A General Framework for Change Detection Using Multimodal Remote Sensing Data
Sanid Chirakkal, Francesca Bovolo, Arundhati Misra, Lorenzo Bruzzone, Avik Bhattacharya
Abstract
A general framework for change detection (CD) is proposed to analyze multi-modal remotely sensed data utilizing the Kronecker product between two data representations (vectors or matrices). The proposed method is sensor independent and provides comparable results to techniques that exist for specific sensors. The proposed fusion technique is a pixel-level approach that incorporates inputs from different modalities, rendering enriched multi-modal data representation. Thus, the proposed hybridization procedure helps to assimilate multi-sensor information in a meaningful manner. A novel change index () is defined for the general multi-modal case. This is then used to quantify the change in bi-temporal remotely sensed data. The article explores the usability, consistency, and robustness of the proposed multimodal fusion framework, including the change index, with proper validation on two multi-modal cases: (1) the dual-frequency (C- and L-band) full-polarimetric Danish EMISAR data, and (2) the dual-polarimetric SAR and multispectral data from the Sentinel constellation satellites. Detailed analysis and validation using extensive ground truth data are presented to establish the proposed framework.