摘要
为提升建筑施工安全管理的智能化水平,文章开展基于计算机视觉的建筑施工危险前兆信息识别研究,介绍基于计算机视觉的危险前兆信息识别方法,详细分析目标检测、数据预处理、模型训练及信息识别等关键技术环节,并提出一种融合YOLOv4目标检测算法与单应性矩阵空间映射的施工危险预警模型。结果表明,该方法能够有效识别施工现场中的危险相关对象,实现对危险前兆信息的准确提取与量化,为施工安全风险的早期预警提供技术支持。
To enhance the intelligent level of construction safety management,this article conducts research on the recognition of construction hazard precursor information based on computer vision.It introduces a method for identifying hazard precursor information using computer vision,and provides a detailed analysis of key technical aspects such as object detection,data preprocessing,model training,and information recognition.Additionally,it proposes a construction hazard warning model that integrates the YOLOv4 object detection algorithm with homography matrix spatial mapping.The results indicate that this method can effectively identify hazardous objects in the construction site,achieve accurate extraction and quantification of hazardous precursor information,and provide reliable technical support for early warning of construction safety risks.
出处
《智能城市》
2025年第11期157-160,共4页
Intelligent City
关键词
计算机视觉
建筑
施工
危险
前兆
信息识别
computer vision
architecture
construction
danger a precursor
information recognition
