MaskSCD improves land cover change maps from satellite images
MaskSCD: Bitemporal Mask Classification for Semantic Change Detection in Remote Sensing Images
Computer Vision and Pattern Recognition
Summary
Semantic change detection finds where and how land cover has changed over time using satellite images. The authors introduce MaskSCD, a new method that looks at regions across two images taken at different times and classifies them together instead of pixel by pixel. This approach ensures that detected changes match actual land-cover transitions and avoids conflicting labels. Tested on three datasets with different resolutions, MaskSCD showed better accuracy than recent methods and eliminated contradictory change predictions.
What this means in practice
- •For remote sensing teams: Produce more accurate and consistent maps of land-cover changes using satellite image pairs for environmental monitoring.
- •For urban planning departments: Detect and analyze semantic land-use changes over time with fewer errors for better city and land management decisions.
Authors
Weikang Yu, Pedram Ghamisi
Abstract
Semantic change detection (SCD) identifies where the land surface has changed and which land-cover transition has taken place. Most deep SCD networks predict the change map and the land-cover map of each date with separate per-pixel decoders. These outputs need not agree, so a pixel can be labeled as changed although it has the same class at both dates, and per-pixel decisions fragment objects. We present MaskSCD, which formulates SCD as the mask classification of a bitemporal label-pair image. Each bitemporal segment, a region of constant land-cover class pair, is predicted as one mask shared by both dates together with one class per date, and the change map is derived from the predicted class pair. This formulation keeps the change and semantic maps consistent and the label space linear in the number of classes. A bitemporal masked-attention transformer decoder implements it by reading each segment in both images and relating the two readings, and a bitemporal fusion pyramid provides per-date and change-aware features. MaskSCD is evaluated on SECOND, Landsat-SCD and Hi-UCD, which span 0.1 to 30 m resolution and four to nine land-cover classes, against fourteen recent SCD models re-trained under one protocol. MaskSCD achieves the best separated kappa, semantic change $F_1$ and transition $F_1$ on all three datasets and exceeds the strongest competitor by about five points in mean separated kappa. It produces no contradictory change pixels, which account for nearly 30% of the predicted change of some multi-task networks. The code is available at https://github.com/EricYu97/MaskSCD.