Adaptive Local Structure Consistency-Based Heterogeneous Remote Sensing Change Detection

Change detection (CD) of heterogeneous remote sensing images is a challenging topic, which plays an important role in natural disaster emergency response. Due to the different imaging mechanisms sensors, it hard directly compare images. To address this challenge, we explore unsupervised CD method based on adaptive local structure consistency (ALSC) between letter, constructs graph representing for each patch one image domain and then projects other measure change level. This exploits fact that share same information ground object, modality-invariant. avoid data confusion, pixelwise calculated by projection. By comparing with some state-of-the-art methods, experimental results show effectiveness proposed ALSC-based method.