钢丝绳
无损检测
漏磁
声学
材料科学
计算机科学
人工智能
计算机视觉
结构工程
工程类
磁铁
机械工程
物理
量子力学
作者
Pengbo Zheng,Juwei Zhang
标识
DOI:10.1080/10589759.2019.1590827
摘要
Detection of magnetic flux leakage (MFL) is widely used in nondestructive testing (NDT) of wire rope. We built and tested an MFL measuring device based on unsaturated magnetic excitation to address the problems of large volume, low accuracy and inconvenience of the traditional MFL instruments for locating and quantifying broken wires in wire ropes. We employed a pseudo-color imaging protocol to convert the MFL information to a pictorial representation. Localization of the breaks in the wire utilized the maximum modulus method. The color moment, statistical texture features, and spectral texture features were extracted from the image, revealing a network of broken wires, which could be localized and quantified. Testing showed that the device has high-detection accuracy and could detect small broken wires. Compared with traditional wire inspection devices, it has the advantages of small size, light weight and portability. We showed that pseudo-color image enhancement can effectively improve the recognition rate of broken wires and reduce the recognition error.
科研通智能强力驱动
Strongly Powered by AbleSci AI