[1]赵振博,付天怡,董红斌,等.基于提案增强的解耦特征挖掘旋转检测器[J].智能系统学报,2025,20(5):1123-1135.[doi:10.11992/tis.202410017]
ZHAO Zhenbo,FU Tianyi,DONG Hongbin,et al.Decoupled feature mining rotational detector based on proposal enhancement[J].CAAI Transactions on Intelligent Systems,2025,20(5):1123-1135.[doi:10.11992/tis.202410017]
点击复制
ZHAO Zhenbo,FU Tianyi,DONG Hongbin,et al.Decoupled feature mining rotational detector based on proposal enhancement[J].CAAI Transactions on Intelligent Systems,2025,20(5):1123-1135.[doi:10.11992/tis.202410017]
基于提案增强的解耦特征挖掘旋转检测器
《智能系统学报》[ISSN 1673-4785/CN 23-1538/TP] 卷:
20
期数:
2025年第5期
页码:
1123-1135
栏目:
学术论文—机器感知与模式识别
出版日期:
2025-09-05
- Title:
- Decoupled feature mining rotational detector based on proposal enhancement
- Keywords:
- remote sensing image; object detection; rotated object detection; small object; high density; oriented bounding box; cross-scale fusion; two-stage detector
- 分类号:
- TP391
- 摘要:
- 小而杂乱的物体交织在一起，在遥感图像中尤为常见，给目标检测带来了巨大挑战。在旋转目标检测任务中这个困难更加突出。鉴于此，本文提出了基于提案增强的解耦特征挖掘旋转检测器(decoupled feature mining rotational detector based on proposal enhancement, PDMDet)。首先，采用单阶段检测器取代两阶段检测器的提案生成网络，通过生成高质量提案以减少背景冗余。其次，在相同维度使用自注意力，不同维度使用交叉注意力，通过对相同维度特征增强，不同维度特征交错融合提升检测器对不同尺寸目标的识别能力。最后，鉴于分类和定向边界框回归任务对特征的敏感性不同，本文提出解耦特征细化处理两个不同任务。通过实验，PDMDet在DOTA-v1.0、DOTA-v1.5和HRSC2016这3个数据集上分别取得单尺度78.37%、72.35%、98.60%的平均精度均值，检测准确率高于其他算法，在复杂的旋转目标检测场景具有一定的竞争力。
- Abstract:
- In remote sensing images, small and cluttered objects often appear intertwined, presenting considerable challenges for object detection. These challenges are even further amplified in rotational object detection tasks. Aiming to address these challenges, this paper proposes a decoupled feature mining rotational detector based on proposal enhancement (PDMDet). First, a single-stage detector is employed to replace the region proposal network of the two-stage detector, generating high-quality proposals to reduce background redundancy. Second, self-attention is applied within the same feature dimensions and cross-attention across different dimensions, aiming to enhance intradimensional features and fuse interdimensional features, thereby improving the capability of the detector to identify objects of varying sizes. Finally, recognizing that classification and oriented bounding box regression tasks have different feature sensitivities, this paper proposes a decoupled feature refinement strategy that processes the two tasks separately. Experiment results demonstrate that PDMDet achieves single-scale mAP scores of 78.37%, 72.35%, and 98.60% on DOTA-v1.0, DOTA-v1.5, and HRSC2016 datasets, respectively, outperforming existing algorithms in terms of detection accuracy and demonstrating strong competitiveness in complex rotational object detection scenarios.
- 参考文献/References:
-
[1] 蓝鑫, 吴淞, 伏博毅, 等. 深度学习的遥感图像旋转目标检测综述[J]. 计算机科学与探索, 2024, 18(4): 861-877.
LAN Xin, WU Song, FU Boyi, et al. Survey on deep learning in oriented object detection in remote sensing images[J]. Journal of frontiers of computer science and technology, 2024, 18(4): 861-877.
[2] REN Shaoqing, HE Kaiming, GIRSHICK R, et al. Faster R-CNN: towards real-time object detection with region proposal networks[J]. IEEE transactions on pattern analysis and machine intelligence, 2017, 39(6): 1137-1149.
[3] REDMON J, DIVVALA S, GIRSHICK R, et al. You only look once: unified, real-time object detection[C]//2016 IEEE Conference on Computer Vision and Pattern Recognition. Las Vegas: IEEE, 2016: 779-788.
