航空学报 > 2026, Vol. 47 Issue (8): 332780-332780   doi: 10.7527/S1000-6893.2025.32780

基于距离损失提升航空图像语义分割研究

马宇卓, 任侃(), 李涛, 陈钱   

  1. 南京理工大学 电子工程与光电技术学院,南京 210094
  • 收稿日期:2025-09-12 修回日期:2025-10-07 接受日期:2025-10-10 出版日期:2025-10-30 发布日期:2025-10-17
  • 通讯作者: 任侃 E-mail:k.ren@njust.edu.cn
  • 基金资助:
    国家自然科学基金(62175111);江苏高校“青蓝工程”工程(2024)

Improving remote sensing image semantic segmentation based on distance loss

Yuzhuo MA, Kan REN(), Tao LI, Qian CHEN   

  1. School of Electronic and Optical Engineering,Nanjing University of Science and Technology,Nanjing 210094,China
  • Received:2025-09-12 Revised:2025-10-07 Accepted:2025-10-10 Online:2025-10-30 Published:2025-10-17
  • Contact: Kan REN E-mail:k.ren@njust.edu.cn
  • Supported by:
    National Natural Science Foundation of China(62175111);Qing Lan Project(2024)

摘要:

深度学习和计算机视觉技术的进展对航空遥感领域产生了深远的影响,使得对航空图像的分析变得更加高效。与常规图像相比,航空图像的目标边界更清晰明显、分布更规律,且具有更强的空间结构性。然而,当前的先进分割方法主要集中于利用复杂的特征提取器以捕捉更强的上下文关系,更多关注单像素分类准确度,这不仅对硬件要求较高,而且忽视了从结构层面进行边界对齐的问题。为了应对这一挑战,提出了一种创新的边界感知损失函数——Lossd,旨在提升航空遥感图像语义分割的性能,尤其在边界精度和目标分割一致性方面。创新性地将结构差异转化为损失,而非传统方法侧重关注的单像素准确性。此外,针对语义分割任务中常见的过切和少切问题,提出了有效的解决方案,并在3个大规模使用的数据集和3个基准模型上进行了广泛的实验验证。实验结果表明,文中方法在不修改原有模型的前提下,显著提升了模型的语义分割性能,在LoveDA上实现了55.8% mIoU(+1.6%),在Uavid上实现了70.8% mIoU(+0.8%),在Potsdam上实现了94.1% mF1(+0.7%),接近并部分超越了当前主流的方法。

关键词: 航空图像, 深度学习, 语义分割, 损失函数, 边界约束

Abstract:

The advancements in deep learning and computer vision technologies have had a profound impact on the field of air-borne remote sensing, making the analysis of aerial images more efficient. Compared to conventional images, the target boundaries in aerial images are clearer and more distinct, with more regular distributions and stronger spatial structure. However, current state-of-the-art segmentation methods mainly focus on utilizing complex feature extractors to capture stronger contextual relationships, placing more emphasis on single-pixel classification accuracy. This not only demands higher hardware requirements but also overlooks the issue of boundary alignment from a structural perspective. To address this challenge, we propose an innovative boundary-aware loss function, Lossd, designed to enhance the performance of semantic segmentation for aerial remote sensing images, particularly in terms of boundary precision and target segmentation consistency. We innovatively translate structural differences into a loss, unlike traditional methods that focus on single-pixel accuracy. Moreover, we propose an effective solution for the common over-segmentation and under-segmentation problems in semantic segmentation tasks. Extensive experimental validation has been conducted on three widely used large-scale datasets and three benchmark models. Experimental results show that our method significantly improves the semantic segmentation performance without modifying the original network. Specialty, our method achieves 55.8% mIoU (+1.6%) on LoveDA, 70.8% mIoU (+0.8%) on UAVid, and 94.1% mF1 (+0.7%) on Potsdam, reaching or partially surpassing the performance of mainstream approaches.

Key words: aerial images, deep learning, semantic segmentation, loss function, boundary constraints

中图分类号: