Search

Classification of Polarimetric Synthetic Aperture Radar (PolSAR) images is an incredibly challenging task. In early days researchers have applied parametric classifiers like maximum likelihood and Wishart classifiers. Later it is found that each class may or may not follow the Gaussian distribution hence researchers have used nonparametric classifiers like ANN, Decision Tree and SVM etc. It is observed from their work that these classifiers give good classification accuracy, but a lot of time is consumed in extracting the features and training the model. To avoid this, in this paper, Semantic Segmentation with the U-net architecture is used. Compared to covariance or coherency matrix, target decomposition techniques provide more information about scattering mechanisms. Hence, the effect of Touzi and Gulab Singh's 4- (G4U), 7- component (7SD) target decomposition techniques is studied on Semantic Segmentation. Comparative results show that G4U and 7SD decomposition gives almost similar classification accuracy. It is also observed that out of many of Touzi parameters, the first and second components of α, Φ, λ and τ can effectively classify various features and the classification accuracy is comparable with G4U and 7 SD decomposition. Moreover, the classification accuracy using nine elements of the coherency matrix is less than all the above-mentioned decomposition techniques. [ABSTRACT FROM AUTHOR]

For a long time, Polarimetric Synthetic Aperture Radar (PolSAR) data was not available free of cost, so the applications were limited. With the recent increase in availability of PolSAR data due to missions like ALOS-2, UAVSAR and Sentinel, the data can be acquired through these sensors periodically. Since the volume of data is large, applying traditional classifiers and conventional machine learning algorithms becomes confounding, as in that case, manual feature extraction must be done by the researcher for training the model. To automate this feature extraction step and accelerate the process, many researchers have used various deep learning algorithms. However, most studies fail to tap into the potential of PolSAR data by not utilizing the complete range of complex data of float-32 bit-depth. The work in this paper suggests the development of a generalized deep learning model for nine elements of coherency matrix [ T 3 ] in the float-32 data-space from ALOS/PALSAR-2 L-Band sensor. Semantic segmentation is used to classify land-covers into various classes like water, settlement, forest, open land, wetlands etc. This is done through modification on the UNet backbone. The deep learning model is trained using data on San Francisco and New Delhi regions and tested on data from Mumbai region. It is observed that the classification accuracy for Mumbai region is " 93.82 % ". This kind of system can assist the decision makers like urban planners to take informed decisions. [ABSTRACT FROM AUTHOR]