基于[KG-*9]Ghost卷积与注意力机制的SAR图像建筑物检测算法

doi:10.12382/bgxb.2021.0425

摘要/Abstract

摘要： 针对深度卷积神经网络存在模型参数量大、占用内存资源等问题，提出了一种基于轻量化网络的SAR图像建筑物检测算法。首先以旋转目标检测算法R-centernet为基础，将主干网络中的传统卷积替换为Ghost卷积，并构建Ghost-ResNet网络，降低模型参数量；其次提出了融合宽高信息的通道注意力模块，增强网络对图像中显著区域的精确定位能力；使用CARAFE上采样代替网络中的DCN模块，在上采样过程中充分结合特征图信息，提高目标检测能力；最后使用改进的R-centernet算法在旋转标注的SAR图像建筑物数据集上进行训练与测试。实验结果表明，相比于原始R-centernet算法，改进后的算法准确率提高了3.8%，召回率提高了1.2%，检测速度提高了12帧/s。

关键词: 轻量化网络, SAR图像建筑物检测, 旋转目标检测, Ghost卷积, 通道注意力, CARAFE上采样

Abstract: A building detection algorithm in SAR images based on a lightweight network is proposed to address the problems caused by the large quantity of model parameters and memory resources involved in deep convolutional neural networks. First, the algorithm is based on the rotating detector R-centernet, and the traditional convolution in backbone network is replaced by Ghost convolution. Then, the Ghost-ResNet is constructed to reduce the number of model parameters. Second, a channel attention module that fuses width and height information is developed to enhance the network's ability to locate significant regions in the images accurately. An up-sampling method named CARAFE is used to replace the DCN module in the network, and the feature map information is fully combined in the up-sampling process to improve target detection. Finally, the improved R-centernet is used to train and test the SAR rotating building dataset. Based on the experimental results, compared with the R-centernet, the improved algorithm has increased detection accuracy by 3.8%, recall by 1.2%, and detection speed by 12 frames per second.

Key words: lightweightnetwork, buildingdetectioninSARimage, rotatingtargetdetection, Ghostconvolution, channelattention, CARAFEup-sampling

中图分类号:

TP751.1

严继伟，苏娟，李义红. 基于[KG-*9]Ghost卷积与注意力机制的SAR图像建筑物检测算法[J]. 兵工学报, 2022, 43(7): 1667-1675.

YAN Jiwei, SU Juan, LI Yihong. Building Detection Algorithm in SAR Images Based on Ghost Convolution and Attention Mechanisms[J]. Acta Armamentarii, 2022, 43(7): 1667-1675.

