A multi-band approach to unsupervised scale parameter selection for multi-scale image segmentation Jian Yang a,c,⇑ , Peijun Li b , Yuhong He c a Department of Geography, University of Toronto, 100 St. George Street, Toronto, ON M5S 3G3, Canada b Institute of Remote Sensing and GIS, School of Earth and Space Sciences, Peking University, Beijing 100871, China c Department of Geography, University of Toronto Mississauga, 3359 Mississauga Rd North, Mississauga, ON L5L 1C6, Canada article info Article history: Received 20 March 2013 Received in revised form 28 February 2014 Accepted 8 April 2014 Keywords: Object based image analysis Multi-scale segmentation Appropriate scale parameter selection Multi-band Spectral angle Segmentation evaluation abstract Image segmentation is one of key steps in object based image analysis of very high resolution images. Selecting the appropriate scale parameter becomes a particularly important task in image segmentation. In this study, an unsupervised multi-band approach is proposed for scale parameter selection in the multi-scale image segmentation process, which uses spectral angle to measure the spectral homogeneity of segments. With the increasing scale parameter, spectral homogeneity of segments decreases until they match the objects in the real world. The index of spectral homogeneity is thus used to determine multiple appropriate scale parameters. The performance of the proposed method is compared to a single-band based method through qualitative visual interpretation and quantitative discrepancy measures. Both methods are applied for segmenting two images: a QuickBird scene of an urban area within Beijing, China and a Woldview-2 scene of a suburban area in Kashiwa, Japan. The proposed multi-band based segmen- tation scale parameter selection method outperforms the single-band based method with the better rec- ognition for diverse land cover objects in different urban landscapes. Ó 2014 International Society for Photogrammetry and Remote Sensing, Inc. (ISPRS) Published by Elsevier B.V. All rights reserved. 1. Introduction The fidelity provided by very high resolution (VHR) data from IKONOS, QuickBird, GeoEye-1 and WorldView-2, has proven useful for numerous applications, such as impervious surface mapping (Lu et al., 2011; Lu and Weng, 2009; Yuan and Bauer, 2006) and tree crown delineation (Ardila et al., 2012; Mallinis et al., 2008; Song et al., 2010). Furthermore, object based image analysis (OBIA) yields better accuracy compared to traditional pixel-based image analysis (Cleve et al., 2008; Yan et al., 2006; Yuan and Bauer, 2006) when high within-class spectral variability occurs (Johansen et al., 2010). Of the first stage of OBIA, the aforemen- tioned literature suggests that image segmentation is the most influential on land cover object recognition. Consequently, several image segmentation algorithms have been proposed including region-growing segmentation (Benz et al., 2004), watershed segmentation (Li et al., 2010; Li and Xiao, 2007; Wang et al., 2004), and mean-shift segmentation (Comaniciu and Meer, 2002). Moreover, many image segmentation algorithms have been expanded to consider objects at multiple scales, since ground objects generally show multi-scale features in high resolution image (Blaschke, 2010; Bruzzone and Carlin, 2006; Hay et al., 2003). For instance, at fine scales a grass field may have spectral variability or patchiness related to micro-moisture regimes or worn patches dues to sporting events while at coarser scales the field in its entirety stands out from the surrounding urban environment. As such, multi-scale image segmentation addresses some of the defi- ciencies associated with single scale segmentation in complex land cover environments (Akçay and Aksoy, 2008; Carleer and Wolff, 2006; De Roeck et al., 2009; Li et al., 2011; Tilton et al., 2012) In image segmentation, the appropriate scale parameter is not readily apparent and is currently chosen by time consuming and subjective trial-and-error (Meinel and Neubert, 2004; Zhang et al., 2008). In lieu of trail-and-error approaches, several scale parameter selection methods have been proposed typically utilizing measures of the dissimilarity between a segmentation result and a reference image in which the optimal scale parameter is the best match to the reference image (Carleer et al., 2005; Chabrier et al., 2006; Liu et al., 2012; Möller et al., 2007; Neubert et al., 2008; Tong et al., 2012). In addition, Espindola et al. (2006) and Johnson and Xie (2011) selected the optimal segmentation scale parameter by assessing http://dx.doi.org/10.1016/j.isprsjprs.2014.04.008 0924-2716/Ó 2014 International Society for Photogrammetry and Remote Sensing, Inc. (ISPRS) Published by Elsevier B.V. All rights reserved. ⇑ Corresponding author at: Department of Geography, University of Toronto, 100 St. George Street, Toronto, ON M5S 3G3, Canada. Tel.: +1 416 978 3375; fax: +1 416 946 3886. E-mail address: yangjian19890528@gmail.com (J. Yang). ISPRS Journal of Photogrammetry and Remote Sensing 94 (2014) 13–24 Contents lists available at ScienceDirect ISPRS Journal of Photogrammetry and Remote Sensing journal homepage: www.elsevier.com/locate/isprsjprs