A cloud based multimodal biometrics framework for road safety

基于云的道路安全多模态生物识别框架

• 批准号: 486201-2015
• 负责人: Guan, Ling
• 金额: $ 7.16万
• 依托单位: Ryerson University
• 依托单位国家: 加拿大
• 项目类别: Collaborative Research and Development Grants
• 财政年份: 2017
• 资助国家: 加拿大
• 起止时间: 2017-01-01 至 2018-12-31
• 项目状态: 已结题
• 来源: https://www.nserc-crsng.gc.ca/ase-oro/Details-Detailles_eng.asp?id=618845
• 关键词: cloud; based; multimodal; biometrics; framework

Alcohol-impaired driving caused 55% of fatal crashes among 16 to 25 years old in Canada in 2009, which is the leading cause of death in this age group. Therefore, ensuring that a driver is not impaired by alcohol is critical to improve road safety in Canada. Drowsy driving is another serious problem on Canada's roadways. Research by the Ontario Ministry of Transportation indicates that about 26% of all fatal and injury crashes are related to drowsiness while driving. Collaborating with Alcohol Countermeasure Systems (ACS) Corp. , we will investigate a novel cloud-based multimodal biometrics framework to address the issues of continuous authentication and drowsiness detection, aiming to enhance road safety in Canada. Technically, we will apply computer vision, pattern recognition, and information theory to address the following research problems: 1) cloud based multimodal biometrics extraction: to extract the driver's multimodal biometrics of face, facial expression, eye movement, electrocardiogram (ECG) signal and Heart Rate Variability (HRV) signal at the cloud server; 2) continuous driver authentication: to continuously identify if the person driving the vehicle is the person who has started the engine based on face, facial expression, and ECG signal; 3) drowsiness detection: to accurately detect the drowsiness status based on eye movement, yawning and HRV signal, and alert the driver to prevent from crashes. The research carried out in this project will result in the development of new road safety devices that will accurately and automatically authenticate the driver or alert a drowsy driver by incorporating multiple biometric traits. The outcome of the project will assist ACS in achieving further technical success and sales of interlock systems and drowsiness detection devices at home and abroad. Furthermore, the success of this project will promote Canada's R&D capability in multimedia and information technology internationally, and strengthen the competitiveness of Canada's products in the world market.

2009年，在加拿大16至25岁的人群中，55%的致命车祸是酒后驾车造成的，这是这一年龄段的主要死亡原因。因此，确保司机不受酒精损害是提高加拿大道路安全的关键。昏昏欲睡的驾驶是加拿大道路上的另一个严重问题。安大略省交通部的研究表明，在所有致命和受伤的车祸中，约26%与驾驶时昏昏欲睡有关。我们将与酒精对抗系统公司(ACS)合作，研究一种新型的基于云的多模式生物识别框架，以解决连续身份验证和嗜睡检测问题，旨在提高加拿大的道路安全。在技术上，我们将应用计算机视觉、模式识别和信息理论来解决以下研究问题：1)基于云的多模式生物特征提取：在云端服务器提取驾驶员的面部、面部表情、眼动、心电(CG)信号和心率变异性(HRV)信号的多模生物特征；2)连续驾驶员身份验证：根据人脸、面部表情和ECG信号，连续识别驾驶车辆的人是否为启动发动机的人；3)嗜睡检测：基于眼动、打哈欠和HRV信号，准确检测驾驶员的困倦状态，并提醒驾驶员防止撞车。在该项目中进行的研究将导致开发新的道路安全设备，该设备将通过结合多种生物特征来准确和自动地验证司机身份或向昏昏欲睡的司机发出警报。该项目的成果将帮助ACS取得进一步的技术成功，并在国内外销售联锁系统和昏昏欲睡检测设备。此外，该项目的成功将促进加拿大在多媒体和信息技术方面的国际研发能力，并增强加拿大产品在世界市场上的竞争力。