隧道衬砌图像清晰度影响因素的模型试验研究

发布时间：2018-04-20 02:01

本文选题：隧道工程 + 隧道衬砌　；参考：《岩石力学与工程学报》2017年S2期

【摘要】：针对隧道衬砌图像清晰度,通过移动机械平台搭载线阵相机、光源(LED或LASER)、工业电脑等设备开展室内模型正交试验,采用一种融合图像幅度谱斜率和图像总变差的图像清晰度量化算法,以方差分析的手段研究光源类型、物面照度、曝光时间、物面直径等四个影响因素的显著性及优水平,并分析不同水平下图像清晰度的变化规律。试验结果表明:影响图像清晰度的主次因素顺序是曝光时间、光源类型、物面照度、物面直径,其中物面直径为不显著因素;试验因素水平下最优组合为LED,6 klux,90μs。研究成果可为基于计算机视觉和机器学习的隧道病害检测装备的研制提供参考依据。
[Abstract]:Aiming at the clarity of tunnel lining image, the indoor model orthogonal test is carried out by means of moving mechanical platform with linear camera, light source LED or LASERA, industrial computer, etc. An image definition quantization algorithm which combines the slope of image amplitude spectrum and the total variation of image is used to study the significance and superior level of four influencing factors, such as the type of light source, illumination of object surface, exposure time and surface diameter, by means of variance analysis. At the same time, the change rule of image sharpness at different levels is analyzed. The experimental results show that the order of primary and secondary factors affecting image definition is exposure time, light source type, illumination of object surface and surface diameter, in which the diameter of object surface is not significant, and the optimal combination of experimental factors is LEDX 6 klux90 渭 s. The research results can provide a reference for the development of tunnel disease detection equipment based on computer vision and machine learning.
【作者单位】：同济大学地下建筑与工程系;同济大学岩土及地下工程教育部重点实验室;
【基金】：国家自然科学基金资助项目(51778474) 上海市科学技术委员会资助项目(16DZ1200402)~~
【分类号】：TP391.41;U456

【相似文献】