RAPID: Automatic and Non-intrusive Screening for Potential Viral Disease Carriers

RAPID：自动、非侵入性筛查潜在病毒性疾病携带者

• 批准号: 1509625
• 负责人: Guodong Guo
• 金额: $ 8.78万
• 依托单位: West Virginia University Research Corporation
• 依托单位国家: 美国
• 项目类别: Standard Grant
• 财政年份: 2014
• 资助国家: 美国
• 起止时间: 2014-12-01 至 2016-11-30
• 项目状态: 已结题
• 来源: https://www.nsf.gov/awardsearch/showAward?AWD_ID=1509625&HistoricalAwards=false
• 关键词: RAPID; Automatic; Non; intrusive; Screening

Viral disease outbreaks such as Ebola epidemic can be a serious concern for the general public. Many these viral diseases cause a temperature increase in human body and can be visually detected from infrared images. This project develops an intelligent system to automatically detect human body temperature for non-intrusively screening potential viral disease carriers from a crowd. The system is built on recent advances in computer vision technologies and thermal infrared sensors. This project addresses an important and interesting application of computer vision. The research leads to development of a prototype system that can be deployed in a public pace to automatically measure human body temperatures from facial images. The research work includes discovering how to localize the facial features accurately and precisely from both visible and infrared images for the purpose of accurate measure of body temperatures and reducing false alarms in screening of potential viral disease carriers. The key approach utilizes both visible light and infrared spectra to achieve a robust solution. It can also process multiple people in a crowd non-intrusively and speed up the screening process. The research in this project can address the critical need for screening of viral disease carriers, and advance the study on machine vision algorithms applicable to an important and practical problem related to public health.

埃博拉疫情等病毒性疾病的爆发可能是公众严重关注的问题。许多这些病毒性疾病会导致人体温度升高，并且可以从红外图像中直观地检测到。本项目开发一种智能系统，自动检测人体温度，以非侵入性地从人群中筛选潜在的病毒性疾病携带者。 该系统建立在计算机视觉技术和热红外传感器的最新进展之上。这个项目解决了计算机视觉的一个重要而有趣的应用。这项研究导致了一个原型系统的开发，该系统可以部署在公共场所，从面部图像中自动测量人体温度。研究工作包括发现如何从可见光和红外图像中准确定位面部特征，以准确测量体温并减少潜在病毒性疾病携带者筛查中的误报。关键方法利用可见光和红外光谱来实现稳健的解决方案。它还可以非侵入性地处理人群中的多个人，并加快筛选过程。本项目的研究可以解决病毒性疾病携带者筛查的迫切需要，并推进适用于公共卫生相关的重要和实际问题的机器视觉算法的研究。