[4] LIU Wei, ANGUELOV D, ERHAN D, et al. SSD: single shot MultiBox detector[C]//European Conference on Computer Vision. Cham: Springer International Publishing, 2016: 21–37.
[5] 徐红明, 王兴华, 方诚, 等. 基于旋转不变性的高分辨率遥感影像船舶检测[J]. 中国航海, 2024, 47(2): 120-127.
XU Hongming, WANG Xinghua, FANG Cheng, et al. Rotation invariant object detection assisted ship identification from high-resolution remote sensing imagery[J]. Navigation of China, 2024, 47(2): 120-127.
[6] 牛为华, 郭迅. 基于改进YOLOv8的船舰遥感图像旋转目标检测算法[J]. 图学学报, 2024, 45(4): 726-735.
NIU Weihua, GUO Xun. Rotating target detection algorithm in ship remote sensing images based on YOLOv8[J]. Journal of graphics, 2024, 45(4): 726-735.
[7] TIAN Zhi, SHEN Chunhua, CHEN Hao, et al. FCOS: fully convolutional one-stage object detection[C]//2019 IEEE/CVF International Conference on Computer Vision. Seoul: IEEE, 2019: 9626-9635.
[8] LIN T Y, GOYAL P, GIRSHICK R, et al. Focal loss for dense object detection[J]. IEEE transactions on pattern analysis and machine intelligence, 2020, 42(2): 318-327.
[9] YANG Ze, LIU Shaohui, HU Han, et al. RepPoints: point set representation for object detection[C]//2019 IEEE/CVF International Conference on Computer Vision. Seoul: IEEE, 2019: 9656–9665.
[10] DING Jian, XUE Nan, LONG Yang, et al. Learning RoI Transformer for oriented object detection in aerial images[C]//2019 IEEE/CVF Conference on Computer Vision and Pattern Recognition. Long Beach: IEEE, 2019: 2844-2853.
[11] YANG Xue, YANG Jirui, YAN Junchi, et al. SCRDet: towards more robust detection for small, cluttered and rotated objects[C]//2019 IEEE/CVF International Conference on Computer Vision. Seoul: IEEE, 2019: 8231-8240.
[12] HAN Jiaming, DING Jian, LI Jie, et al. Align deep features for oriented object detection[J]. IEEE transactions on geoscience and remote sensing, 2021, 60: 5602511.
[13] YANG Xue, YAN Junchi, FENG Ziming, et al. R3det: refined single-stage detector with feature refinement for rotating object[C]//Proceedings of the AAAI Conference on Artificial Intelligence. [S. l. ]: AAAI Press, 2021: 3163–3171.
[14] XIE Xingxing, CHENG Gong, WANG Jiabao, et al. Oriented R-CNN for object detection[C]//2021 IEEE/CVF International Conference on Computer Vision. Montreal: IEEE, 2021: 3500-3509.
[15] DAI Linhui, LIU Hong, TANG Hao, et al. AO2-DETR: arbitrary-oriented object detection Transformer[J]. IEEE transactions on circuits and systems for video technology, 2023, 33(5): 2342-2356.
[16] 杨学, 严骏驰. 基于特征对齐和高斯表征的视觉有向目标检测[J]. 中国科学: 信息科学, 2023, 53(11): 2250-2265.
YANG Xue, YAN Junchi. Visual oriented object detection via feature alignment and Gaussian parameterization[J]. Scientia sinica (informationis), 2023, 53(11): 2250-2265.
[17] 涂可龙, 卿雅娴, 李真强, 等. EllipticNet: 基于椭圆方程的遥感有向目标检测[J]. 遥感学报, 2025, 29(3): 713-727.
TU Kelong, QING Yaxian, LI Zhenqiang, et al. EllipticNet: equation-based remote sensing oriented object detection network[J]. National remote sensing bulletin, 2025, 29(3): 713-727.
[18] 毛泽勇, 陈欣易, 丁俊峰, 等. 基于不确定性感知旋转目标检测的二次接线质检[J]. 计算机技术与发展, 2024, 34(10): 178-185.
MAO Zeyong, CHEN Xinyi, DING Junfeng, et al. Uncertainty-aware oriented object detection for trustworthy quality inspection of secondary wiring[J]. Computer technology and development, 2024, 34(10): 178-185.
