Please wait a minute...
Journal of ZheJiang University (Engineering Science)  2019, Vol. 53 Issue (11): 2118-2128    DOI: 10.3785/j.issn.1008-973X.2019.11.009
Computer Technology and Control Engineering     
Estimation of river surface flow velocity based on conditional boundary equilibrium generative adversarial network
Wan-liang WANG1(),Sheng-lan YANG1,Yan-wei ZHAO2,Zhuo-rong LI1
1. College of Computer Science and Technology, Zhejiang University of Technology, Hangzhou 310023, China
2. College of Mechanical Engineering, Zhejiang University of Technology, Hangzhou 310023, China
Download: HTML     PDF(1959KB) HTML
Export: BibTeX | EndNote (RIS)      


Aiming at the difficulty in image classification due to the high similarity between different flow speeds, a conditional boundary equilibrium generative adversarial network and a convolutional classification network based on the multi-feature fusion were proposed to realize the generation and the classification of flow velocity images, respectively. A labeling mechanism and a verification module were introduced to realize the fitting and generation of corresponding category images, in order to achieve data enhancement. To enhance the impact of different texture features on velocity estimation, a multi-feature fusion layer was introduced to realize the feature extraction and the flow velocity recognition so as to realize the classification for images with small differences. The proposed method was applied to the actual river surface velocity estimation. Results demonstrate that the added tag information and the verification module can guide the data generation of corresponding class to a certain extent in the image generation module. Compared with other methods, the multi-feature fusion mechanism makes the proposed classifier more robust in identifying flow velocity images with small differences.

Key wordsflow rate estimation      generative adversarial network      feature fusion      velocity category      image verification     
Received: 20 September 2018      Published: 21 November 2019
CLC:  TP 391  
Cite this article:

Wan-liang WANG,Sheng-lan YANG,Yan-wei ZHAO,Zhuo-rong LI. Estimation of river surface flow velocity based on conditional boundary equilibrium generative adversarial network. Journal of ZheJiang University (Engineering Science), 2019, 53(11): 2118-2128.

URL:     OR


针对不同流速类类间差异小而造成的分类困难问题,提出条件边界平衡生成对抗网络和多特征融合的卷积分类网络,分别进行流速图像的生成和分类. 为了达到数据增强效果,引入标签机制和验证模块实现相应类别图像数据的拟合与生成;为了加强图像不同纹理特征信息对流速估测的影响,引入多特征融合机制对所有真实样本和生成伪样本进行特征提取和流速识别,实现对差异性较小的图像的分类. 将该方法应用于实际的河流表面流速估测,结果表明,在图像生成模块中,引入的标签信息和验证机制在一定程度上能强制引导模型的数据生成方向;在图像识别模块中,引入的多特征融合机制使所提出方法相较于其他方法,在差异性较小的水流图像的识别上更具鲁棒性.

