Automatic Adaptive Signature Generalization in R
Abstract
The automatic adaptive signature generalization (AASG) algorithm overcomes many of the limitations associated with classification of multitemporal imagery. By locating stable sites between two images and using them to adapt class spectral signatures from a high-quality reference classification to a new image, AASG mitigates the impacts of radiometric and phenological differences between images and ensures that class definitions remain consistent between the two classifications. Here, I provide source code (in the R programming environment), as well as a comprehensive user guide, for the AASG algorithm. See Dannenberg, Hakkenberg and Song (2016) for details of the algorithm. Dannenberg, MP, CR Hakkenberg, and C Song (2016), Consistent classification of Landsat time series with an improved automatic adaptive signature generalization algorithm, Remote Sensing 8(8): 691.
File (1)
Content uploaded by Matthew Dannenberg
Author content
Supplementary resource (1)
... Once stable sites located, class spectral signatures wereextracted from the target image,labelled to their respective classes using the reference map andused as input into a Maximum Likelihood classifier, as well as the three depth-invariant bands, to classify the target image. Class specific signature extraction and target image classification were conducted in the software R, v.3.3.1 (R Core Team 2017) using the AASG, raster and rgdal packages (Hijmans et al. 2015;Dannenberg et al. 2017;Bivand et al. 2018). ...
The recent world-wide loss of seagrasses, which are critical components of coastal ecosystems, has ignited an effort among scientists and resource managers to develop effective monitoring tools. Although Landsat time-series is considered one of the most cost-effective options for monitoring landscapes, its application to monitor seagrasses remains scarce due to many factors including difficulties obtaining accurate ground-truth data and perceived limitations in mapping nearshore marine ecosystems. Here, we report on the use of archived Landsat multispectral imagery and the automatic adaptive signature generalization (AASG) to evaluate eelgrass (Zostera marina) distribution and abundance between 1984 and 2017, in an estuary located in northeastern New Brunswick, Canada. The AASG algorithm, a novel cost-efficient approach for satellite imagery time-series analysis that requires limited ground truth data, was used to produce fourteen maps, four of which had accuracies ranging from 75 to 85%. The results indicated that eelgrass meadows near the beach barrier were highly dynamic, exhibiting high abundance fluctuations between years and a conversion of dense eelgrass to medium-low eelgrass near the main coastline. This study demonstrates the feasibility of using the AASG algorithm to map seagrass and advantages of including satellite time-series in monitoring programmes to investigate seagrass dynamics and long-term trends.
ResearchGate has not been able to resolve any references for this publication.