可视化识别技术在进境旅邮检动植物源风险监管中的应用  

Application of visual identification technology in risk supervision sourced from animal and plant in entry mail and entry passger belongings

作　　者：胡琳子 李凯兵 袁淑珍 刘扬东 王祥 刘海军 Hu Linzi;Li Kaibing;Yuan Shuzhen;Liu Yangdong;Wang Xiang;Liu Haijun(Guangzhou Customs District,Guangzhou 510623,China;Foshan Customs House;Erlianhaote Customs House;Baiyun Airport Customs House;Maoming Institute of Forestry Science)

机构地区：[1]广州海关,广东广州510623 [2]佛山海关 [3]二连浩特海关 [4]白云机场海关 [5]茂名市林业科学研究所

出　　处：《植物检疫》2020年第4期17-21,共5页Plant Quarantine

基　　金：国家重点研发计划(2018YFC0809200)。

摘　　要：本文综述了当前可用的非接触式可视化识别技术及中国口岸的应用现状,在刘扬东等提出的针对检疫物的CT图像着色和自动识别算法基础上,利用传统的X光机和CT检查装置,汇集其他安检领域过往的图像与数据,研发识别动植物源产品的算法,提高旅邮检中动植物产品的识别率,减少人为误差。This paper summarizes the currently available non-contact visual identification technology and the application status of Chinese ports. Basing on Liu Yangdong’s CT image coloring and automatic recognition algorithm for quarantine objects,using the traditional X-ray machine and CT inspection device,combining with images and data collected in other security inspection fields for past years,we develop the inspection device to identify animal and plant products,so as to improve the identification rate and reduce human error in entry mail quarantine and entry passger belongings quarantine.

关 键 词：可视化X光 动植物产品 图像处理 图像识别 人工智能 

分 类 号：S41[农业科学—植物保护]