availability. In this paper, we present a vision-based TRN that estimates the position of a vehicle by comparing the terrain elevation extracted from aerial images with a terrain elevation database. Compared to the radar altimeter commonly used in TRN, image sensors have the advantage of being built into most platforms and being able to relatively more terrain elevations at a time. This paper proposes the use of position fix measures through an iterative closest point (ICP) algorithm by utilizing the advantage that the image sensor can acquire terrain elevations over a large area compared to the radio altimeter. Three-dimensional terrain elevation information is extracted through a stereo matching technique using continuously acquired aerial images from an image sensor installed downward on the vehicle. The acquired three-dimensional terrain information is compared with a terrain elevation database using the ICP algorithm to obtain a position fix. The position of the vehicle is estimated by applying a filtering algorithm that uses the obtained position fix as a measure. To demonstrate the validity and feasibility of the proposed method, numerical simulations were performed using aircraft motion data and aerial images. Inertial navigation data was generated in consideration of the specification of the inertial navigation device and the images were captured along the flight trajectory through an image generation tool. Monte-Carlo (MC) simulations were performed to evaluate the performance of the proposed method.
Published in:	Proceedings of the ION 2024 Pacific PNT Meeting April 15 - 18, 2024 Hilton Waikiki Beach Honolulu, Hawaii
Pages:	132 - 142
Cite this article:	Kim, Taeyun, Nam, Seongho, Sung, Changky, Park, Junpyo, "Vision-Based Terrain-Referenced Navigation with Iterative Closest Point Algorithm," Proceedings of the ION 2024 Pacific PNT Meeting, Honolulu, Hawaii, April 2024, pp. 132-142. https://doi.org/10.33012/2024.19635
Full Paper:	ION Members/Non-Members: 1 Download Credit Sign In