Rinki Deo

Classification Accuracy of Multi-Frequency and Multi-Polarization SAR Images for Various Land Covers

This paper presents the land cover classification capabilities of fully versus partially polarimetric SAR data for C- and L-band frequencies. Maximum Likelihood classifier with complex Wishart distribution and artificial neural network classifier (ANN) have been used for classification. The change in accuracy due to the phase information of SAR data is also assessed by comparing the classified results of intensity and complex images for all the possible polarization combinations at L- and C-band. In all the combinations, fully polarimetric data provides highest accuracy and it is not much different from that of complex partial polarimetric (HH, VV) combination. The accuracies obtained with various partial polarimetric combinations are dependent on the land cover types. Among L-, C- and X-bands, L-band offers better accuracy. By combining all bands data, accuracy improved by 7%.The accuracy has been improved slightly by combining the three components of van Zyl decomposition with the combination of X-, C-and L-band. IRS-P6 optical data over the same area has been used to compare the classification accuracy between optical and SAR data.

Framework for Fusion of Ascending and Descending Pass TanDEM-X Raw DEMs

A novel method for calculating optimum incidence angle for the TanDEM-X system using any available digital elevation model (DEM) for the given area is proposed in this study. This method includes the plotting of slopes and aspect of the test area in a statistical way and applying mathematical approach using acquisition geometry in ascending and descending pass TanDEM-X data to optimize the incidence angle for obtaining precise DEM. Furthermore, the TanDEM-X raw DEMs in ascending and descending pass over Mumbai, India are combined using a simple weighted fusion algorithm and the quality of fused DEM thus generated is enhanced. The method adopted for fusion is just an experimental study. The problem of optimum weight selection for fusion has been addressed using height error map and a robust layover shadow mask calculated in “Integrated TanDEM-X Processor” (ITP) during TanDEM-X DEM generation. The height error map is calculated from the interferometric coherence with geometrical considerations and the robust layover and shadow map is calculated using TanDEM-X DEM and the corresponding slant range. Results show a significant reduction in the number of invalid pixels after fusion. In the fused DEM, invalids are only 2.14%, while ascending and descending pass DEMs have 5.02% and 6.34%, respectively. Statistical analysis shows a slight improvement in standard deviation of the error in fused DEM by 8% in urban area and about 5% for the whole scene. Only slight improvement in accuracy of fused DEM can be attributed to the coarse resolution of the SRTM-X DEM used as reference.

Fusion of ascending and descending pass raw TanDEM-X DEM

This paper deals with the fusion of TanDEM-X raw DEMs in ascending and descending pass over Mumbai test area and enhance its quality. Before applying fusion method, a robust layover and shadow map has been calculated in ITP using TanDEM-X DEM and the corresponding slant range image. The selection of optimum weights for fusion has been based on height error map calculated from interferometric coherence. Results show a substantial reduction in number of invalid pixels after fusion. In the fused DEM, invalid is only 1.2%, while ascending and descending pass DEMs have 6.7% and 5.7% respectively. The improvement in accuracy of the DEM is very slight in this case which is due to the coarse resolution of the SRTM DEM used as reference.

Comparison of classification accuracy between fully polarimetric and dual-polarization SAR images

Polarimetric Synthetic Aperture Radar (PolSAR) data is available at different frequencies and polarizations from various sensors like ALOS-PALSAR, Envisat ASAR, TerraSAR-X. This study compares the classification accuracies obtained with fully polarimetric and dual-polarization L-band ALOS-PALSAR data over Mumbai and Sundarban area. We have also compared dual polarized ALOS-PALSAR L-band TerraSAR-X X-band and Envisat C-band SAR data acquired over Mumbai area. IRS-P6 optical data over the same area has been used to compare the classification accuracy between optical and SAR data. The change in accuracy due to the phase information of SAR data is also assessed by comparing the classified results of intensity and complex images for all the possible polarization combinations such as (HH, HV), (HH, VV) and (HV, VV) at L-band. The results show that the fully polarimetric mode gives maximum classification accuracy (above 95%). It is observed that among dual polarized data, complex (HH, VV) combination gives the maximum (above 92%) accuracy.

Application of persistent scatterer interferometry for identification of landslide areas of Himalayan region

ENVISAT ASAR images acquired during 2003-2007 were processed for landslide movement using PSInSAR and SBAS methods over Himalayan region. The mean velocity of the landslide movement obtained with ascending and descending pass data using PSInSAR is around ± 16 mm/yr and SBAS method gives -25 to 15 mm/yr with descending pass and -39 to 38 mm/yr with ascending pass data sets. The number of PS points found with SBAS method is much less than that of obtained using PSInSAR. The effectiveness of the result shows highest movement in steeply sloped region.

Comparison of TanDEM-X and Cartosat-1 stereo DEMs over different terrains of India

In this paper, the accuracy of TanDEM-X DEMs is evaluated for different terrains of India and is also compared with that of Cartosat-1 optical stereo DEMs. Using accurate GPS points as reference, TanDEM-X DEMs of Mumbai area with flat as well as low hilly terrain, Koyna area with high hilly terrain, rugged Himalayan terrain of Manali and Katerniaghat area with forest cover over flat terrain were evaluated. The results show an RMSE of 3.3 m, 4.9 m, 12.8 m and 8.7 m for the four test areas respectively. Cartosat-1 DEM over these areas give an RMSE of 4.8 m, 7.9 m, 11.4 m and 8.7 m respectively. The height error also shows its dependence on slope. The difference between the DEMs obtained from the two techniques was also calculated for all the test sites in order to compare their accuracies. For all the test sites except Manali, an RMSE <; 4 m with 90% confidence level was observed between the two DEMs. The rugged terrain of Manali area is highly affected due to layover and shadow effect in TanDEM-X DEM and hence showed higher RMSE when compared with Cartosat-1 DEM. As both DEMs have spatial resolution and high accuracy, gaps in the TanDEM-X DEM may be filled with Cartosat-1 DEM.

Matlab based SAR signal processor for educational use

This paper is an attempt to help the beginners and the students to learn the basics of the SAR signal processing in a simple and easy way using MATLAB. The signal processing of SAR data is very important for generating various data products for different applications and analysis of land features. The SAR raw data for signal processing is a two dimensional array which contain sampled echoes in the complex form. Pulse compression techniques are used to compress energy of echoes and thereby increasing the resolution of SAR. To obtain the final image the range compression, Doppler centroid frequency estimation, range cell migration correction, azimuth compression etc. are implemented in such a way that any user can understand very easily. Advantage of the developed code in MATLAB is that it does not need a vast programming knowledge to be able to customize it. The developed code will be available to all users for educational and training purposes.