Human behaviour detection dataset (HBDset) using computer vision for evacuation safety and emergency management

作     者：Yifei Ding Xinghao Chen Zilong Wang Yuxin Zhang Xinyan Huang 

作者机构：Research Centre for Fire Safety Engineering Department of Building Eviroment and Energy EngineeringThe Hong Kong Polytechnic UniversityHong Kong China State Key Laboratory of Disaster Reduction in Civil EngineeringTongji UniversityShanghaiChina Departnent of Geotechnical EngineeringTongji UniversityShanghaiChina 

出 版 物：《Journal of Safety Science and Resilience》 (安全科学与韧性（英文）)

年 卷 期：2024年第5卷第3期

页      面：355-364页

核心收录：

学科分类：08[工学] 0812[工学-计算机科学与技术（可授工学、理学学位）] 

基　　金：funded by the Hong Kong Research Grants Council Theme-based Research Scheme(T22-505/19-N) the National Natural Science Foundation of China(52204232) MTR Research Fund(PTU-23005) 

主　　题：Image dataset Object detection Human behaviour Public safety Evacuation process 

摘      要：During emergency evacuation,it is crucial to accurately detect and classify different groups of evacuees based on their behaviours using computer *** object detection models trained on standard image databases often fail to recognise individuals in specific groups such as the elderly,disabled individuals and pregnant women,who require additional assistance during *** address this limitation,this study proposes a novel image dataset called the Human Behaviour Detection Dataset(HBDset),specifically collected and anno-tated for public safety and emergency response *** dataset contains eight types of human behaviour categories,*** normal adult,child,holding a crutch,holding a baby,using a wheelchair,pregnant woman,lugging luggage and using a mobile *** dataset comprises more than 1,5o0 images collected from various public scenarios,with more than 2,9oo bounding box *** images were carefully selected,cleaned and subsequently manually annotated using the Labellmg *** demonstrate the effectiveness of the dataset,classical object detection algorithms were trained and tested based on the HBDset,and the average detection accuracy exceeds 90%,highlighting the robustness and universality of the *** developed open HBDset has the potential to enhance public safety,provide early disaster warnings and prioritise the needs of vulnerable individuals during emergency evacuation.