直升机电力巡检系统中自动高效的塔台检测
The Study of Automatic and Efficient Tower Detection in the Helicopter Power Line Inspection System
DOI: 10.12677/JISP.2018.73014, PDF,    国家自然科学基金支持
作者: 赵 凡, 吉 璐, 张海燕, 杨 丹：西安理工大学印刷包装与数字媒体学院信息科学系，陕西 西安
关键词: 目标检测；机器学习；直升机电力巡检系统；AdaBoost分类器；CNN分类器；Object Detection； Machine Learning； Helicopter Power Line Inspection System； AdaBoost Classifier； CNN Classifier
摘要: 针对直升机电力巡检系统中塔台的自动高效检测问题，提出了基于学习的两级塔台检测方法，首先利用AdaBoost分类器在多尺度上对塔台进行粗检，得到的初检结果作为候选塔台，再利用基于深度学习的CNN分类器对候选目标进行确认，从而对飞行视频中的塔台进行定位。通过实验仿真，证明了算法的高效性。
Abstract: In order to solve the problem of automatic and efficient detection of tower in the helicopter power line inspection system, this paper proposes two-level tower detection method based on machine learning. Firstly, the trained AdaBoost classifier is used for coarse detections of towers in multi scales. Secondly, as candidate towers, the detected results by AdaBoost classifier are put to learned CNN classifier for identification of the candidate targets, and power towers are located thereby. The simulation results show that the algorithm is efficient.
文章引用：赵凡, 吉璐, 张海燕, 杨丹. 直升机电力巡检系统中自动高效的塔台检测[J]. 图像与信号处理, 2018, 7(3): 119-127. https://doi.org/10.12677/JISP.2018.73014

参考文献

[1] 张元军, 李清华. 电力铁塔运行状态智能在线监测的研究及应用[J]. 科技视界, 2016(22): 9-10.
[2] Fortes, M.Z., Ferreira, V.H., et al. (2015) Fault Diagnosis in Transmission Lines: Trends and Main Research Areas. IEEE Latin America Transactions, 13, 3324-3332.
[Google Scholar] [CrossRef] 
[3] 夏凌云, 安中印. 基于轨迹导航的电力巡检业务应用[J]. 软件导刊, 2014, 13(12): 86-88.
[4] Baker, L., Mills, S., et al. (2016) Power Line Detection Using Hough Transform and Line Tracing Techniques. 2016 International Conference on Image and Vision Computing New Zealand (IVCNZ), Palmerston North, 21-22 No-vember 2016, 1-6.
[Google Scholar] [CrossRef] 
[5] Steiger, O., Lucas, E. and Maret, Y. (2014) Automatic Detection of Transmission Towers. Sensors, 2014 IEEE, Valencia, 2-5 November 2014, 1034 -1037.
[Google Scholar] [CrossRef] 
[6] 吴华, 郭贞, 杨国田, 等. 基于全局自相似描述子的电塔检测[J]. 华中科技大学学报: 自然科学版, 2011, 39(S2): 437-440.
[7] 吴晓东, 李瑛, 刘晓伟. 基于多尺度联合分割的SAR图像电线走廊变化检测[J]. 电力科学与技报, 2012, 27(4): 51-56.
[8] 张雪峰, 周赞东, 龚浩, 等. 基于SIRV杂波模型的SAR图像输电铁塔检测[J]. 武汉大学学报: 工学版, 2013, 46(6): 737-741.
[9] Sampedro, C., Martinez, C., Chauhan, A., et al. (2014) A Supervised Approach to Electric Tower Detection and Classification for Power Line Inspection. 2014 International Joint Conference on Neural Networks (IJCNN), Beijing, 6-11 July 2014, 1970-1977.
[10] Tang, J., Ding, R., et al. (2013) Real-Time Target Detection Algorithm Based on Background Molding. 2013 4th International Conference on Digital Manufacturing & Automation, Qingdao, 29-30 June 2013, 970-974.
[Google Scholar] [CrossRef] 
[11] 张明军, 俞文静, 袁志, 等. 视频中目标检测算法研究[J]. 软件, 2016, 37(4): 40-45.
[12] Freund, Y. and Schipare, R.E. (1997) A Decision-Theoretic Generalization of On-Line Learning and an Application to Boosting. Journal of Computer and System Sciences, 55, 119-139.
[Google Scholar] [CrossRef] 
[13] Guo, W., Xiao, Y. and Zhang, G.D. (2014) Multi-Scale Pedestrian Detection by Use of AdaBoost Learning Algorithm. 2014 International Conference on Virtual Reality and Visualization, Shenyang, 30-31 August 2014, 266-271.
[Google Scholar] [CrossRef] 
[14] Ojala, T. (2002) Multi-Resolution Gray-Scale and Rotation Invariant Texture Classification with Local Binary Patterns. IEEE Transactions on Pattern Analysis and Machine Intelligence, 24, 971-987.
[Google Scholar] [CrossRef] 
[15] 王德才, 秦亮曦. 基于LBP特征和AdaBoost的行人检测方法[J]. 河北软件职业技术学院学报, 2016, 18(3): 51-54.
[16] Hinton, G.E. and Salakhutdinov, R.R. (2006) Reducing the Dimensionality of Data with Neural Networks. Science, 313, 504-507.
[Google Scholar] [CrossRef] [PubMed]
[17] Krizhevsky, A., Sutskever, I. and Hinton, G.E. (2012) Image Net Classification with Deep Convolutional Neural Networks. Proceeding of Advances in Neural Information Processing Systems, 25, 1097-1105.
[18] 张昭旭. CNN深度学习模型用于表情特征提取方法探究[J]. 现代计算机(专业版), 2016(3): 41-44.
[19] 王茜, 张海仙. 深度学习框架Caffe在图像分类中的应用[J]. 现代计算机: 专业版, 2016(5): 72-75.