Remote Sensing Semantic Change Detection Model for Improving Objects Completeness

Abstract

Semantic change detection (SCD) extends beyond binary change detection by not only discerning the locations of change areas, but also offering the alterations in land-cover/land-use types. This refined change information is concernful for various applications. While deep learning methods have made significant progress in SCD, accurately capturing the integrity of targets remains challenging in intricate scenarios. Therefore, this article proposes a deformable multiscale composite transformer network (DMCTNet). This network effectively models relevant semantic information and spatio-temporal dependencies. DMCTNet leverages a variant vision foundation models encoder to learn specific knowledge, facilitating effective visual representation in remote sensing images. A multiscale feature aggregator module is developed to discern both the “what” and “where” of changes by integrating features across different scales. Subsequently, a masked decoder through queries to convey rich semantic change information, guided by conspicuous change potential locations to decode. A substantial volume of experimental results consistently demonstrate that this model achieves more accurate and reliable results in change areas, improving the intersection over the union of change by 2.65% and 1.67% on the SECOND and Landsat-SCD datasets, respectively. Code will be made available.

Keywords

• Remote sensing (RS)
• semantic change detection (SCD)
• semantic segmentation
• visual foundation models (VFMs)