A Multi-scale Segmentation Approach to Mapping Seagrass Habitats Using Airborne Digital Camera Imagery

Abstract

The purpose of this study was to map the areal extent and density of submerged aquatic vegetation, principally the seagrasses, Zostera marina and Ruppia maritima, as part of ongoing monitoring for the Barnegat Bay, New Jersey National Estuary Program. We examine the utility of multiscale image segmentation/object-oriented image classification using the eCognition software to map seagrass across our 36,000 ha study area. The multi-scale image segmentation/ object oriented classification approach closely mirrored our conceptual model of the spatial structure of the seagrass habitats and successfully extracted the features of ecological interest. The agreement between the mapped results and the original field reference was 68 percent (Kappa ϭ 56.5 percent) for the four category map and 83 percent (Kappa ϭ 63.1 percent) for the presence/absence map; the agreement between the mapped results and the independent reference data was 71 percent (Kappa ϭ 43.0 percent) for a simple presence/absence map. While the aerial digital camera imagery employed in this study had the advantage of flexible acquisition, suitable image scale, fast processing return time, and comparatively low cost, it had inconsistent radiometric response from image to image. This inconsistency made it difficult to develop a rule-based classification that was universally applicable across the 14 individual image mosaics. However, within the individual scene mosaics, using the eCognition software in a "manual classification" mode provided a flexible and time effective approach to mapping seagrass habitats.

Multiple Stable States and Catastrophic Shifts in Coastal Wetlands: Progress, Challenges, and Opportunities in Validating Theory Using Remote Sensing and Other Methods

“…The broadest extent of literature has emphasized defining individual plant reflectance spectra and optimizing spectral discrimination of vegetation classes [54][55][56][57][58][59][60][61][62][63][64]. Other studies have tested innovative approaches to classifying vegetation, including comparing use of different remote sensing data sources [65][66][67][68][69], testing object-based image analysis techniques [69][70][71][72][73][74][75], and other methods [76][77][78]. In addition to automated analysis methods, vegetation classes have long been digitized by hand [79][80][81].…”

Object-oriented classification of QuickBird data for mapping seagrass spatial structure

“…The present study utilized Definiens Developer 7 software. Previous versions of the software have already been used to analyze seabed cover based on optical (Green & Lopez 2007, Lathrop et al 2006 or acoustical data (Lucieer 2008). The workflow is as follows (Definiens Developer 7, 2007).…”

“…In order to create continuous high-resolution maps of seabed vegetation, which consists mainly of coral reefs, seagrasses and macroalgae, optical and acoustical remote sensing methods are supplemented with field surveys. Most of the recent studies are based on optical methods, analyzed satellite data received from Landsat, IKONOS and SPOT satellites (Andréfouët et al 2003, Elvidge et al 2004, Fornes et al 2006, Pasqualini et al 2005 or images from multi-spectral airborne cameras (Green & Lopez 2007, Lathrop et al 2006. The images were usually preliminarily transformed using geometric and radiometric corrections.…”

“…This method has been used to create maps of SAV but only using images obtained from airborne digital cameras. Lathrop et al (2006), following Robbins & Bell's (1994) approach, considered habitat structure at three different levels: (a) meadow -a spatially continuous area of seagrass beds of varying percent of cover composition; (b) bed -a spatially continuous area of overall similar percent of cover composition; and (c) patch -a small discrete clump of seagrass-or gapan area within a seagrass bed not occupied by plants. They applied both segmentation procedures and object-based analysis in order to aid the manual creation of a seabed cover map.…”

Application of Geobia to map the seafloor

“…Among the most frequently mapped habitats are mangroves (Conchedda et al, 2008;Heumann, 2011;Wang et al, 2004), saltmarsh (Moffett and Gorelick, 2013;Ouyang et al, 2011), seagrass (Lathrop et al, 2006;Lyons et al, 2012;Roelfsema et al, 2014) and coral reefs (Benfield et al, 2007;Knudby et al, 2011;Phinn et al, 2012).…”