Quantitative comparison of classification capability: fully-polarimetric versus partially polarimetric SAR

Lee, J.S.; Grunes, M.R.; Ainsworth, T.L.; Pottier, Éric; Krogager, E.; Boerner, W.-M.

doi:10.1109/igarss.2000.858035

Cited by 28 publications

(28 citation statements)

References 0 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Or using more descriptors, the polarimetric information is converted into three parameters (H, α and Anisotropy A) to which a physical interpretation is associated [15]. Some other classification methods are based on statistical characteristics of data and coherency matrix and derive a distance measure based on the complex Wishart distribution [7] or the complex Gaussian distribution. We use it for maximumlikelihood (ML) classification of single-look polarimetric SAR data.…”

Section: Methodsmentioning

confidence: 99%

“…For that purpose, an adapted likelihood distance measurement based on the previous statistical analysis is introduced in a Maximum Likelihood algorithm. In order to assess the performance of our method, our result is compared with other results from supervised classification such as SVM [6] or supervised ML Wishart classification [7]. The experiments are held on a POLSAR image provided by the RAMSES SAR sensor of ONERA over a suburban area.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Maximum Likelihood Classification of High-Resolution Sar Images in Urban Area

Majd¹,

Simonetto²,

Praly³

2012

Int. Arch. Photogramm. Remote Sens. Spatial Inf. Sci.

View full text Add to dashboard Cite

ABSTRACT:In this work, we propose a state-of-the-art on statistical analysis of polarimetric synthetic aperture radar (SAR) data, through the modeling of several indices. We concentrate on eight ground classes which have been carried out from amplitudes, co-polarisation ratio, depolarization ratios, and other polarimetric descriptors. To study their different statistical behaviours, we consider Gauss, lognormal, Beta I, Weibull, Gamma, and Fisher statistical models and estimate their parameters using three methods: method of moments (MoM), maximum-likelihood (ML) methodology, and log-cumulants method (MoML). Then, we study the opportunity of introducing this information in an adapted supervised classification scheme based on Maximum-Likelihood and Fisher pdf. Our work relies on an image of a suburban area, acquired by the airborne RAMSES SAR sensor of ONERA. The results prove the potential of such data to discriminate urban surfaces and show the usefulness of adapting any classical classification algorithm however classification maps present a persistant class confusion between flat gravelled or concrete roofs and trees.

show abstract

Section: Methodsmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Maximum Likelihood Classification of High-Resolution Sar Images in Urban Area

Majd¹,

Simonetto²,

Praly³

2012

Int. Arch. Photogramm. Remote Sens. Spatial Inf. Sci.

View full text Add to dashboard Cite

show abstract

“…A total of 297 patches were collected from the ground truth map to train the decision tree classifier. The inferred decision tree Downloaded by [New York University] Lee et al (2001). is shown in Figure 5, and the corresponding mutually exclusive rule set is listed in Table 1.…”

Section: Decision Tree Classificationmentioning

confidence: 99%

Land Cover Classification from Polarimetric SAR Data Based on Image Segmentation and Decision Trees

Zhang

et al. 2015

Canadian Journal of Remote Sensing

View full text Add to dashboard Cite

“…Wang et al (2009) briefly introduced the Phased Array type L-band Synthetic Aperture Radar (PALSAR) on board Advanced Land Observing Satellite (ALOS), and RADARSAT-2 operated by the Canadian Space Agency (CSA) and MacDonald Dettwiler and Associates Ltd (MDA). Both could provide fully polarized SAR data to support PolSAR (Polarimetric SAR) technology (i.e., PolSAR decomposition), which has achieved promising results in many environmental researches (e.g., Lee et al, 2001;McNairn et al, 2009;Shimoni et al, 2009). Light detection and ranging (Lidar), also called Laser altimetry, is an active remote sensing technology that utilizes a laser to illuminate a target object and a photodiode to register the backscatter radiation (Lim et al, 2003;Hyyppa et al, 2009).…”

Section: Radar and Lidarmentioning

confidence: 99%