|
|
论文:2024,Vol:42,Issue(2):335-343 |
|
|
引用本文： |
|
|
王啸林, 方厚章, 李雪婷, 吴辰星, 王黎明. 逐像素注意力驱动的红外小目标检测网络[J]. 西北工业大学学报 |
|
|
WANG Xiaolin, FANG Houzhang, LI Xueting, WU Chenxing, WANG Liming. Pixel-wise attention driven infrared small target detection network[J]. Journal of Northwestern Polytechnical University |
|
|
|
|
|
|
|
| 逐像素注意力驱动的红外小目标检测网络 |
|
| 王啸林, 方厚章, 李雪婷, 吴辰星, 王黎明 |
|
| 西安电子科技大学 计算机科学与技术学院, 陕西 西安 710071 |
| 摘要: |
| 红外小目标检测在军事和民用领域获得了广泛应用,但其存在目标尺度小、细节少、复杂背景干扰等问题,现有经典深度学习检测方法往往适用于通用目标检测,对红外小目标适配性较差。针对上述问题,构建了一种新的基于U形注意力块和逐像素注意力块的红外小目标检测网络。设计了U形注意力块,在单层级内通过局部U形子网络提取多尺度特征,并通过逐像素注意力精细化增强小目标特征,丰富多尺度小尺度目标特征表示,提升网络对小尺度目标判别能力;通过稠密融合方式进一步保留小目标信息,缓解不同层特征融合时的语义鸿沟,降低漏检率;将空间与通道2个维度逐像素注意力块应用于融合后的特征图,避免小目标特征被衰减,同时抑制复杂背景干扰。实验结果表明,提出的网络在2个红外小目标数据集NUDT-SIRST与IRSTD-1k上交并比、检测概率、虚警率指标均超过最新基准方法。此外,所提网络在检测精度和效率上也达到较好平衡。 |
| 关键词:
红外小目标检测
U形注意力块
逐像素注意力
|
|
| Pixel-wise attention driven infrared small target detection network |
|
| WANG Xiaolin, FANG Houzhang, LI Xueting, WU Chenxing, WANG Liming |
|
| School of Computer Science and Technology, Xidian University, Xi'an 710071, China |
| Abstract: |
| Infrared small target detection has been widely used in the military and civil fields. The detection difficulties lie in small target size, few details and complex background interference. Existing classic deep learning detection methods are commonly suitable for generic target detection, but have poor adaptability to infrared small targets. To handle these problems, this article build a new infrared small target detection network based on U-shaped attention block and pixel-wise attention block. Firstly, a residual U-shaped attention block to extract multi-scale features through local U-shaped subnetworks in a single layer level and extract multiscale features to enrich the representation of small-scale target features is designed, so as to enhance the network's ability to discriminate small-scale targets. Then, small target information is further preserved through dense fusion method to alleviate the semantic gap feature fusion between different layers and reduce the miss detection rate. Finally, the pixel-wise attention in space and channel dimensions is applied to the fused feature map to enhance small targets and suppress complex background interference. The experimental results show that the present network outperforms the latest benchmark method in the intersection over union, detection probability and false alarm rate of two infrared small target data sets NUDT-SIRST and IRSTD-1k. Moreover, the present network also achieves a good balance between the detection accuracy and the efficiency. |
| Key words:
small infrared target detection
U-shaped attention
pixel-wise attention
|
|
| 收稿日期: 2023-03-07
修回日期:
|
| DOI: 10.1051/jnwpu/20244220335 |
| 基金项目: 国家自然科学基金(41501371)、多谱信息处理技术国家级重点实验室开放课题(6142113220303)、中央高校基本科研业务费专项资金与西安电子科技大学研究生创新基金(YJSJ24015)资助 |
| 通讯作者: 王黎明(1982—),副教授 e-mail:wanglm@mail.xidian.edu.cn
Email：wanglm@mail.xidian.edu.cn |
| 作者简介: 王啸林(1998—),博士研究生
|
|
|
|
|
|
|
|
参考文献: |
|
|
