Title: Automatic remote sensing image mosaic technology of small surveying and mapping UAV based on edge computing

Authors: Yong Liu

Addresses: Zhejiang Tongji Vocational College of Science and Technology, Hangzhou, Zhejiang, China

Abstract: Aiming at the problems of slow running speed and poor real-time performance of traditional UAV RSI algorithm, this paper proposes an improved algorithm with good real-time performance and fast data processing speed. By studying the applicable conditions and ranges of direct average fusion, weighted average fusion, Poisson fusion and random sample consistency (RANSAC), this paper calculates the image mosaic results, and proposes an RSI mosaic fusion optimisation algorithm based on edge computing for small surveying drones. The experimental results show that the RANSAC splicing method adopted in this paper is shorter than the direct splicing algorithm, and the operation efficiency is increased by 70.8%. Compared with the weighted average fusion method, the smoothness is better and the operation efficiency is increased by 60%. Compared with the Poisson fusion algorithm, RANSAC algorithm has less difference in the brightness of the mosaic, and the efficiency is increased by 40%.

Keywords: unmanned aerial vehicle remote sensing image; edge computing; Poisson fusion; automatic image mosaic.

DOI: 10.1504/IJGUC.2024.140108

International Journal of Grid and Utility Computing, 2024 Vol.15 No.3/4, pp.220 - 230

Received: 13 Feb 2023
Accepted: 31 Mar 2023

Published online: 24 Jul 2024 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article