On the Effect of Imperfect Reference Images in SAR Change Detection Based on Bayes' Theorem
Lucas Pedroso Ramos, Rômulo Fernandes da Costa, Diego da Silva de Medeiros, Paulo Branco da Silva, Dimas Irion Alves, Renato Machado
Evento: XL Simpósio Brasileiro de Telecomunicações e Processamento de Sinais (SBrT2022)
Keywords: Synthetic Aperture Radar (SAR) ground scene prediction (GSP) wavelength-resolution SAR images change detection method
AbstractThis paper presents a study regarding the impact of imperfect reference images containing targets and image formation issues in the performance of change detection methods for wavelength-resolution SAR images. The presented analysis uses a change detection method based on Bayes' Theorem recently proposed. The experimental evaluation is carried out using real data obtained using the CARABAS II SAR system. The experimental setup is evaluated using imperfect reference images and considering images obtained using ground scene prediction (GSP) methods. The GSP-generated images tend not to contain targets and are not significantly affected by image formation issues. Results indicate that the use of reference images obtained by using GSP provided a false alarm reduction in the evaluated scenarios when compared with the CDA implementation associated with imperfect reference images. For instance, in an evaluated setup in this paper, a FAR reduction from 0.667 to 0.229 is observed.