4. A multiple classifier system to improve classification accuracy of wetlands using SAR data

So far, numerous classification algorithms have been developed to classify various land covers, each containing its own advantages and limitations. Random Forest algorithm has proved its high potential for wetland classification in many studies (e.g. [40, 26, 61]). However, the most promising approach to obtain a high classification accuracy is fusing different classifiers in a way that the advantages of each are ensembled. The obtained ensemble classifier is called multiple classifier system (MCS [38, 95]). The system is more important when classifying complex landscapes, such as wetlands, because achieving high accuracy for individual classes is significantly challenging in these cases. This becomes even more serious when only SAR data are applied for discriminating wetlands. There are several studies which developed new MCSs to improve the classification accuracy of similar

Landsat-8, Sentinel-2A, and ASTER) and SAR (Sentinel-1, RADARSAT-2, and ALOS-2) data in NL, Canada. As is clear from this table, the ratio features provided the highest separability

of HH/HV and HH/TP obtained from RADARSAT-2 full-polarimetric data are the most impor-

Comparing the optical spectral bands, the NIR and Red Edge bands are most effective for discriminating wetland classes. Two main characteristics of wetlands are vegetation and water, which can be efficiently studied by these two bands. This demonstrates that it is more efficient to use the optical satellites, in which both NIR and Red Edge bands are included (e.g. Sentinel-2A and RapidEye). In this regard, Sentinel-2A, which provides free imagery, is superior for employment in operational wetland mapping and monitoring. The red band is also helpful in separating wetlands, especially discrimination between bog and other wetlands, because of bogs' red appearance. Additionally, there is a high overlap between the spectral signatures of wetlands in the green, SWIR, and TIR bands, and thus, there is a difficulty in using these bands for wetland studies. Finally, the blue band is not very useful

Comparing various decomposition methods, including Freeman-Durden, Cloude-Pottier, Touzi, Van Zyl, Yamaguchi, and Krogager, it is observed that coherent decomposition techniques, such Krogager, are not recommended for wetland classification. The reason is that the coherent decompositions are mostly applicable for detecting man-made features in urban areas and less useful for naturally distributed targets such as wetland classes [93]. In addition, the Cloude-Pottier and Freeman-Durden methods are most optimum for separating wetland species. In this regard, the volume scattering component of Freeman-Durden and Anisotropy element of Cloude-Pottier are generally the best. Moreover, some SAR features extracted from the eigenvalue/eigenvector of the coherency matrix demonstrated a high potential for separating wetland class pairs and all wetland classes. In this regard, the serd, normalized serd and normalized derd, introduced by [94], are frequently selected for wet-

4. A multiple classifier system to improve classification accuracy of

So far, numerous classification algorithms have been developed to classify various land covers, each containing its own advantages and limitations. Random Forest algorithm has proved its high potential for wetland classification in many studies (e.g. [40, 26, 61]). However, the most promising approach to obtain a high classification accuracy is fusing different classifiers in a way that the advantages of each are ensembled. The obtained ensemble classifier is called multiple classifier system (MCS [38, 95]). The system is more important when classifying complex landscapes, such as wetlands, because achieving high

Brightness ratios are most efficient regarding the optical data, and the ratios

measures. NIR

in most of the cases.

lands separation.

wetlands using SAR data

Brightness and Red Edge

tant SAR features for separating wetlands.

120 Wetlands Management - Assessing Risk and Sustainable Solutions

Figure 3. Proposed multiple classifier system by Amani et al. [38] to improve the classification accuracy of the complex environments.

landscapes (e.g. [96, 97]). Regarding wetland classification, Amani et al. [38] proposed a novel MCS to increase wetland classification accuracy using only SAR data in NL, Canada, in terms of both individual class and overall accuracies. The system initially removes poor classifiers and selects the best classification algorithm to identify each wetland class. Then, the final label is selected for each random pixel/object using the class label decision criteria introduced by the authors. The flowchart of the proposed MCS along with the corresponding criteria is illustrated in Figure 3. The proposed MCS outperformed the single classifiers and produced the highest producer and user accuracies for almost all wetland and non-wetland classes. It also increased the overall classification accuracy and kappa coefficient by 5–8 and 9–16%, respectively.

Conflict of interest

for reviewing this chapter.

Sahel Mahdavi<sup>1</sup> and Brian Brisco<sup>2</sup>

Author details

Bahram Salehi<sup>1</sup>

References

Thanks

The authors declare no conflict of interest.

\*, Masoud Mahdianpari<sup>1</sup>

2 Canada Centre for Remote Sensing, Ottawa, Ontario, Canada

\*Address all correspondence to: bsalehi@esf.edu

DOI: 10.1080/15481603.2017.1419602

cation. Wetland Science. 2010;8:299-304

Wildlife Service; 1979. 131 p

Thanks to Brian Huberty, U.S. Fish & Wildlife Service, National Wetland Inventory program

, Meisam Amani1

[1] Kingsford RT, Basset A, Jackson L. Wetlands: Conservation's poor cousins. Aquatic

[2] Pan Y, Birdsey RA, Phillips OL, Jackson RB. The structure, distribution, and biomass of the world's forests. Annual Review of Ecology, Evolution, and Systematics. 2013;44:593-622

[3] Mahdavi S, Salehi B, Granger JE, Amani M, Brisco B, Huang W. Remote sensing for wetland classification: A comprehensive review. GIS & Remote Sensing. 2018;55:623-658.

[4] National Wetlands Working Group. Wetlands of Canada. Eco-logical Land Classifi cation Series, No. 24. Sustainable Development Branch, Environment Canada, Ottawa,

[5] Hongjun C, Guoping W, Xianguo L. Wetland definitions: Creation, evolution and appli-

[6] Cowardin LM, Carter V, Golet FC, LaRoe ET. Classification of Wetlands and Deepwater Habitats of the United States. Washington, DC: U.S. Department of Interior, Fish and

1 State University of New York, College of Environmental Science and Forestry

Conservation: Marine and Freshwater Ecosystems. 2016;6:892-916

Ontario, and Polyscience, Montreal, Québec. 1988. 452 p

, Fariba M. Manesh1

A Collection of Novel Algorithms for Wetland Classification with SAR and Optical Data

http://dx.doi.org/10.5772/intechopen.80688

, Jean Granger<sup>1</sup>

,

123
