Share Email Print

Journal of Applied Remote Sensing

Extended census transform histogram for land-use scene classification
Author(s): Baohua Yuan; Shijin Li
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

With the popular use of high-resolution satellite images, more and more research efforts have been focused on land-use scene classification. In scene classification, effective visual features can significantly boost the final performance. As a typical texture descriptor, the census transform histogram (CENTRIST) has emerged as a very powerful tool due to its effective representation ability. However, the most prominent limitation of CENTRIST is its small spatial support area, which may not necessarily be adept at capturing the key texture characteristics. We propose an extended CENTRIST (eCENTRIST), which is made up of three subschemes in a greater neighborhood scale. The proposed eCENTRIST not only inherits the advantages of CENTRIST but also encodes the more useful information of local structures. Meanwhile, multichannel eCENTRIST, which can capture the interactions from multichannel images, is developed to obtain higher categorization accuracy rates. Experimental results demonstrate that the proposed method can achieve competitive performance when compared to state-of-the-art methods.

Paper Details

Date Published: 25 April 2017
PDF: 15 pages
J. Appl. Rem. Sens. 11(2) 025003 doi: 10.1117/1.JRS.11.025003
Published in: Journal of Applied Remote Sensing Volume 11, Issue 2
Show Author Affiliations
Baohua Yuan, Hohai Univ. (China)
Nanjing Univ. of Science and Technology (China)
Shijin Li, Hohai Univ. (China)

© SPIE. Terms of Use
Back to Top