西安工程大学学报2025，Vol.39Issue(2)：47-56,10.DOI:10.13338/j.issn.1674-649x.2025.02.006
基于SuNet的公共交通安检违禁品的检测
Prohibited items detection in public transportation security inspection based on SuNet
摘要
Abstract
In the scenario of security inspection in public transportation,the overlapping of pro-hibited and non-prohibited items made it difficult for existing models to effectively identify ob-scured prohibited item categories.To address this issue,a prohibited item detection model based on SuNet was proposed in this paper.Firstly,an augmented attention localization feature pyramid network(AALFPN)was designed to enhance the semantic information of prohibited items and fuse the localization information and semantic information of prohibited items to guide the model in accurately locating obscured prohibited items,enhancing the feature contour of prohibited i-tems.Secondly,a dense attention mechanism(DAM)was introduced to effectively identify and extract obscured prohibited items.Finally,the SmoothL1 Loss loss function was introduced to ad-dress the problem of loss of prohibited item category information during regression.To verify Su-Net's ability to effectively identify obscured prohibited item categories,this study conducted ex-periments on the PIDray dataset.To assess SuNet's generalization on other prohibited item data-sets,this study conducted experiments on the CLCXray dataset.Experimental results show that on the PIDray dataset,compared to the RoIAttn model,the SuNet improves by 2.9%,4.4%and 3.3%on the AP@0.5∶0.95,AP@0.5 and AP@0.75 metrics,respectively.On the CLCXray da-ta set,compared to the RoIAttn model,the SuNet improves by 1.4%,1.4%and 0.4%on the AP@0.5∶0.95,AP@0.5 and AP@0.75 metrics,respectively.The experimental results demon-strate that SuNet not only effectively identifies obscured prohibited item categories but also ex-hibits good generalization performance on other prohibited item datasets,providing an effective solution for prohibited item detection in the public transportation security inspection scenario.关键词
违禁品检测/SuNet/强化注意力定位特征金字塔网络/密集注意力机制/SmoothL1 LossKey words
prohibited item detection/SuNet/augmented attention localization feature pyramid network(AALFPN)/dense attention mechanism(DAM)/SmoothL1 Loss分类
计算机与自动化引用本文复制引用
张缓缓,刘鹏程,姜萌,王雨欣..基于SuNet的公共交通安检违禁品的检测[J].西安工程大学学报,2025,39(2):47-56,10.基金项目
国家自然科学基金青年项目(61902302) （61902302）
陕西省科技厅重点研发计划项目(2024GX-YBXM-231) （2024GX-YBXM-231）
浙江省博士后科研项目择优资助(ZJ2022154) （ZJ2022154）
