Share Email Print

Journal of Applied Remote Sensing

Aeromagnetic gradient compensation method for helicopter based on ϵ -support vector regression algorithm
Author(s): Peilin Wu; Qunying Zhang; Chunjiao Fei; Guangyou Fang
Format Member Price Non-Member Price
PDF $20.00 $25.00
cover GOOD NEWS! Your organization subscribes to the SPIE Digital Library. You may be able to download this paper for free. Check Access

Paper Abstract

Aeromagnetic gradients are typically measured by optically pumped magnetometers mounted on an aircraft. Any aircraft, particularly helicopters, produces significant levels of magnetic interference. Therefore, aeromagnetic compensation is essential, and least square (LS) is the conventional method used for reducing interference levels. However, the LSs approach to solving the aeromagnetic interference model has a few difficulties, one of which is in handling multicollinearity. Therefore, we propose an aeromagnetic gradient compensation method, specifically targeted for helicopter use but applicable on any airborne platform, which is based on the ϵ -support vector regression algorithm. The structural risk minimization criterion intrinsic to the method avoids multicollinearity altogether. Local aeromagnetic anomalies can be retained, and platform-generated fields are suppressed simultaneously by constructing an appropriate loss function and kernel function. The method was tested using an unmanned helicopter and obtained improvement ratios of 12.7 and 3.5 in the vertical and horizontal gradient data, respectively. Both of these values are probably better than those that would have been obtained from the conventional method applied to the same data, had it been possible to do so in a suitable comparative context. The validity of the proposed method is demonstrated by the experimental result.

Paper Details

Date Published: 8 June 2017
PDF: 12 pages
J. Appl. Rem. Sens. 11(2) 025012 doi: 10.1117/1.JRS.11.025012
Published in: Journal of Applied Remote Sensing Volume 11, Issue 2
Show Author Affiliations
Peilin Wu, Institute of Electronics (China)
Univ. of Chinese Academy of Sciences (China)
Qunying Zhang, Institute of Electronics (China)
Chunjiao Fei, Institute of Electronics (China)
Univ. of Chinese Academy of Sciences (China)
Guangyou Fang, Institute of Electronics (China)

© SPIE. Terms of Use
Back to Top