Fast Thick Cloud Removal for Multi-Temporal Remote Sensing Imagery via Representation Coefficient Total Variation
2023; Multidisciplinary Digital Publishing Institute; Volume: 16; Issue: 1 Linguagem: Inglês
10.3390/rs16010152
ISSN2072-4292
AutoresShuang Xu, Jilong Wang, Jialin Wang,
Tópico(s)Image Enhancement Techniques
ResumoAlthough thick cloud removal is a complex task, the past decades have witnessed the remarkable development of tensor-completion-based techniques. Nonetheless, they require substantial computational resources and may suffer from checkboard artifacts. This study presents a novel technique to address this challenging task using representation coefficient total variation (RCTV), which imposes a total variation regularizer on decomposed data. The proposed approach enhances cloud removal performance while effectively preserving the textures with high speed. The experimental results confirm the efficiency of our method in restoring image textures, demonstrating its superior performance compared to state-of-the-art techniques.
Referência(s)