计算机科学 ›› 2019, Vol. 46 ›› Issue (3): 159-163.doi: 10.11896/j.issn.1002-137X.2019.03.024
赵振兵1,崔雅萍1,戚银城1,杜丽群1,张珂1,翟永杰2
ZHAO Zhen-bing1,CUI Ya-ping1,QI Yin-cheng1,DU Li-qun1,ZHANG Ke1,ZHAI Yong-jie2
摘要： 航拍巡线图像中的绝缘子目标存在部分遮挡的情况,利用区域全卷积网络(Region-based Fully Convolutional Networks,R-FCN)模型对其进行检测,出现了绝缘子目标检测效果较差且检测框无法完全贴合目标的问题。基于此,文中提出了一种基于改进的R-FCN航拍巡线图像中的绝缘子目标检测方法。首先,根据绝缘子目标的宽高比特征,将R-FCN模型中RPN的建议框的宽高比修改为1∶4,1∶2,1∶1,2∶1,4∶1;然后,针对遮挡问题,在R-FCN模型中引入对抗空间丢弃网络(Adversarial Spatial Dropout Network,ASDN)层,对特征图的部分位置生成掩码以获得目标特征的不完整样本,从而提高模型对目标特征较差的样本检测性能。在包含7433个绝缘子目标框的数据集中,R-FCN模型的平均检测率达到了77.27%,而改进的R-FCN检测方法的平均检测率达到了84.29%,性能提升了7.02%,且检测框更贴合目标。
中图分类号:
| [1]MA X Y.Research on automatic inspection technology of transmission lines based on aerial image[D].Beijing:North China Electric Power University,2016.(in Chinese) 马兴誉.基于航拍图像的输电线路自动巡检相关技术研究[D].北京:华北电力大学,2016. [2]JIANG S,JIANG W,HUANG W,et al.UAV-Based oblique photogrammetry for outdoor data acquisition and offsite visual inspection of transmission line[J].Remote Sensing,2017,9(3):278-303. [3]MATIKAINEN L,LEHTOMAKI M,AHOKAS E,et al.Remote sensing methods for power line corridor surveys[J].Isprs Journal of Photogrammetry & Remote Sensing,2016,119:10-31. [4]KRIZHEVSKY A,SUTSKEVER I,HINTON G E.ImageNet classification with deep convolutional neural networks[J].Communications of the Acm,2012,60(2):1097-1105. [5]DAI J F,LI Y,HE K M,et al.R-FCN:Object Detection via Region-based Fully Convolutional Networks[C]∥Conference and Workshop on Neural Information Processing Systems(NIPS).2016:379-387. [6]HOSANG J,BENENSON R,DOLLAR P,et al.What Makes for Effective Detection Proposals?[J].IEEE Transactions on Pattern Analysis & Machine Intelligence,2016,38(4):814. [7]GIRSHICK R,DONAHUE J,DARRELL T,et al.Rich Feature Hierarchies for Accurate Object Detection and Semantic Segmentation[C]∥IEEE Conference on Computer Vision and Pattern Recognition(CVPR).2013:580-587. [8]SANDE K,UIJLINGS J,GEVERS T,et al.Segmentation as selective search for object recognition[C]∥IEEE International Conference on Computer Vision (ICCV).2012:1879-1886. [9]HE K M,ZHANG X,REN S Q,et al.Spatial pyramid pooling in deep convolutional networks for visual recognition[J].IEEE Transactions on Pattern Analysis & Machine Intelligence,2014,37(9):1904-1916. [10]GIRSHICK R.Fast R-CNN[C]∥IEEE International Confe- rence on Computer Vision(ICCV).2015:1440-1448. [11]REN S Q,HE K M,GIRSHICK R,et al.Faster R-CNN:to- wards real-Time object detection with region proposal networks[J].IEEE Transactions on Pattern Analysis & Machine Intelligence,2015,39(6):1137-1149. [12]LIU W,ANGUELOV D,ERHAN D,et al.SSD:single shot multiboxdetector[C]∥European Conference on Computer Vision(ECCV).2015:21-37. [13]HUANG X N,ZHANG Z L.A method to extract insulator image from aerial image of helicopter patrol[J].Power System Technology,2010,34(1):194-197.(in Chinese) 黄宵宁,张真良.直升机巡检航拍图像中绝缘子图像的提取算法[J].电网技术,2010,34(1):194-197. [14]XU Y L,ZHANG S C,YANG N,et al.An algorithm to extract insulator image from aerial image[J].Journal of Shanghai University of Electric Power,2011,27(5):515-518.(in Chinese) 徐耀良,张少成,杨宁,等.航拍图像中绝缘子的提取算法[J].上海电力学院学报,2011,27(5):515-518. [15]YAO C Y,JIN L J,YAN S J.Recognition of insulator string in power grid patrol images[J].Journal of System Simulation,2012,24(9):1818-1822.(in Chinese) 姚春羽,金立军,闫书佳.电网巡检图像中绝缘子的识别[J].系统仿真学报,2012,24(9):1818-1822. [16]XU X J,WANG S P,JI Q C,et al.Insulator infrared image recognition method based on gaussian scale-space and GHT[J].Infrared Technology,2014,36(7):596-599.(in Chinese) 徐向军,王生鹏,纪青春,等.基于高斯尺度空间GHT的绝缘子红外图像识别方法[J].红外技术,2014,36(7):596-599. [17]ZHAO Z B,WANG L.Aerial insulator string image automatic location method[J].Chinese Journal of Scientific Instrument,2014,35(3):558-565.