Deep Convolutional Neural Networks for Identification, Tracking, and Classification of Transport Vehicles using Multiple-view Imaging Systems

使用多视图成像系统的深度卷积神经网络对运输车辆进行识别、跟踪和分类

• 批准号: 536767-2018
• 负责人: Mehrandezh, Mehran
• 金额: $ 2.51万
• 依托单位: University of Regina
• 依托单位国家: 加拿大
• 项目类别: Collaborative Research and Development Grants
• 财政年份: 2019
• 资助国家: 加拿大
• 起止时间: 2019-01-01 至 2020-12-31
• 项目状态: 已结题
• 来源: https://www.nserc-crsng.gc.ca/ase-oro/Details-Detailles_eng.asp?id=685427
• 关键词: Deep; Convolutional; Neural; Networks; Identification

This project revolves around the concept of Identification, Tracking, and Classification (ITC) of transport vehicles passing through inspection stations using imaging sensors. Multiple-view machine-learning-based algorithms will be developed that is robust to partial occlusion in the imaging sensors and to the light scatter due to differential illumination and weather conditions for the first time.A close collaboration between IRD and the University of Regina (UofR) will generate a unique partnership between industry and academia that can pave the way towards training highly-qualified personnel at both institutes. Furthermore, the transport vehicle inspection facilities, such as the one in the Global Transportation Hub (GTH) will be used as a living lab located at close proximity to the UofR, where our research results and findings can be tested and utilized. This project has the potential to lead to developing an emerging engineering discipline in intelligent transportation in Saskatchewan for the first time.

该项目围绕着使用成像传感器对通过检查站的运输车辆进行识别、跟踪和分类（ITC）的概念。 将开发基于多视图机器学习的算法，该算法对成像传感器中的部分遮挡以及由于不同照明和天气条件引起的光散射具有鲁棒性。IRD与里贾纳大学（UofR）之间的密切合作将在工业界和学术界之间建立独特的伙伴关系，为两个研究所培养高素质人才铺平道路。 此外，运输车辆检测设施，如全球运输中心（GTH）的检测设施，将被用作靠近UofR的生活实验室，我们的研究成果和发现可以在那里得到测试和利用。该项目有可能导致在萨斯喀彻温省首次开发一个新兴的智能交通工程学科。