[19] HAN Jiaming, DING Jian, XUE Nan, et al. ReDet: a rotation-equivariant detector for aerial object detection[C]//2021 IEEE/CVF Conference on Computer Vision and Pattern Recognition. [S. l. ]: IEEE, 2021: 2786-2795.
[20] FENG Xiaoxu, YAO Xiwen, CHENG Gong, et al. Weakly supervised rotation-invariant aerial object detection network[C]//2022 IEEE/CVF Conference on Computer Vision and Pattern Recognition. New Orleans: IEEE, 2022: 14126-14135.
[21] 方婷婷, 刘斌, 陈春晖, 等. 视图一致性网络下的弱监督遥感影像旋转目标检测[J]. 遥感学报, 2024, 28(12): 3213-3230.
FANG Tingting, LIU Bin, CHEN Chunhui, et al. View-consistency network for weakly supervised oriented object detection in remote sensing images[J]. National remote sensing bulletin, 2024, 28(12): 3213-3230.
[22] ZHOU Yanzhao, YE Qixiang, QIU Qiang, et al. Oriented response networks[C]//2017 IEEE Conference on Computer Vision and Pattern Recognition. Honolulu: IEEE, 2017: 4961-4970.
[23] ALEXEY D, LUCAS B, ALEXANDER K, et al. An image is worth 16x16 words Transformers for image recognition at scale[C]//International Conference on Learning Representations. [S. l. ]: OpenReview.net, 2021: 1-21.
[24] LIU Ze, LIN Yutong, CAO Yue, et al. Swin Transformer: hierarchical vision Transformer using shifted windows[C]//2021 IEEE/CVF International Conference on Computer Vision. Montreal: IEEE, 2021: 9992-10002.
[25] WU Yuhuan, LIU Yun, ZHAN Xin, et al. P2T: pyramid pooling Transformer for scene understanding[J]. IEEE transactions on pattern analysis and machine intelligence, 2023, 45(11): 12760-12771.
[26] ZHOU Xingyi, KOLTUN V, KR?HENBüHL P. Probabilistic two-stage detection[EB/OL]. (2021-3-12)[2024-11-21]. https://arxiv.org/abs/2103.07461v1.
[27] KIM K, LEE H S. Probabilistic anchor assignment with IoU prediction for object detection[C]//European Conference on Computer Vision. Cham: Springer International Publishing, 2020: 355-371.
[28] 付天怡, 杨本翼, 董红斌, 等. 基于深度重塑的航拍目标检测增强网络[J]. 模式识别与人工智能, 2024, 37(7): 652-662.
FU Tianyi, YANG Benyi, DONG Hongbin, et al. Depth-reshaping based aerial object detection enhanced network[J]. Pattern recognition and artificial intelligence, 2024, 37(7): 652-662.
[29] ASHISH V, NOAM S, NIKI P, et al. Attention is all you need[C]//Advances in Neural Information Processing Systems. Long Beach: Springer, 2017: 1-11.
[30] ZHOU Yue, YANG Xue, ZHANG Gefan, et al. MMRotate: a rotated object detection benchmark using PyTorch[C]//Proceedings of the 30th ACM International Conference on Multimedia. Lisboa: ACM, 2022: 7331–7334.
[31] XIA Guisong, BAI Xiang, DING Jian, et al. DOTA: a large-scale dataset for object detection in aerial images[C]//2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition. Salt Lake City: IEEE, 2018: 3974-3983.
[32] LIU Zikun, YUAN Liu, WENG Lubin, et al. A high resolution optical satellite image dataset for ship recognition and some new baselines[C]//Proceedings of the 6th International Conference on Pattern Recognition Applications and Methods. Porto: SCITEPRESS-Science and Technology Publications, 2017: 324-331.
[33] LI Yuxuan, HOU Qibin, ZHENG Zhaohui, et al. Large selective kernel network for remote sensing object detection[C]//2023 IEEE/CVF International Conference on Computer Vision. Paris: IEEE, 2023: 16748-16759.
[34] YANG Xue, YAN Junchi, MING Qi, et al. Rethinking rotated object detection with Gaussian Wasserstein distance loss[C]//International Conference on Machine Learning. [S. l. ]: PMLR, 2021: 11830–11841.
[35] YANG Xue, YANG Xiaojiang, YANG Jirui, et al. Learning high-precision bounding box for rotated object detection via kullback-leibler divergence[C]//Advances in Neural Information Processing Systems. [S. l. ]: Curran Associates Inc., 2021: 18381–18394.
