Abstract:Synthetic aperture radar (SAR) images have been one of the important tools to support earth observations and topographic measurements. It means that SAR images are essentially rich in structures. However, the single spatial relationship is difficult to deal with the heterogeneous structures of the SAR images. In this paper, we propose an adaptive hierarchical multinomial latent model with hybrid kernel function for SAR image semantic segmentation. In the proposed approach, we design a hybrid kernel function combing Gaussian radial basis function (GRBF) and ridgelet kernel function to adaptively describe the spatial relationships between the central pixel and the surrounding pixels. Then, based on the hybrid kernel function, adaptive methods are proposed for semantic segmentation. Specifically, an SAR image is divided into different characteristics subspaces, homogeneous, structural, and aggregated subspaces, by SAR hierarchical semantic model. For the homogeneous subspace, GRBF is used to describe the isotropic spatial relationships. Then, multilayer multinomial latent model with GRBF is used for segmentation to improve the labeling consistency and reduce the wrong segmentation. For the structural subspace, the ridgelet kernel function is used to describe the anisotropic spatial relationships. Then, we adopt the singlelayer multinomial latent model with ridgelet kernel function for segmentation to preserve the details (such as edge, lines, and small objects). For aggregated subspace, bag-of-words model is used to extract the features of the aggregated portions, and then affinity propagation cluster is used for segmentation. Finally, the segmentation results of different subspaces are integrated together to obtain the final segmentation result. Comprehensive experiments on both synthetic and real SAR images demonstrate that the segmentation results by our proposed approach achieve the semantic consistency, labeling consistency, and detail preservation simultaneously.

SAR Image Compression Using Discretized Gaussian Adaptive Model and Generalized Subtractive Normalization

Content-Adaptive Residual Learning and Context-Aware Entropy Model for SAR Image Compression

Synthetic Aperture Radar Image Compression Based on a Variational Autoencoder

Synthetic Aperture Radar Image Compression Based on Low-Frequency Rejection and Quality Map Guidance

SAR Image Compression With Inherent Denoising Capability Through Knowledge Distillation

Dynamic Range Compression Self-Adaption Method for SAR Image Based on Deep Learning

Learned Compression Framework With Pyramidal Features and Quality Enhancement for SAR Images

A Two-Stage Method for Joint Denoising and Compression of SAR Images

End-to-End SAR Deep Learning Imaging Method Based on Sparse Optimization

Adaptive Hierarchical Multinomial Latent Model with Hybrid Kernel Function for SAR Image Semantic Segmentation

Conditional Diffusion for SAR to Optical Image Translation

Learning-Based Sparse Recovery Algorithm for 3D SAR Imaging

A novel lossless compression encoding framework for SAR remote sensing images

Compressed Sensing SAR Imaging Based on Centralized Sparse Representation

Synthetic Aperture Radar Deep Statistical Imaging Through Diffusion Generative Model Conditional Inference

Nonsparse SAR Scene Imaging Network Based on Sparse Representation and Approximate Observations

SAR Nonsparse Scene Reconstruction Network via Image Feature Representation Learning

Target Oriented High Resolution SAR Image Formation via Semantic Information Guided Regularizations

Remote Sensing Image Recognition Based on Generalized Regularized Auto-Encoders

A SAR-to-Optical Image Translation Method Based on Conditional Generation Adversarial Network (cGAN)