The paper addresses subsurface anomaly detection using radar data. It proposes to enhance the Segment Anything Model with reservoir computing, combining visual segmentation with wave analysis. This approach improves the accuracy of detecting and categorizing underground issues while reducing manual effort.
- Xiren Zhou
- Shikang Liu
- Huanhuan Chen