基于计算机视觉的钢结构表面缺陷智能识别研究综述

Research review on intelligent recognition of steel structure surface defects based on computer vision

我国是钢结构工程建造量和保有量最多的国家,传统的钢结构缺陷工程诊断方法已无法满足当前的需求。而随着计算机技术的快速发展,将人工智能引入工程诊断领域逐步成为研究热点。根据对国内外研究现状的总结,提供了常见的钢结构表面缺陷基于计算机视觉的智能识别方法的研究综述,并结合国内实际情况总结出一套通用的技术框架。首先总结了表面缺陷智能检测与识别的通用流程,即:图像采集、目标提取和缺陷识别,然后以此为框架介绍了计算机视觉各种方法的内容和原理。接下来分别介绍了钢结构三种常见表面缺陷:高强螺栓缺失与松动、表面锈蚀、表面裂缝智能识别技术最新的研究进展和当前方法的不足。最后讨论和展望了基于计算机视觉的表面缺陷智能识别未来的研究方向。

China is the country with the largest number of steel structure construction and ownership.The traditional engineering defect diagnosis method of steel structure can not meet the current demand.With the rapid development of computer technology,the introduction of artificial intelligence into the field of engineering diagnosis has gradually become a research hotspot.According to the summary of the research status at home and abroad,a review of the research on intelligent recognition of common steel structure surface defects based on computer vision was provided,and a set of general technical framework combined with the actual situation in China was summarized.Firstly,the general process of surface defect intelligent detection and recognition were summarized,namely:image acquisition,target extraction and defect recognition,and then uses this as a framework to introduce the content and principles of various computer vision methods.Next,three kinds of common surface defects of steel structure were introduced:lack and loose-ness of high-strength bolts,surface corrosion,and the latest research progress of intelligent identification technology of surface cracks and the shortcomings of current methods.Finally,the future research direction of surface defect intelligent recognition based on computer vision was discussed and prospected.