SBIR Phase I: Advanced Terahertz Image Processing Systems

小企业创新研究(SBIR)第一阶段项目的更广泛影响将通过解决两个问题来加强公共安全和公共卫生：1)私营和公共部门每年收到约3,500封含有白色粉末的信件和超过5,500份爆炸威胁；这些可能是导致临时关闭和生产力损失的恶作剧；或者更严重的是，它们可能含有炭疽、蓖麻毒素、芬太尼或爆炸装置。2)阿片类药物的流行是由小包、药丸或液体配方中包含的粉末的国际递送助长的；所有这些都需要进行非侵入性检测。该项目将开发新的图像处理技术，利用太赫兹成像来检测隐藏在信件和包裹中的有害物品。如果成功，由这些图像分析算法驱动的太赫兹扫描仪可能会像复印机、打印机和文档扫描仪一样在任何公司办公室无处不在，解决关键的公共卫生和安全挑战。拟议的项目将开发图像处理和计算机视觉技术，以利用通过台式太赫兹扫描仪捕获的多维、动态3-D太赫兹成像数据集。不存在商业上可行的方法，能够从用于安全应用的太赫兹成像数据流中实时自动检测威胁。该项目旨在开发强大的技术，既增强操作员的能力，又自动检测特定的威胁。一个不小的挑战是在移动的参照系的背景下检测液体/粒子的运动--例如，隐藏在移动的包裹中的小袋子内的液体的运动。太赫兹成像系统能够探测单滴液体(50毫克)，并实时成像炸药中粉末的运动。该项目的另一个重点领域将是将3-D太赫兹数据集与从其他传感器输入收集的上下文信息以及由于设备(物联网)的连接性质而收集的外部来源相结合。这些信息与太赫兹数据的融合将提供无与伦比的检测能力，并在动态、3-D成像领域产生深远影响，远远超出威胁检测。这一奖项反映了NSF的法定使命，并通过使用基金会的智力优势和更广泛的影响审查标准进行评估，被认为值得支持。

The broader impact of this Small Business Innovation Research (SBIR) Phase I project will enhance both public safety and public health by addressing two problems: 1) Approximately 3,500 letters containing white powders and over 5,500 explosive threats are received by private and public offices each year; these may be hoaxes that result in temporary shut-down and lost productivity; or more critically, they may contain anthrax, ricin, fentanyl, or an explosive device. 2) The opioid epidemic is fueled by international delivery of powders contained in small packets, pills, or liquid formulations; all require to non-invasive detection. This project will develop novel image processing techniques utilizing terahertz imaging to detect harmful items concealed within letters and packages. If successful, terahertz scanners powered by these image analysis algorithms may become as ubiquitous as copiers, printers, and document scanners in any corporate office, addressing critical public health and safety challenges.The proposed project will develop image processing and computer vision techniques to capitalize on the multidimensional, dynamic 3-D terahertz imaging dataset captured via desktop-based terahertz scanners. No commercially-viable approaches exist, capable of automatically detecting threats in real-time from terahertz imaging data streams for security applications. This project aims to develop robust techniques to both enhance the capabilities of a human operator and automatically detect specific threats. One non-trivial challenge is the detection of liquid/particle motion within the context of a moving reference frame - such as the motion of liquid within a small pouch, itself concealed within a moving package. The terahertz imaging system is capable of detecting single drops of liquid (50 mg) and imaging the movement of powders within explosives in real-time. Another focus area for the project will be the integration of the 3-D terahertz data set together with contextual information gleaned from other sensor inputs, as well as external sources due to the connected nature of the device (internet of things). The fusion of this information with the terahertz data will provide unparalleled detection capabilities and have far reaching impacts in the field of dynamic, 3-D imaging, well beyond threat detection.This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.