Slender Object Scene Segmentation in Remote Sensing Image Based on Learnable Morphological Skeleton with Segment Anything Model
This work addresses the specific challenge of segmenting slender objects in remote sensing images for applications like urban planning and environmental monitoring, representing an incremental improvement by enhancing SAM with morphological priors.
The paper tackles the problem of slender object segmentation in remote sensing images, where existing deep learning models like SAM lose fine details during downsampling, and proposes integrating a learnable morphological skeleton prior into SAM to preserve structural details, resulting in improved performance and generalization over the original SAM on datasets including buildings, roads, and water.
Morphological methods play a crucial role in remote sensing image processing, due to their ability to capture and preserve small structural details. However, most of the existing deep learning models for semantic segmentation are based on the encoder-decoder architecture including U-net and Segment Anything Model (SAM), where the downsampling process tends to discard fine details. In this paper, we propose a new approach that integrates learnable morphological skeleton prior into deep neural networks using the variational method. To address the difficulty in backpropagation in neural networks caused by the non-differentiability presented in classical morphological operations, we provide a smooth representation of the morphological skeleton and design a variational segmentation model integrating morphological skeleton prior by employing operator splitting and dual methods. Then, we integrate this model into the network architecture of SAM, which is achieved by adding a token to mask decoder and modifying the final sigmoid layer, ensuring the final segmentation results preserve the skeleton structure as much as possible. Experimental results on remote sensing datasets, including buildings, roads and water, demonstrate that our method outperforms the original SAM on slender object segmentation and exhibits better generalization capability.