[1]沙爱民,蔡若楠,高杰,等.基于级联卷积神经网络的公路路基病害识别[J].长安大学学报(自然科学版),2019,39(02):1-9.
　SHA Ai min,CAI Ruo nan,GAO Jie,et al.Subgrade distresses recognition based on convolutional neural network[J].Journal of Chang’an University (Natural Science Edition),2019,39(02):1-9.
点击复制

基于级联卷积神经网络的公路路基病害识别()
分享到：

长安大学学报（自然科学版）[ISSN:1006-6977/CN:61-1281/TN]

卷:
第39卷
期数:
2019年02期
页码:
1-9
栏目:
道路工程
出版日期:
2019-03-31

文章信息/Info

Title:
Subgrade distresses recognition based on convolutional neural network
作者:
沙爱民; 蔡若楠; 高杰; 童峥; 李帅
(1. 长安大学 公路学院，陕西 西安 710064； 2. 长安大学 材料科学与工程学院，陕西 西安 710061)
Author(s):
SHA Aimin;  CAI Ruonan;  GAO Jie;  TONG Zheng;  LI Shuai
(1. School of Highway, Changan University, Xian 710064, Shaanxi, China;2. School of Materials Science and Engineering, Changan University, Xian 710061, Shaanxi, China)
关键词:
道路工程; 卷积神经网络; 探地雷达; 路基病害检测; 图像处理
Keywords:
road engineering;  convolutional neural network;  GPR;  subgrade distresses detection;  image processing
文献标志码:
A
摘要:
针对当前公路路基病害识别中探地雷达（GPR）技术的数据分析还依赖于人工识别，识别效率低、准确性差问题，建立了级联卷积神经网络来实现自动识别探地雷达图像所反映路基病害的任务。级联卷积神经网络系统由2个卷积神经网络组成，分别用于识别低分辨率和高分辨率探地雷达图像。神经网络的建立包括训练、验证和测试3个步骤。通过训练和测试的结果验证了级联卷积神经网络系统的稳定性，并将级联卷积神经网络和Sobel边缘检测，K值聚类分析进行比较，以论证其优越性。结果表明：级联卷积神经网络在路基病害分类训练中的识别准确率为97.46％，验证中的识别准确率为95.80％，其识别路基病害的精度较高；级联卷积神经网络对发射频率300、500、900 MHz 的图像分类准确率分别为94.20％、93.89％、94.57％，对不同公路结构的图像分类准确率分别为94.80％、94.78％、94.28％、94.21％，可见级联卷积神经网络的识别准确性不受雷达发射频率和路面结构的影响；当图像分辨率较低时，Sobel边缘检测和K值聚类分析无法准确提取路基病害几何特征信息，级联卷积神经网络可通过分类器2准确识别；当图像分辨率较高时，Sobel边缘检测和K值聚类分析仅能提取路基病害的部分特征，后续病害类型的识别需要人工完成。可见级联卷积神经网络较其他算法在路基病害识别方面更准确、高效。
Abstract:
Aimed at the data analysis of ground penetrating radar (GPR) technology for detecting subgrade distresses relies on manual identification，which classifying distresses was inefficient and inaccurate. The application of convolutional neural networks (CNN) was put forward to classify subgrade distresses automatically. The cascaded CNN consists of two convolutional neural networks, which were separately utilized to recognize distresses using lowresolution images and highresolution images. The processes for developing convolutional neural networks mainly include training, validating and testing. After developing the cascaded CNN, training and testing results were used to verify the stability of the cascaded CNN. The cascaded CNN was then compared with Sobel edge detection and Kvalue clustering analysis to demonstrate its superiority. The results show that the accuracies of the cascaded CNN in the training and validating processes were 97.46％ and 95.80%. The cascaded CNN has a high accuracy in identifying subgrade distress. When the cascaded CNN was operated at frequencies of 300, 500 and 900 MHz, the accuracies of image classification were 94.20%, 93.89%, and 94.57%, respectively. When dealing with different highway structures, the accuracies of the images were 94.80%, 94.78%, 94.28%, and 94.21%. The cascaded CNN has great stability with respect to both emission frequencies and highway structures. When the image resolution is low, Sobel edge detection and Kvalue clustering analysis cannot extract the distress information accurately. However, CNN can extract the distress information accurately with classifier 2. When the image resolution is high, Sobel edge detection and Kvalue clustering analysis can extract only some of the subgrade distress. The remaining distress information needs to be extracted manually. The cascaded CNN can identify subgrade cracks accurately and efficiently，compared with other algorithms. 3 tabs, 10 figs, 28 refs.

相似文献/References:

[1]武建民,祝伟,马士让,等.应用加权密切值法评价基质沥青抗老化性能[J].长安大学学报(自然科学版),2012,32(01):0.
[2]张宜洛,袁中山.SMA混合料结构参数的影响因素[J].长安大学学报(自然科学版),2012,32(01):0.
[3]陈璟,袁万杰,郝培文,等.微观指标对沥青热稳定性能的影响[J].长安大学学报(自然科学版),2012,32(01):0.
[4]周兴业,刘小滔,王旭东,等.基于轴载谱的沥青路面累计当量轴次换算[J].长安大学学报(自然科学版),2012,32(01):0.
[5]李祖仲,王伯禹,陈拴发,等.轴载对复合式路面应力吸收层荷载应力的影响[J].长安大学学报(自然科学版),2012,32(01):0.
[6]关博文,刘开平,陈拴发,等.水镁石纤维路面混凝土路用性能[J].长安大学学报(自然科学版),2012,32(01):0.
[7]翁效林,王玮,张留俊,等.拓宽路基荷载下管桩复合地基沉降变形模式[J].长安大学学报(自然科学版),2012,32(01):0.
[8]穆柯,王选仓,柳志军,等.基于非饱和渗流原理的路基含水率预估[J].长安大学学报(自然科学版),2012,32(01):0.
[9]李振霞,陈渊召.不同类型半刚性基层材料性能的试验与分析[J].长安大学学报(自然科学版),2012,32(01):0.
[10]马 骉,马 晋,周宇鹏.沥青混合料降温收缩断裂特性[J].长安大学学报(自然科学版),2012,32(03):1.
　MA Biao,MA Jin,ZHOU Yu-peng.Cooling shrinkage fracture characteristic of asphalt mixture[J].Journal of Chang’an University (Natural Science Edition),2012,32(02):1.

更新日期/Last Update: 2019-04-01