[36] PU Yifan, WANG Yiru, XIA Zhuofan, et al. Adaptive rotated convolution for rotated object detection[C]//2023 IEEE/CVF International Conference on Computer Vision. Paris: IEEE, 2023: 6566-6577.
[37] WANG Jiangshan, PU Yifan, HAN Yizeng, et al. GRA: detecting oriented objects through group-wise rotating and attention[C]//European Conference on Computer Vision. Cham: Springer Nature Switzerland, 2024: 298-315.
[38] CAI Xinhao, LAI Qiuxia, WANG Yuwei, et al. Poly kernel inception network for remote sensing detection[C]//2024 IEEE/CVF Conference on Computer Vision and Pattern Recognition. Seattle: IEEE, 2024: 27706-27716.
[39] HE Kaiming, GKIOXARI G, DOLLáR P, et al. Mask R-CNN[C]//2017 IEEE International Conference on Computer Vision. Venice: IEEE, 2017: 2980-2988.
[40] CHEN Kai, PANG Jiangmiao, WANG Jiaqi, et al. Hybrid task cascade for instance segmentation[C]//2019 IEEE/CVF Conference on Computer Vision and Pattern Recognition. Long Beach: IEEE, 2019: 4969-4978.
[41] LI Wentong, CHEN Yijie, HU Kaixuan, et al. Oriented RepPoints for aerial object detection[C]//2022 IEEE/CVF Conference on Computer Vision and Pattern Recognition. New Orleans: IEEE, 2022: 1819-1828.
[42] EVERINGHAM M, VAN G, WILLIAMS C K I, et al. The PASCAL visual object classes challenge 2007 (VOC2007) results[EB/OL]. (2007-06-07)[2024-12-20]. http://www.pascalnetwork.org/challenges/VOC/voc2007/workshop/index.html.
[43] EVERINGHAM M, VAN G, WILLIAMS C K I, et al. The PASCAL visual object classes challenge 2012 (VOC2012) results[EB/OL]. (2012-05-18)[2024-12-20]. http://www.pascalnetwork.org/challenges/VOC/voc2012/workshop/index.html.
[44] KOLEN J F, KREMER S C. Gradient flowin recurrent nets: the difficulty of learning longterm dependencies[J]. A field guide to dynamical recurrent networks, 2001, 237-243.
- 相似文献/References:
-
[1]刘富,于鹏,刘坤.采用独立分量分析Zernike矩的遥感图像飞机目标识别[J].智能系统学报,2011,6(1):51.
LIU Fu,YU Peng,LIU Kun.Research concerning aircraft recognition of remote sensing images based on ICA Zernike invariant moments[J].CAAI Transactions on Intelligent Systems,2011,6():51.
[2]胡光龙,秦世引.动态成像条件下基于SURF和Mean shift的运动目标高精度检测[J].智能系统学报,2012,7(1):61.
HU Guanglong,QIN Shiyin.High precision detection of a mobile object under dynamic imaging based on SURF and Mean shift[J].CAAI Transactions on Intelligent Systems,2012,7():61.
[3]韩峥,刘华平,黄文炳,等.基于Kinect的机械臂目标抓取[J].智能系统学报,2013,8(2):149.[doi:10.3969/j.issn.1673-4785.201212038]
HAN Zheng,LIU Huaping,HUANG Wenbing,et al.Kinect-based object grasping by manipulator[J].CAAI Transactions on Intelligent Systems,2013,8():149.[doi:10.3969/j.issn.1673-4785.201212038]
[4]韩延彬,郭晓鹏,魏延文,等.RGB和HSI颜色空间的一种改进的阴影消除算法[J].智能系统学报,2015,10(5):769.[doi:10.11992/tis.201410010]
HAN Yanbin,GUO Xiaopeng,WEI Yanwen,et al.An improved shadow removal algorithm based on RGB and HSI color spaces[J].CAAI Transactions on Intelligent Systems,2015,10():769.[doi:10.11992/tis.201410010]
[5]龙海侠,吴淑雷,吕雁.基于多样性变异的QPSO算法的遥感图像分类[J].智能系统学报,2015,10(6):938.[doi:10.11992/tis.201507045]