[1] 张凯, 刘昊, 杨曦, 等. 基于关键点检测网络的空中红外目标要害部位识别算法[J]. 西北工业大学学报, 2020,38(6): 1154-1162 ZHANG Kai, LIU Hao, YANG Xi, et al. Identification algorithm based on key-point detection network for vital parts of infrared aerial target[J]. Journal of Northwestern Polytechnical University, 2020, 38(6): 1154-1162 (in Chinese)
[2] ZHA Y, WU M, QIU Z, et al. Online semantic subspace learning with siamese network for UAV tracking[J]. Remote Sensing, 2020, 12(2): 325
[3] DAI Y, WU Y. Reweighted infrared patch-tensor model with both nonlocal and local priors for single-frame small target detection[J]. IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing, 2017, 10(8): 3752-3767
[4] ZHANG L, PENG Z. Infrared small target detection based on partial sum of the tensor nuclear norm[J]. Remote Sensing, 2019, 11(4): 382
[5] 胡磊力, 张君昌, 张亮中. 基于人类视觉系统的实时红外目标检测方法[J]. 西北工业大学学报, 2017,35(5): 910-914 HU Leili, ZHANG Junchang, ZHANG Liangzhong. Real-time infrared target detection method based on human vision system[J]. Journal of Northwestern Polytechnical University, 2017, 35(5): 910-914 (in Chinese)
[6] 龚任杰, 郑智辉, 丛龙剑, 等. 小样本条件下异源图像迁移学习的红外目标检测与识别[J]. 西北工业大学学报, 2021, 39(增刊1): 84-88 GONG Renjie, ZHENG Zhihui, CONG Longjian, et al. Infrared target detection and recognition based on transfer learning with small samples[J]. Journal of Northwestern Polytechnical University, 2021,39(suppl 1): 84-88 (in Chinese)
[7] MCINTOSH B, VENKATARAMANAN S, MAHALANOBIS A. Infrared target detection in cluttered environments by maximization of a target to clutter ratio(TCR) metric using a convolutional neural network[J]. IEEE Trans on Aerospace and Electronic Systems, 2020, 57(1): 485-496
[8] RONNEBERGER O, FISCHER P, BROX T. U-net: convolutional networks for biomedical image segmentation[C]//Medical Image Computing and Computer-Assisted Intervention, 2015
[9] LI B, XIAO C, WANG L, et al. Dense nested attention network for infrared small target detection[J]. IEEE Trans on Image Processing, 2022, 32: 1745-1758
[10] ZHANG M, ZHANG R, YANG Y, et al. ISNET: shape matters for infrared small target detection[C]//Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2022: 877-886
[11] QIN X, ZHANG Z, HUANG C, et al. U2-Net: going deeper with nested U-structure for salient object detection[J]. Pattern recognition, 2020, 106: 107404
[12] WOO S, PARK J, LEE J Y, et al. CBAM: convolutional block attention module[C]//Proceedings of the European Conference on Computer Vision, 2018: 3-19
[13] WANG H, ZHOU L, WANG L. Miss detection vs. false alarm: adversarial learning for small object segmentation in infrared images[C]//Proceedings of the IEEE/CVF International Conference on Computer Vision, 2019: 8509-8518
[14] DAI Y, WU Y, ZHOU F, et al. Asymmetric contextual modulation for infrared small target detection[C]//Proceedings of the IEEE/CVF Winter Conference on Applications of Computer Vision, 2021: 950-959
[15] DAI Y, WU Y, ZHOU F, et al. Attentional local contrast networks for infrared small target detection[J]. IEEE Trans on Geoscience and Remote Sensing, 2021, 59(11): 9813-9824
[16] JIANG N, KUIRAN W, XIAOKE P, et al. Anti-UAV: a large-scale benchmark for vision-based uav tracking[J]. IEEE Trans on Multimedia, 2023, 25: 486-500 |
|
|
|
|
|
|
|
