Intelligent real-time traffic analysis from live video streams

实时视频流的智能实时流量分析

• 批准号: 561391-2020
• 负责人: Elgazzar, Khalid
• 金额: $ 1.97万
• 依托单位: University of Ontario Institute of Technology
• 依托单位国家: 加拿大
• 项目类别: Alliance Grants
• 财政年份: 2021
• 资助国家: 加拿大
• 起止时间: 2021-01-01 至 2022-12-31
• 项目状态: 已结题
• 来源: https://www.nserc-crsng.gc.ca/ase-oro/Details-Detailles_eng.asp?id=721994
• 关键词: Intelligent; real; time; traffic; analysis

This project will develop novel and highly accurate AI-powered techniques for real-time traffic analysis from HD live video streams (>25 FPS) to enable smart and intelligent service provisioning to customers. The project will focus on developing algorithms to detect and classify vehicles on the road at different lanes, calculate average road speeds at high accuracy, shareholder recognition (vehicle class, motorcycle, cyclist, pedestrian, etc.). These AI services will be developed using custom machine learning models built specifically for these services and trained on the state-of-the-art datasets at different image quality and varying resolutions to maximize model accuracy in real-time fitting.Ontario Tech will leverage the research and development outcomes of this project to help North Line attain a minimum viable product of the proposed technology and bring another great feature to the intelligent transportation market in Canada and beyond. This will also help North Line to expand their business nationally and internationally with a unique technology that provides an edge over competitors, combining the excellence of intelligent sensing, real-time analytics from video streams and the power of AI to enable predictive analytics. The developed AI-powered traffic analysis will push the boundaries of smart and intelligent transportation to a new level with a "Made in Canada" integrated solution.

该项目将开发新颖和高精度的人工智能支持的技术，用于从高清直播视频流(&gt；25FPS)进行实时交通分析，以实现向客户提供智能和智能服务。该项目将专注于开发算法，以检测和分类不同车道上的车辆，高精度计算平均道路速度，股东识别(车辆类别，摩托车，骑自行车，行人等)。这些人工智能服务将使用专门为这些服务构建的定制机器学习模型来开发，并在不同图像质量和不同分辨率的最先进数据集上进行培训，以最大限度地提高实时拟合的模型准确性。安大略省理工学院将利用该项目的研究和开发成果，帮助北线实现拟议技术的最低可行产品，并为加拿大及其他地区的智能交通市场带来另一个伟大的功能。这也将有助于北线通过一种独特的技术在国内和国际上扩大他们的业务，该技术提供了相对于竞争对手的优势，结合了智能传感、视频流的实时分析和人工智能实现预测分析的能力的卓越之处。开发的人工智能支持的交通分析将把智能和智能交通的边界推向一个新的水平，拥有一种“加拿大制造”的集成解决方案。