参考文献

［1］ HU L, ZHANG J, GAO F. A building extraction method using shadow in high resolution multispectral images［C］∥Proceedings of International Geoscience and Remote Sensing Symposium. Vancouver, BC, Canada: IEEE, 2011: 24-29.
［2］ GIRSHICK R, DONAHUE J, DARRELL T, et al. Rich feature hierarchies for accurate object detection and semantic segmentation［C］∥Proceedings of IEEE Conference on Computer Vision and Pattern Recognition. Columbus, OH, US：IEEE, 2014: 580-587.
［3］ GIRSHICK R. Fast R-CNN［C］∥Proceedings of IEEE International Conference on Computer Vision. Santiago, Chile: IEEE, 2015: 1440-1448.
［4］ REN S, HE K, GIRSHICK R, et al. Faster R-CNN: towards real-time object detection with region proposal networks［C］∥Proceedings of the 2015 neural information processing systems(NIPS). New York，NY，US: Curran Associates Inc., 2015: 91- 99.
［5］ REDMON J, DIVVALA S, GIRSHICK R, et al. You only look once: unified, real-time object detection［C］∥Proceedings of IEEE Conference on Computer Vision and Pattern Recognition. Las Vegas, NV, US：IEEE, 2016: 779-788.
［6］ REDMON J, FARHADI A. YOLO9000: better, faster, stronger［C］∥Proceedingts of IEEE Conference on Computer Vision and Pattern Recognition. Honolulu, HI, US: IEEE, 2017: 6517-6525.
［7］ REDMON J, FARHADI A. YOLOv3: an incremental improvement［EB/OL］.［2021-04-05］.http:∥arxiv.org/abs/1804.02767.
［8］ BOCHKOVSKIY A, WANG C Y, LIAO H. YOLOv4: optimal speed and accuracy of object detection［EB/OL］. ［2021-04-05］.http:∥arxiv.org/abs/2004.10934.
［9］ LIU W, ANGUELOV D, ERHAN D, et al. SSD: single shot multibox detector［C］∥Proceedings of European Conference on Computer Vision. Berlin，Germany: Springer, 2016: 21-37.
［10］ ZHOU X Y, WANG D Q, KRHENB?HL P. Objects as points［EB/OL］.［2019-04-26］.http:∥arxiv.org/abs/1904.07850.
［11］ SHAHZAD M, MAURER M, FRAUNDORFER F, et al. Buildings detection in VHR SAR images using fully convolution neural networks ［J］. IEEE Transactions on Geoscience and Remote Sensing, 2019, 57(2): 1100-1116.
［12］ ZHU X X, BAMLER R. Very high resolution space borne SAR tomography in urban environment ［J］. IEEE Transactions on Geoscience and Remote Sensing, 2010, 48(12): 4296-4308.
［13］ LI J, ZHANG R, LI Y. Multiscale convolutional neural network for the detection of built-up areas in high-resolution SAR images［C］∥Proceedings of 2016 IEEE International Geoscience and Remote Sensing Symposium (IGARSS). Piscataway，NJ，US: IEEE, 2016: 910-913.
［14］ WU Y, ZHANG R, LI Y. The detection of built-up areas in high-resolution SAR images based on deep neural networks［C］∥Proceedings of International Conference on Image and Graphics. Berlin，Germany：Springer, 2017: 646-655.
［15］ AN Q, PAN Z, LIU L, et al. DRBox-v2: an improved detector with rotatable boxes for target detection in SAR images ［J］. IEEE Transactions on Geoscience and Remote Sensing, 2019, 57(99): 8333-8349.
［16］ YANG X, YANG J, YAN J, et al. SCRDet: Towards More Robust Detection for Small, Cluttered and Rotated Objects［C］∥Proceedings of 2019 IEEE/CVF International Conference on Computer Vision (ICCV). Piscataway，NJ，US: IEEE, 2019: 8231-8240.
［17］ YANG X, LIU Q, YAN J, et al. R3Det: refined single-stage detector with feature refinement for rotating object［EB/OL］.［2021-04-05］.http:∥arxiv.org/abs/1908.05612.
［18］ ZHOU X Y, YAO C, WEN H, et al. EAST: an efficient and accurate scene text detector［C］∥Proceedings of IEEE Conference on Computer Vision and Pattern Recognition. Honolulu, HI，US:IEEE, 2017: 2642-2651.
［19］ LIU X B, LIANG D, YAN S, et al. FOTS: fast oriented text spotting with a unified network［C］∥Proceedings of IEEE Conference on Computer Vision and Pattern Recognition. Salt Lake City, UT，US：IEEE, 2018: 5676-5685.
［20］ HOWARD A G, ZHU M L, CHEN B, et al. MobileNets: efficient convolutional neural networks for mobile vision applications［EB/OL］.［2017-05-27］.http:∥arxiv.org/abs/1704.04861.
［21］ SANDLER M, HOWARD A G, ZHU M L, et al. Inverted residuals and linear bottlenecks: mobile networks for classification, detection and segmentation［EB/OL］. ［2018-01-12］.http:∥arxiv.org/abs/1801.04381.
［22］ HOWARD A, SANDLER M, CHU G, et al. Searching for MobileNetV3［C］∥Proceedings of IEEE International Conference on Computer Vision. Seoul, South Korea：IEEE, 2019: 1314-1324.
［23］ HAN K, WANG Y H, TIAN Q, et al. GhostNet: more features from cheap operations［C］∥Proceedings of IEEE Conference on Computer Vision and Pattern Recognition. Seattle, WA，US：IEEE, 2020: 1577-1586.
［24］ ZHANG X Y, ZHOU X Y, LIN M X, et al. ShuffleNet: an extremely efficient convolutional neural network for mobile devices［C］∥Proceedings of IEEE Conference on Computer Vision and Pattern Recognition. Salt Lake City, UT，US：IEEE, 2018: 6848-6856.
［25］ LI X, SU J, YANG L. Building detection in SAR images based on bi-dimensional empirical mode decomposition algorithm ［J］. IEEE Geoscience and Remote Sensing Letters, 2020, 17(4): 641-645.
［26］ HU J, SHEN L, ALBANIE S, et al. Squeeze-and- excitation networks ［J］. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2020, 42(8): 2011-2023.
［27］ WANG Q, WU B, ZHU P, et al. ECA-Net: efficient channel attention for deep convolutional neural networks［C］∥Proceedings of IEEE/CVF Conference on Computer Vision and Pattern Recognition. Seattle, WA，US：IEEE, 2020: 11531-11539.
［28］ HOU Q B, ZHOU D Q, FENG J S. Coordinate attention for efficient mobile network design［EB/OL］.［2021-03-15］. https:∥dblp.org/rec/journals/corr/abs-2103-02907.
［29］ WANG J Q, CHEN K, XU R, et al. CARAFE: content-aware reassembly of features［C］∥Proceedings of IEEE International Conference on Computer Vision. Seoul, South Korea：IEEE, 2019: 3007-3016.
［30］李响, 苏娟, 杨龙. 基于改进YOLOV3的合成孔径雷达图像中建筑物检测算法［J］. 兵工学报, 2020, 41(7): 1347-1359.
LI X, SU J, YANG L. A SAR image building detection algorithm based on improved YOLOV3 ［J］. Acta Armamentarii, 2020, 41(7): 1347-1359. (in Chinese)

