[1]陈云菲,王飞,况立群,等.自适应加权组合卷积的遥感舰船目标检测方法[J].计算机技术与发展,2025,(10):71-80.[doi:10.20165/j.cnki.ISSN1673-629X.2025.0118]
　CHEN Yun-fei,WANG Fei,KUANG Li-qun,et al.Remote Sensing Ship Target Detection in Remote Sensing via Adaptive Weighted Combinatorial Convolution[J].,2025,(10):71-80.[doi:10.20165/j.cnki.ISSN1673-629X.2025.0118]
点击复制

自适应加权组合卷积的遥感舰船目标检测方法()

《计算机技术与发展》[ISSN:1006-6977/CN:61-1281/TN]

卷:
期数:
2025年10期
页码:
71-80
栏目:
媒体计算
出版日期:
2025-10-10

文章信息/Info

Title:
Remote Sensing Ship Target Detection in Remote Sensing via Adaptive Weighted Combinatorial Convolution
文章编号:
1673-629X(2025)10-0071-10
作者:
陈云菲; 王飞; 况立群; 韩燮; 郭耀武
机器视觉与虚拟现实山西省重点实验室,山西太原 030051
Author(s):
CHEN Yun-fei; WANG Fei; KUANG Li-qun; HAN Xie; GUO Yao-wu
Shanxi Provincial Key Laboratory of Machine Vision and Virtual Reality,Taiyuan 030051,China
关键词:
细粒度; 遥感舰船图像; 目标检测; 多尺度融合; 深度学习
Keywords:
fine-grained; remotely sensed ship images; object detection; multi-scale fusion; deep learning
分类号:
TP751.1
DOI:
10.20165/j.cnki.ISSN1673-629X.2025.0118
摘要:
遥感舰船目标检测与分类任务旨在准确识别和定位遥感图像中的舰船目标,从而为军事侦察、海上救援、海洋监控以及海上交通管理等领域提供有效的技术支持。为解决复杂背景和相似类别中的精细化船舶分类问题,该文提出了一种基于两阶段细粒度精细化检测模型的自适应加权组合卷积舰船目标检测方法。一方面,结合普通卷积在提取目标基本特征方面的高效性和膨胀卷积在扩大感受野、捕捉多尺度上下文信息方面的优势,实现了局部与全局特征的有效融合,加强突出舰船目标,减少背景干扰的能力。另一方面,模型通过计算方差、空间和边缘信息多尺度自适应调整特征提取,同时结合全局和历史数据优化阈值计算权重动态调整两种卷积的分配,提高模型对复杂场景的适应性。实验结果表明,该检测方法在ShipRSImageNet数据集、MAR20数据集上的平均准确率达到55.83%、59.06%,比基准模型提高了3.56百分点、1.68百分点,证明了其在复杂环境下船舶细粒度识别的准确性和有效性。
Abstract:
Remote sensing ship target detection and classification aims to accurately identify and locate ship targets in remote sensing images,thereby providing effective technical support for military reconnaissance,maritime rescue,ocean monitoring,and maritime traffic management. To address the challenge of fine-grained ship classification amid complex backgrounds and similar categories,we propose an adaptive weighted combination convolutional ship target detection method based on a two-stage fine-grained detection model. On the one hand, by integrating the efficiency of standard convolution in extracting basic target features with the advantages of dilated convolution in enlarging the receptive field and capturing multi-scale contextual information,the proposed method achieves an effective fusion of local and global features,which enhances the emphasis on ship targets while reducing background interference. On the other hand,the model adaptively adjusts feature extraction at multiple scales by computing variance, spatial, and edge information, and dynamically reassigns the convolutional operations through optimizing threshold calculation weights with the aid of both global and historical data,thereby improving its adaptability in complex scenarios. Experimental results show that the proposed method achieves average accuracies of 55.83% and 59.06% on the ShipRSImageNet and MAR20 datasets,surpassing the baseline by 3.56 percentage points and 1.68 percentage points,respectively,demonstrating its effectiveness and robustness in fine-grained ship recognition under complex scenarios.

相似文献/References:

[1]肖威 程文青 许炜.基于Web的工作流细粒度授权框架[J].计算机技术与发展,2006,(12):240.
　XIAO Wei,CHENG Wen-qing,XU Wei.A Web-Based Workflow Authorization Architecture with Fine Granularity[J].,2006,(10):240.
[2]郝小龙. 改进的RBAC模型在电网视频监控平台中的应用[J].计算机技术与发展,2014,24(12):212.
　HAO Xiao-long. Application of Improved RBAC Model in Grid Video Monitoring Platform[J].,2014,24(10):212.
[3]陈彦竹,郝天曙. 基于角色信任度动态监控的访问控制研究[J].计算机技术与发展,2017,27(10):106.
　CHEN Yan-zhu,HAO Tian-shu. Research on Access Control of Dynamic Monitoring with Role Trustrank[J].,2017,27(10):106.
[4]王继业,范 永,余文豪,等.基于细粒度访问控制的大数据安全防护方法[J].计算机技术与发展,2019,29(10):134.[doi:10. 3969 / j. issn. 1673-629X. 2019. 10. 027]
　WANG Ji-ye,FAN Yong,YU Wen-hao,et al.Big Data Security Protection Based on Fine-grained Access Control[J].,2019,29(10):134.[doi:10. 3969 / j. issn. 1673-629X. 2019. 10. 027]
[5]白瑜颖,刘宁钟,姜晓通.结合注意力混合裁剪的细粒度分类网络[J].计算机技术与发展,2021,31(10):38.[doi:10. 3969 / j. issn. 1673-629X. 2021. 10. 007]
　BAI Yu-ying,LIU Ning-zhong,JIANG Xiao-tong.Fine Grained Image Classification Network Combined with Attention CutMix[J].,2021,31(10):38.[doi:10. 3969 / j. issn. 1673-629X. 2021. 10. 007]
[6]苗壮,王培龙,崔浩然,等.融合全局上下文关联特征的细粒度图像分类[J].计算机技术与发展,2024,34(06):29.[doi:10.20165/j.cnki.ISSN1673-629X.2024.0067]
　MIAO Zhuang,WANG Pei-long,CUI Hao-ran,et al.Fine-grained Image Classification Based on Fusion of Global Contextual Features[J].,2024,34(10):29.[doi:10.20165/j.cnki.ISSN1673-629X.2024.0067]

更新日期/Last Update: 2025-10-10