A Classification-Aware Super-Resolution Framework for Ship Targets in SAR Imagery

2025-08-08

AI Summary

Low-resolution Synthetic Aperture Radar (SAR) imagery limits the accuracy of ship classification. To address this, we propose a classification-aware super-resolution (SR) framework that explicitly integrates ship classification into the SR reconstruction process for the first time. Our method formulates a multi-task joint optimization objective combining pixel-level reconstruction loss with a classification-confidence-driven perceptual loss, enabling end-to-end deep network training to simultaneously enhance image fidelity and classification discriminability. The key innovation lies in establishing a reconstruction paradigm explicitly optimized for downstream recognition performance—departing from conventional fidelity-centric SR approaches. Experiments on public SAR ship datasets demonstrate that, with only marginal improvements in PSNR (+0.12 dB) and SSIM (+0.008), our method achieves substantial gains in classification accuracy (+3.2%–5.8%), validating the effectiveness and practicality of classification-guided SR.

Technology Category

• Machine Learning: Multi-class/Multi-label Learning & Extreme Classification
• Computer Vision: Learning & Optimization for CV
• Intelligent Robots: Multimodal Perception & Sensor Fusion

Application Category

• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs
• User Modeling, Personalization and Recommendation: Fairness-aware retrieval and ranking

Abstract

High-resolution imagery plays a critical role in improving the performance of visual recognition tasks such as classification, detection, and segmentation. In many domains, including remote sensing and surveillance, low-resolution images can limit the accuracy of automated analysis. To address this, super-resolution (SR) techniques have been widely adopted to attempt to reconstruct high-resolution images from low-resolution inputs. Related traditional approaches focus solely on enhancing image quality based on pixel-level metrics, leaving the relationship between super-resolved image fidelity and downstream classification performance largely underexplored. This raises a key question: can integrating classification objectives directly into the super-resolution process further improve classification accuracy? In this paper, we try to respond to this question by investigating the relationship between super-resolution and classification through the deployment of a specialised algorithmic strategy. We propose a novel methodology that increases the resolution of synthetic aperture radar imagery by optimising loss functions that account for both image quality and classification performance. Our approach improves image quality, as measured by scientifically ascertained image quality indicators, while also enhancing classification accuracy.

Problem

• Enhancing SAR image resolution for better classification accuracy
• Exploring SR-classification relationship via integrated algorithmic strategy
• Optimizing loss functions for both image quality and classification

Innovation

• Classification-aware super-resolution for SAR imagery
• Joint optimization of image quality and classification
• Enhanced resolution and classification accuracy simultaneously