高校餐厅食品加工违规行为视频分析算法与系统设计

Algorithm and system design of video analysis for the violations in food processing in university restaurants

中文关键词:  食品安全  违规行为  视频分析系统  视频分析算法  混合编程

英文关键词:food safety  violations  video analysis system  video analysis algorithm  hybrid programming

中文摘要:

      目的 开发视频分析子系统分析食品加工违规行为, 提高高校餐厅的食品安全风险预防能力。方法 介绍了该视频分析子系统的组成和违规行为分析过程的主要工作流程; 详细阐述了检测员工不着工作服、用素食案板加工生肉2种典型违规行为的视频分析算法的基本原理、主要公式和算法流程; 并对2种典型违规行为进行了大量测试。结果 本视频分析系统对员工未着工作服行为的检测准确率达到了99%, 对员工在素食案板上加工生肉行为的检测准确率达到了98%。结论 本研究所开发的视频分析子系统可提高预防食品安全风险的能力。

英文摘要:

      Objective To develop video analysis subsystem for analyzing the food processing violations, in order to improve the food safety risk prevention ability of university restaurants. Methods The composition of the video analysis subsystem and the main working process of the violation analysis process were introduced. The basic principle, main formula and algorithm flow of video analysis algorithm for detecting 2 typical violations (non-working clothes and raw meat processed on vegetarian chopping board) were elaborated. A large number of tests were conducted on 2 typical irregularities. Results The video analysis system established in this study was 99% accurate in detecting the behaviors of employees without working clothes, and 98% accurate in detecting the behaviors of employees processing raw meat on the vegetarian chopping board. Conclusion The developed video analysis subsystem can improve the ability to prevent food safety risks.