LONG Haixia,WU Shulei,LYU Yan.Classification of multispectral remote sensing image based on QPSO and diversity-mutation[J].CAAI Transactions on Intelligent Systems,2015,10():938.[doi:10.11992/tis.201507045]
[6]曾宪华,易荣辉,何姗姗.流形排序的交互式图像分割[J].智能系统学报,2016,11(1):117.[doi:10.11992/tis.201505037]
ZENG Xianhua,YI Ronghui,HE Shanshan.Interactive image segmentation based on manifold ranking[J].CAAI Transactions on Intelligent Systems,2016,11():117.[doi:10.11992/tis.201505037]
[7]吴诗婳,吴一全,周建江.直线截距直方图城区遥感图像多阈值分割[J].智能系统学报,2018,13(2):227.[doi:10.11992/tis.201609012]
WU Shihua,WU Yiquan,ZHOU Jianjiang.Multi-level thresholding for remote sensing image of urban area based on line intercept histogram[J].CAAI Transactions on Intelligent Systems,2018,13():227.[doi:10.11992/tis.201609012]
[8]葛园园,许有疆,赵帅,等.自动驾驶场景下小且密集的交通标志检测[J].智能系统学报,2018,13(3):366.[doi:10.11992/tis.201706040]
GE Yuanyuan,XU Youjiang,ZHAO Shuai,et al.Detection of small and dense traffic signs in self-driving scenarios[J].CAAI Transactions on Intelligent Systems,2018,13():366.[doi:10.11992/tis.201706040]
[9]李亚飞,董红斌.基于卷积神经网络的遥感图像分类研究[J].智能系统学报,2018,13(4):550.[doi:10.11992/tis.201706078]
LI Yafei,DONG Hongbin.Classification of remote-sensing image based on convolutional neural network[J].CAAI Transactions on Intelligent Systems,2018,13():550.[doi:10.11992/tis.201706078]
[10]莫宏伟,汪海波.基于Faster R-CNN的人体行为检测研究[J].智能系统学报,2018,13(6):967.[doi:10.11992/tis.201801025]
MO Hongwei,WANG Haibo.Research on human behavior detection based on Faster R-CNN[J].CAAI Transactions on Intelligent Systems,2018,13():967.[doi:10.11992/tis.201801025]
[11]赵文清,康怿瑾,赵振兵,等.改进YOLOv5s的遥感图像目标检测[J].智能系统学报,2023,18(1):86.[doi:10.11992/tis.202203013]
ZHAO Wenqing,KANG Yijin,ZHAO Zhenbing,et al.A remote sensing image object detection algorithm with improved YOLOv5s[J].CAAI Transactions on Intelligent Systems,2023,18():86.[doi:10.11992/tis.202203013]
[12]梁礼明,冯耀,龙鹏威,等.基于MobileViT和多尺度特征聚合的遥感图像目标检测[J].智能系统学报,2024,19(5):1168.[doi:10.11992/tis.202310022]
LIANG Liming,FENG Yao,LONG Pengwei,et al.Remote sensing image object detection based on MobileViT and multiscale feature aggregation[J].CAAI Transactions on Intelligent Systems,2024,19():1168.[doi:10.11992/tis.202310022]
[13]赵文清,赵振寰,巩佳潇.结合倒残差自注意力机制的遥感图像目标检测[J].智能系统学报,2025,20(1):64.[doi:10.11992/tis.202312001]
ZHAO Wenqing,ZHAO Zhenhuan,GONG Jiaxiao.Remote sensing image object detection based on inverted residual self-attention mechanism[J].CAAI Transactions on Intelligent Systems,2025,20():64.[doi:10.11992/tis.202312001]
备注/Memo
收稿日期:2024-10-24。
基金项目:国家自然科学基金项目(82374621).
作者简介:赵振博，硕士研究生，主要研究方向为深度学习、旋转目标检测。E-mail：zhao_zhenbo@hrbeu.edu.cn。;付天怡，博士研究生，主要研究方向为深度学习、计算机视觉。E-mail：futianyi@hrbeu.edu.cn。;董红斌，教授，博士生导师，中国计算机学会高级会员，主要研究方向为多智能体系统、机器学习。主持和完成国家自然科学基金项目、工信部基础研究项目、黑龙江省自然科学基金项目，荣获黑龙江省高校科学技术奖和黑龙江省优秀高等教育科学成果奖。发表学术论文 90 余篇，主编教材 2 部。E-mail：donghongbin@hrbeu.edu.cn。
通讯作者:董红斌. E-mail：donghongbin@hrbeu.edu.cn
更新日期/Last Update:
2025-09-05