(in Chinese) 赵振兵,王乐.一种航拍绝缘子串图像自动定位方法[J].仪器仪表学报,2014,35(3):558-565. [18]ZHAO Z B,XU G Z,QI Y C,et al.Multi-patch deep features for power line insulator status classification from aerial images[C]∥IEEE International Joint Conference on Neural Networks (IJCNN).2016:3187-3194. [19]ZHAO Z B,FAN X Q,XU G Z,et al.Aggregating deep convolutional feature maps for insulator detection in infrared images[J].IEEE Access,2017,5(1):21831-21839. [20]ZHAO Z B,XU G Z,QI Y C.Representation of binary feature pooling for detection of insulator strings in infrared images[J].IEEE Transactions on Dielectrics and Electrical Insulation,2016,23(5):2858-2866. [21]WANG X,SHRIVASTAVA A,GUPTA A.A-Fast-RCNN: hard positive generation via adversary for object detection[C]∥IEEE Conference on Computer Vision and Pattern Recognition(CVPR).2017:3039-3048. |
| [1] | 戴禹, 许林峰. 基于文本行匹配的跨图文本阅读方法 Cross-image Text Reading Method Based on Text Line Matching 计算机科学, 2022, 49(9): 139-145. https://doi.org/10.11896/jsjkx.220600032 |
| [2] | 颜敏, 罗晓清, 张战成. 基于光传输模型学习的红外和可见光图像融合网络设计 Infrared and Visible Image Fusion Network Based on Optical Transmission Model Learning 计算机科学, 2022, 49(4): 215-220. https://doi.org/10.11896/jsjkx.210200174 |
| [3] | 江昊琛, 魏子麒, 刘璘, 陈俊. 非均衡数据分类经典方法综述与面向医疗领域的实验分析 Imbalanced Data Classification:A Survey and Experiments in Medical Domain 计算机科学, 2022, 49(1): 80-88. https://doi.org/10.11896/jsjkx.210200124 |
| [4] | 郑小萌, 高猛, 滕俊元. 航天器软件缺陷预测数据集构建方法研究 Research on Construction Method of Defect Prediction Dataset for Spacecraft Software 计算机科学, 2021, 48(6A): 575-580. https://doi.org/10.11896/jsjkx.200900133 |
| [5] | 和青芳, 王慧, 程光. 自适应小数据集乳腺癌病理组织分类研究 Research on Classification of Breast Cancer Pathological Tissues with Adaptive Small Data Set 计算机科学, 2021, 48(6A): 67-73. https://doi.org/10.11896/jsjkx.201000188 |
| [6] | 张曼, 李杰, 朱新忠, 沈霁, 成昊天. 基于改进DCGAN算法的遥感数据集增广方法 Augmentation Technology of Remote Sensing Dataset Based on Improved DCGAN Algorithm 计算机科学, 2021, 48(6A): 80-84. https://doi.org/10.11896/jsjkx.200700185 |
| [7] | 周彦, 陈少昌, 吴可, 宁明强, 陈宏昆, 张鹏. SCTD1.0:声呐常见目标检测数据集 SCTD 1.0:Sonar Common Target Detection Dataset 计算机科学, 2021, 48(11A): 334-339. https://doi.org/10.11896/jsjkx.210100138 |
| [8] | 王萧萧, 王亭雯, 马玉玲, 范佳奕, 崔超然. 基于深度森林的P2P网贷借款人信用风险评估方法 Credit Risk Assessment Method of P2P Online Loan Borrowers Based on Deep Forest 计算机科学, 2021, 48(11A): 429-434. https://doi.org/10.11896/jsjkx.201000013 |
| [9] | 杨连平, 孙玉波, 张红良, 李封, 张祥德. 基于编解码残差的人体关键点匹配网络 Human Keypoint Matching Network Based on Encoding and Decoding Residuals 计算机科学, 2020, 47(6): 114-120. https://doi.org/10.11896/jsjkx.200300079 |
| [10] | 宋传鸣, 洪旭, 王相海. 空-频域联合投票的交通视频阴影去除方法 Shadow Removal of Traffic Surveillance Video by Joint Voting in Spatial-Frequency Domain 计算机科学, 2020, 47(5): 129-136. https://doi.org/10.11896/jsjkx.190400040 |
| [11] | 张鹏飞, 李冠宇, 贾彩燕. 面向自然语言推理的基于截断高斯距离的自注意力机制 Truncated Gaussian Distance-based Self-attention Mechanism for Natural Language Inference 计算机科学, 2020, 47(4): 178-183. https://doi.org/10.11896/jsjkx.190600149 |
| [12] | 蔡强, 邓毅彪, 李海生, 余乐, 明少锋. 基于深度学习的人体行为识别方法综述 Survey on Human Action Recognition Based on Deep Learning 计算机科学, 2020, 47(4): 85-93. https://doi.org/10.11896/jsjkx.190300005 |
| [13] | 王莹, 郑丽伟, 张禹尧, 张晓妘. 面向中文APP用户评论数据的软件需求挖掘方法 Software Requirement Mining Method for Chinese APP User Review Data 计算机科学, 2020, 47(12): 56-64. https://doi.org/10.11896/jsjkx.201200031 |
| [14] | 沈琦, 陈逸伦, 刘枢, 刘利刚. 基于两级网络的三维目标检测算法 3D Object Detection Algorithm Based on Two-stage Network 计算机科学, 2020, 47(10): 145-150. https://doi.org/10.11896/jsjkx.190900172 |
| [15] | 张昉, 赵书良, 武永亮. 面向多尺度数据挖掘的数据尺度划分方法 Data Scaling Method for Multi-scale Data Mining 计算机科学, 2019, 46(4): 57-65. https://doi.org/10.11896/j.issn.1002-137X.2019.04.009 |
|
||
首页