The Segment Anything Model (SAM) for Remote Sensing Applications: From Zero to One Shot
It addresses segmentation challenges in remote sensing for applications like aerial and orbital image analysis, but is incremental as it adapts an existing model to a new domain.
This study applied the Segment Anything Model (SAM) to remote sensing image segmentation, testing it with multi-scale datasets and various prompts, and improved accuracy by combining text-prompt-derived examples with one-shot training, reducing the need for manual annotation.
Segmentation is an essential step for remote sensing image processing. This study aims to advance the application of the Segment Anything Model (SAM), an innovative image segmentation model by Meta AI, in the field of remote sensing image analysis. SAM is known for its exceptional generalization capabilities and zero-shot learning, making it a promising approach to processing aerial and orbital images from diverse geographical contexts. Our exploration involved testing SAM across multi-scale datasets using various input prompts, such as bounding boxes, individual points, and text descriptors. To enhance the model's performance, we implemented a novel automated technique that combines a text-prompt-derived general example with one-shot training. This adjustment resulted in an improvement in accuracy, underscoring SAM's potential for deployment in remote sensing imagery and reducing the need for manual annotation. Despite the limitations encountered with lower spatial resolution images, SAM exhibits promising adaptability to remote sensing data analysis. We recommend future research to enhance the model's proficiency through integration with supplementary fine-tuning techniques and other networks. Furthermore, we provide the open-source code of our modifications on online repositories, encouraging further and broader adaptations of SAM to the remote sensing domain.