Analysis of Complex Valued Neural Network for SAR Image Classification
Synthetic Aperture Radar (SAR) is one of the vital sensors in the field of remote sensing. Its day, night and all-weather observation capability make it unique as compared to other passive sensors. Nature of SAR data is complex, to process such kind of data in complex domain is essential to preserve its physical nature. A complex valued neural network (CVNN) has been utilized to process the images in the complex domain. A full polarimetric data set ALOS-PALSAR 1 has been utilized to check the robustness of the proposed algorithm. An accuracy of 99.25% has been found on full polarimetric data set which is better as compared to different mode of polarization and other approaches.
Authors
- Mayank Swarnkar (ORCID: https://orcid.org/0000-0001-7351-0341)
- Gaurav Kumar Dashondhi (ORCID: https://orcid.org/0000-0002-7075-3919)
- Girjesh Dasaundhi (ORCID: https://orcid.org/0009-0007-2934-7904)
- P. V. Arun
Institutions
- Indian Institute of Technology Bombay (IN)
- Bennett University (IN)
- Indian Institute of Technology BHU (IN)
- International Institute of Information Technology, Hyderabad (IN)
Publication Details
- Published
- 2024-12-02
- DOI
- https://doi.org/10.1109/ingarss61818.2024.10984177
- Primary Topic
- Neural Networks and Applications
- Type
- article
- Field-Weighted Citation Impact
- 0.00