关键词: 流速估测,  生成式对抗网络,  特征融合,  流速类别,  图像验证 
Fig.1 Structure diagram of generative adversarial networks
Fig.2 Schematic diagram of generator structure
Fig.3 Schematic diagram of discriminator structure
Fig.4 Schematic diagram of classification network structure
Fig.5 Framework diagram of flow velocity estimation based on CBEGAN
Fig.6 Schematic diagram of flow velocity measurement
Fig.7 Water image after pretreatment
Fig.8 Generated images of water without histogram equalization
Fig.9 Generated images of water with histogram equalization
Fig.10 Accuracy of water flow generated samples and analysis of category label loss function
Fig.11 Performance comparison of proposed model and VGG-16
Va/(m?s?1 W 流速估测结果
A:V=0~0.5 m/s B:V=0.5~1.0 m/s C:V=1.0~1.5 m/s D:V=1.5~2.0 m/s E:V=2.0~2.5 m/s
0.3 3.672 1.477 ?0.097 ?6.042 ?6.034 A
0.8 ?3.348 8.604 ?1.979 ?5.569 ?12.128 B
1.2 ?3.839 3.303 15.051 0.236 ?1.108 C
1.8 ?14.221 ?7.087 ?5.498 3.204 0.674 D
2.3 ?21.496 ?15.273 6.009 10.042 13.874 E
Tab.1 Flow rate estimation at different flow rates
识别方法 P/%
V0=0.2 m/s V0=0.5 m/s V0=0.8 m/s
BP神经网络 63.91 71.33 72.11
SVM 67.63 73.34 76.71
SRC(40维) 52.82 54.17 55.32
RRC(40维) 63.36 65.33 65.97
VGG-16 79.68 81.27 84.12
基于CBEGAN的流速识别算法 86.79 90.38 91.67
Tab.2 Accuracy comparison of multiple recognition methods at different flow rate resolutions
[1]   尚淑丽, 顾正华, 赵世凯, 等 基于模糊逻辑的水电工程生态效应综合评价[J]. 浙江大学学报:工学版, 2014, 48 (9): 1603- 1609
SHANG Shu-li, GU Zheng-hua, ZHAO Shi-kai, et al Comprehensive evaluation of ecological effects of hydropower projects based on fuzzy logic[J]. Journal of Zhejiang University: Engineering Science, 2014, 48 (9): 1603- 1609
[2]   周李智, 王万良, 徐新黎, 等 基于生态惩罚的混联梯级小水电群优化调度[J]. 水力发电学报, 2015, 34 (7): 45- 56
ZHOU Li-zhi, WANG Wan-liang, XU Xin-li, et al Optimal scheduling of hybrid cascade hydropower groups based on ecological penalty[J]. Journal of Hydroelectric Engineering, 2015, 34 (7): 45- 56
[3]   鲍鑫, 贺治国, 王振宇, 等 台风暴雨影响区溃坝洪灾淹没损失评估[J]. 浙江大学学报: 工学版, 2012, 46 (9): 1638- 1646
BAO Xin, HE Zhi-guo, WANG Zhen-yu, et al Evaluation of flooding damage caused by dam flood in typhoon storm-affected area[J]. Journal of Zhejiang University: Engineering Science, 2012, 46 (9): 1638- 1646
[4]   吴俊, 丁甡奇, 余葵, 等 光电非接触式表面流速测量[J]. 光学精密工程, 2010, 18 (2): 349- 356
WU Jun, DING Shen-qi, YU Kui, et al Photoelectric non-contact surface velocity measurement[J]. Optics and Precision Engineering, 2010, 18 (2): 349- 356
[5]   李笠. 考虑生态因素的小水电优化调度算法研究[D].杭州: 浙江工业大学, 2014.
LI Li. Research on small hydropower optimal scheduling algorithm considering ecological factors [D]. Hangzhou: Zhejiang University of Technology, 2014.
[6]   FULTON J, JH O N, OSTROWSKI J, et al Measuring real-time streamflow using emerging technologies: radar, hydroacoustics, and the probability concept[J]. Journal of Hydrology, 2008, 357 (1): 1- 10
[7]   COSTA J E, CHENG R T, HAENI F P, et al Use of radars to monitor stream discharge by noncontact methods[J]. Water Resources Research, 2006, 42 (7): 887- 896
[8]   HOU Y, WEN B, YANG Y, et al Two-dimensional river flow patterns observed with a pair of UHF radar system[J]. International Journal of Antennas and Propagation, 2017, 4792324
[9]   MONIN A S, YAGLOM A M. Statistical fluid mechanics, volume Ⅱ: mechanics of turbulence [M]. New York: Courier Corporation, 2007.
[10]   FUJITA I, KOMURA S Application of video image analysis for measurements of river-surface flows[J]. Proceedings of Hydraulic Engineering, 1994, 38: 733- 738
doi: 10.2208/prohe.38.733
[11]   阮晓东, 刘志皓, 瞿建武 粒子图像测速技术在两相流测量中的应用研究[J]. 浙江大学学报:工学版, 2005, 39 (6): 785- 788
RUAN Xiao-dong, LIU Zhi-hao, QU Jian-wu Application of particle image velocimetry in two-phase flow measurement[J]. Journal of Zhejiang University: Engineering Science, 2005, 39 (6): 785- 788
[12]   MESELHE E A, MUSTE M, PEEVA T Large scale particle image velocimetry for low velocity and shallow water flows[J]. Journal of Hydraulic Engineering, 2004, 130 (9): 937- 940
doi: 10.1061/(ASCE)0733-9429(2004)130:9(937)
[13]   TAURO F, GRIMALDI S, PORFIRI M Unraveling flow patterns through nonlinear manifold learning[J]. Plos One, 2014, 9 (3): e91131
doi: 10.1371/journal.pone.0091131
[14]   王万良, 邱虹, 郑建炜 基于压缩感知图像分析的河流表面流速估计方法[J]. 水力发电学报, 2018, 37 (5): 69- 79
WANG Wan-liang, QIU Hong, ZHENG Jian-wei Estimation method of river surface velocity based on compressed sensing image analysis[J]. Journal of Hydroelectric Engineering, 2018, 37 (5): 69- 79
doi: 10.11660/slfdxb.20180507