[1]	马鹏阁, 魏宏光, 孙俊灵, 陶然, 庞栋栋, 单涛, 蔡志勇, 刘兆瑜. 基于LOG滤波的增强局部对比度红外小目标检测方法[J]. 兵工学报, 0, (): 0-0.
[2]	赵继东，盖振伟，李晶，杨志来，王巍. 车辆通信设备集成化体系结构[J]. 兵工学报, 2022, 43(S1): 21-25.
[3]	王珣，高亮，魏永勇，李广运，张天赐. 基于任务级-交战级体系仿真平台的并行实验研究[J]. 兵工学报, 2022, 43(S1): 183-188.
[4]	李娟，郭雷平，王宁，焦雯，胡冰甦. 装备体系建设现状分析[J]. 兵工学报, 2022, 43(S1): 222-226.
[5]	李静, 刘哲, 黄文准. 基于快速非局部均值和超分辨率重建的图像降噪算法[J]. 兵工学报, 2021, 42(8): 1716-1727.
[6]	李响，苏娟，杨龙. 基于改进YOLOv3的合成孔径雷达图像中建筑物检测算法[J]. 兵工学报, 2020, 41(7): 1347-1359.
[7]	杨龙，苏娟，李响. 基于生成式对抗网络的合成孔径雷达舰船数据增广在改进单次多盒检测器中的应用[J]. 兵工学报, 2019, 40(12): 2488-2496.
[8]	谌德荣，宫久路，陈乾，曹旭平. 基于样本分割的快速高光谱图像异常检测支持向量数据描述方法[J]. 兵工学报, 2008, 29(9): 1049-1053.
[9]	李怀琼，陈钱，隋修宝. 基于边缘保护的红外图像插值放大算法[J]. 兵工学报, 2006, 27(4): 655-658.
[10]	1:罗强，杨万海；2：罗莉，何鸿君；3：任庆利. 一种基于小波变换的卫星SAR海洋图像舰船目标检测方法[J]. 兵工学报, 2002, 23(4): 500-503.
[11]	1:程相权，郭治，余海,戚国庆;2:李显峰,殷开宝. ー种快速目标的图像跟踪综合算法[J]. 兵工学报, 2001, 22(4): 496-499.