工业管道内壁损伤重建与检测  

Reconstruction and Inspection of the Inner Wall Damage of Industrial Pipelines

作　　者：赵谦[1] 张莉华 钱渠 ZHAO Qian;ZHANG Li-hua;QIAN Qu(College of Communication and Information Engineering, Xi'an University of Science and Technology, Xi'an 710054, China;Intelligent Remote Sensing Division, Nantong Institute of Intelligent Perception, Nantong 226400, China)

机构地区：[1]西安科技大学通信与信息工程学院,西安710054 [2]南通智能感知研究院智能遥感事业部,南通226400

出　　处：《科学技术与工程》2021年第25期10796-10805,共10页Science Technology and Engineering

基　　金：国家自然科学基金(61801373);陕西省科技计划工业攻关项目(2017GY-073);西安市科技计划高校人才服务企业项目(2019217714GXR C013CG014-GXYD13.5)。

摘　　要：工业管道内壁的检测可以提前发现安全隐患,降低危害。针对管道内壁分辨率低、适用性差等特征,提出了工业管道内壁的重建方法。首先搭建内窥镜视频采集平台对管道内壁进行实时采集,采用张正友平面标定法进行标定处理,为了提高工业管道内壁检测的时效性,研究了几何成像关系对管道内表面图像进行径向拉伸的算法;采用改进的尺度不变特征转换(scale-invariant feature transform,SIFT)算法进行特征点提取与匹配,融合随机抽样一致(random sample consensus,RANSAC)算法和加权融合算法剔除误匹配对以及消除拼接缝,实现了管道内表面损伤图像的完整重建。实验结果表明,该算法时效性仅有1.5625 s,准确率可高达96%以上,为工业管道内壁损伤检测提供了新的思路。The detection technology of the inner wall of the industrial pipeline can detect the hidden danger of the pipeline in advance.A distortion-free reconstruction method of the inner wall of the industrial pipeline was proposed to overcome the characteristics of low resolution and poor applicability of the inner wall of the pipeline.Firstly,an endoscope video acquisition platform was built to collect the video of the inner wall of the pipeline,employing Zhang Zhengyou's plane calibration method to calibrate the camera.In order to improve the timeliness of the inspection of the inner surface of industrial pipeline,the radial stretching algorithm of the inner surface image of pipelines based on geometric imaging relation was studied.The improved scale-invariant feature transform(SIFT)algorithm was used to extract and match the feature points,using random sample consensus(RANSAC)algorithm and weighted fusion algorithm to eliminate the mismatched pairs and eliminate the splicing joints,so as to realize the complete reconstruction of the damaged image of the inner surface of the pipeline.Experimental results exhibit that the timeliness of the algorithm is only 1.5625 s,and the accuracy rate is greater than 96%,which provides a new idea for the detection of damage to the inner wall of industrial pipelines.

关 键 词：管道检测 径向拉伸 全景展开 SIFT特征提取 图像拼接 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]