[关键词]
[摘要]
为提升城市轨道交通安检系统判图效率及准确率,本文设计一种新颖的判图模式将AI(人工智能)图像识别技术与人工集中判图技术有效结合。首先针对逢液必检现状,引入液体检测算法避免对安全液体的开包查验,其次按安检品风险等级分类处理,最后结合AI置信度判断、人工抽检或必检判图对安检品进行判定。本判图模式可根据不同阶段下AI图像识别准确度和判图需求灵活调节AI介入深度,随着AI图像识别准确度的不断提升,逐渐从AI为辅人工为主判图模式向AI为主人工为辅判图模式平滑渐进迭代,最终实现完全智能判图模式。通过案例分析可知,在不降低城市轨道交通车站安检水平的前提下,本判图模式可以进一步达到前端过检增速后端降本增效的效果。
[Key word]
[Abstract]
In order to improve the efficiency and accuracy of urban rail transit security inspection system, this paper designs a novel pattern recognition mode to effectively combine AI ( artificial intelligence ) image recognition technology with manual centralized pattern recognition technology.Firstly, according to the current situation of liquid inspection, the liquid detection algorithm is introduced to avoid the open inspection of safe liquid. Secondly, the security products are classified according to the risk level. Finally, the AI confidence judgment, manual sampling or necessary inspection chart are combined to determine.The pattern recognition mode can flexibly adjust the depth of AI intervention according to the accuracy of AI image recognition and the needs of pattern recognition at different stages. With the continuous improvement of the accuracy of AI image recognition, it gradually changes from AI-assisted manual-based pattern recognition mode to AI-based manual-assisted pattern recognition mode, and finally realizes the fully intelligent pattern recognition mode.Through case analysis, it can be seen that the judgment graph model can further achieve the effect of rapid security inspection, cost reduction and efficiency increase without reducing the safety inspection level of urban rail transit stations.
[中图分类号]
[基金项目]