Fusion-based approach to change detection to reduce the effect of the trade-off parameter in the active contour model

This study proposes an approach to unsupervised change detection in which two different change maps are fused using different trade-off parameters of an active contour model. First, the change vector analysis method is conducted to produce a difference image from multitemporal and multispectral remotely sensed images. Second, two change maps are obtained based on the difference image using an active contour model using two different values of the trade-off parameter. Finally, an advantage fusion strategy is proposed to yield a final change map by fusing the two change maps, thereby reducing false alarms and preserving the accurate outlines of the changed regions. Two experiments are conducted with Landsat-7 Enhanced Thematic Mapper Plus and Landsat-5 Thematic Mapper data sets to evaluate the performance of the proposed method. Results confirm the effectiveness of the proposed approach vis-à-vis some of the state-of-the-art methods. This work contributes to the reduction of the effect of the trade-off parameter on the accuracy of the change map.