PDF全文阅读

李海滨、纪晓鹏、郑祥、杨英、吴喜元

（中建深圳装饰有限公司，广东 深圳518000）

摘 要：为提高玻璃幕墙结构热损伤区域图像识别精度，提出基于改进机器视觉的玻璃幕墙结构热损伤区域图像识别方法。根据玻璃幕墙结构损伤热波传播原理，构建温度传递函数和玻璃幕墙结构热损伤区域概率模型，利用改进机器视觉增强处理玻璃幕墙结构热损伤图像，结合锐化处理实现对玻璃幕墙结构热损伤图像的识别。结果表明：方法识别出玻璃幕墙结构损伤区域与正常区域表面温度差别较大，其变化趋势符合玻璃幕墙实际情况，能够识别玻璃幕墙结构热损伤区域图像；在适当的热激励时间下，可以提高玻璃幕墙结构热损伤区域识别精度。

关键词：改进机器视觉；图像识别；热损伤；热波传播；玻璃幕墙

中图分类号:TP751            

文献标志码:A      

文章编号：1005-8249（2024）02-0164-06

DOI：10.19860/j.cnki.issn1005-8249.2024.02.030

 

LI Haibin, JI Xiaopeng,ZHENG Xiang, YANG Ying, WU Xiyuan

( China Construction Shenzhen Decoration Co.,Ltd,GUANGDONG SHENZHEN,518000,China)

Abstract: To improve the accuracy of image recognition of thermal damage areas in glass curtain wall structures, a method based on improved machine vision for image recognition of thermal damage areas in glass curtain wall structures is proposed. Based on the principle of thermal wave propagation in glass curtain wall structure damage, a temperature transfer function and a probability model of thermal damage area in glass curtain wall structure are constructed. The improved machine vision is used to enhance the processing of thermal damage images in glass curtain wall structure, and combined with sharpening processing, the recognition of thermal damage images in glass curtain wall structure is achieved. The results show that the method proposed in the article identifies a significant difference in surface temperature between the damaged and normal areas of the glass curtain wall structure, and its changing trend is consistent with the actual situation of the glass curtain wall. It can identify images of thermal damage areas in the glass curtain wall structure; Under appropriate thermal excitation time, the accuracy of identifying thermal damage areas in glass curtain wall structures can be improved.

Keywords: improve machine vision; image recognition; thermal damage; thermal wave propagation; glass screen wall

 

作者简介：李海滨（1981-），男，本科，工程师，主要研究方向：室内外装饰装修及结构力学施工技术。

收稿日期：2023-11-21