[15]   陈金鹏, 李贵山, 李双科, 等 基于机器视觉的流体速度量化研究[J]. 自动化仪表, 2012, 33 (12): 5- 9
CHEN Jin-peng, LI Gui-shan, LI Shuang-ke, et al Quantitative study of fluid velocity based on machine vision[J]. Automation Instrumentation, 2012, 33 (12): 5- 9
doi: 10.3969/j.issn.1000-0380.2012.12.002
[16]   SHABAN H, TAVOULARIS S Measurement of gas and liquid flow rates in two-phase pipe flows by the application of machine learning techniques to differential pressure signals[J]. International Journal of Multiphase Flow, 2014, 67: 106- 117
doi: 10.1016/j.ijmultiphaseflow.2014.08.012
[17]   王万良, 人工智能及其应用: 第三版 [M]. 北京: 高等教育出版社, 2016.
[18]   GOODFELLOW I J, POUGET-ABADIE J, MIRZA M, et al. Generative adversarial nets [C]// Advances in Neural Information Processing Systems. Montreal: ACM, 2014: 2672-2680.
[19]   CRESWELL A, WHITE T, DUMOULIN V, et al Generative adversarial networks: an overview[J]. IEEE Signal Processing Magazine, 2018, 35 (1): 53- 65
doi: 10.1109/MSP.2017.2765202
[20]   王坤峰, 苟超, 段艳杰, 等 生成式对抗网络GAN的研究进展与展望[J]. 自动化学报, 2017, 43 (3): 321- 332
WANG Kun-feng, GOU Chao, DUAN Yan-jie, et al Research progress and prospects of generative anti-network GAN[J]. Acta Automatica Sinica, 2017, 43 (3): 321- 332
[21]   王万良, 李卓蓉 生成式对抗网络研究进展[J]. 通信学报, 2018, 39 (2): 135- 149
WANG Wan-liang, LI Zhuo-rong Research progress of generating countermeasure networks[J]. Journal on Communications, 2018, 39 (2): 135- 149
doi: 10.3969/j.issn.1001-2400.2018.02.023
[22]   RATLIFF L J, BURDEN S A, SASTRY S S. Characterization and computation of local Nash equilibria in continuous games [C]// Communication, Control, and Computing. Monticello: IEEE, 2013: 917-924.
[23]   BERTHELOT D, SCHUMM T, METZ L. BEGAN: boundary equilibrium generative adversarial networks [EB/OL]. (2017-05-31).
[24]   MIRZA M, OSINDERO S. Conditional generative adversarial nets [EB/OL]. (2014-11-06).
[25]   ARJOVSKY M, CHINTALA S, BOTTOU L. Wasserstein GAN [EB/OL]. (2017-12-06).
[26]   LECUN Y, BENGIO Y Convolutional networks for images, speech, and time series[J]. Handbook of Brain Theory and Neural Networks, 1995, 3361 (10):
[27]   KRIZHEVSKY A, SUTSKEVER I, HINTON G E. ImageNet classification with deep convolutional neural networks [C]// International Conference on Neural Information Processing Systems. Lake Tahoe: ACM, 2012, 25(2): 1097-1105.
[28]   HE K, ZHANG X, REN S, et al. Deep residual learning for image recognition [C] // Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition. Las Vegas: IEEE, 2016: 770-778.
[29]   SIMONYAN K, ZISSERMAN A Very deep convolutional networks for large-scale image recognition[J]. Computer Science, 2014, 1 (2): 3
[30]   WRIGHT J, YANG A Y, GANESH A, et al Robust face recognition via sparse representation[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2009, 31 (2): 210- 227
doi: 10.1109/TPAMI.2008.79
[1] Hui-ya HU,Shao-yan GAI,Fei-peng DA. Face frontalization based on generative adversarial network[J]. Journal of ZheJiang University (Engineering Science), 2021, 55(1): 116-123.
[2] Pu ZHENG,Hong-yang BAI,Wei LI,Hong-wei GUO. Small target detection algorithm in complex background[J]. Journal of ZheJiang University (Engineering Science), 2020, 54(9): 1777-1784.
[3] Deng-wen ZHOU,Jin-yue TIAN,Lu-yao MA,Xiu-xiu SUN. Lightweight image semantic segmentation based on multi-level feature cascaded network[J]. Journal of ZheJiang University (Engineering Science), 2020, 54(8): 1516-1524.
[4] Kun LIU,Xi WEN,Min-ming HUANG,Xin-xin YANG,Jing-kun MAO. Solar cell defect enhancement method based on generative adversarial network[J]. Journal of ZheJiang University (Engineering Science), 2020, 54(4): 684-693.
[5] Hong LIN,Yao-yao LU. Scale differentiated text detection method focusing on hard examples[J]. Journal of ZheJiang University (Engineering Science), 2019, 53(8): 1506-1516.
[6] Ran DUAN,Deng-wen ZHOU,Li-juan ZHAO,Xiao-liang CHAI. Image super-resolution reconstruction based on multi-scale feature mapping network[J]. Journal of ZheJiang University (Engineering Science), 2019, 53(7): 1331-1339.
[7] Xiang-hao CHENG,Fei-peng DA,Liang WANG. Feature fusion based constrained local model for three-dimensional facial landmark localization[J]. Journal of ZheJiang University (Engineering Science), 2019, 53(4): 770-776.
[8] BAO Bi-sai, WU Jian-rong , LOU Xiao-jun, LIU Hai-tao. Feature fusion algorithm based on two-dimensional
feature matrix
[J]. Journal of ZheJiang University (Engineering Science), 2012, 46(11): 2081-2088.