Pham Xuan Rebecca, Xiong Yuxin, Brietzke Adrian, Marker Stefanie
Volkswagen AG Group Innovation, Letterbox 011/17773, 38436, Wolfsburg, Germany; Technical University Berlin, Naturalistic Driving Observation for Energetic Optimization and Accident Avoidance, Institute of Land and Sea Transport Systems, Gustav-Meyer-Allee 25, 13355, Berlin, Germany.
Volkswagen AG Group Innovation, Letterbox 011/17773, 38436, Wolfsburg, Germany; Institute of Mechatronic Systems, Gottfried Wilhelm Leibniz University Hannover, 30167, Hannover, Germany.
J Therm Biol. 2021 Feb;96:102806. doi: 10.1016/j.jtherbio.2020.102806. Epub 2020 Dec 9.
Motion Sickness is associated with a variety of symptoms, which differ in occurrence rate and intensity between individuals. In order to research the cause of car sickness and develop countermeasures, it is important to determine symptoms and their severity objectively. A tool for this purpose could be the assessment of physiological reactions due to motion sickness. This paper describes and discusses a methodology to identify changes in facial skin temperatures in a real-driving study. Common techniques had to be adjusted in order to meet the requirements given by the challenges of in-car-recording. The examined data was generated in a previous study, which was designed to research motion sickness in a driving environment. A pre-processing technique had to be developed to magnify features on the face and subsequently improve the tracking in thermal imagery. After the pre-processing, regions of interest (ROI) were manually marked and tracked in thermal images. The thereby assessed facial skin temperatures were compared to tympanic temperatures. Derived temperatures from the forehead as well as from the 20 hottest pixels within the face indicated a better tracking, while the nose tip was more affected by detection errors. The correlation of the three features with the tympanic temperature showed remarkable differences between a baseline measurement and the actual driving. Less than 10% of the data derived during the driving and up to 30% of the data during the baseline measurement correlated highly. It is concluded that detecting changes in facial skin temperature using thermal infrared imaging in a moving car is challenging and results are hardly comparable to tympanic temperatures. Future research should aim at the different influencing factors of skin and tympanic temperature, while enhancing tracking or detection of ROI could be achieved by reducing the passengers' movements or choosing the target area more carefully.
晕动病与多种症状相关，这些症状在个体之间的发生率和强度有所不同。为了研究晕车的原因并制定应对措施，客观地确定症状及其严重程度很重要。为此目的的一种工具可以是评估晕动病引起的生理反应。本文描述并讨论了一种在实际驾驶研究中识别面部皮肤温度变化的方法。必须调整常用技术，以满足车内记录挑战所提出的要求。所检查的数据是在先前一项旨在研究驾驶环境中晕动病的研究中生成的。必须开发一种预处理技术来放大面部特征，随后改善热成像中的跟踪。预处理后，在热图像中手动标记并跟踪感兴趣区域（ROI）。将由此评估的面部皮肤温度与鼓膜温度进行比较。从前额以及面部内20个最热像素得出的温度显示出更好的跟踪效果，而鼻尖受检测误差的影响更大。这三个特征与鼓膜温度的相关性在基线测量和实际驾驶之间显示出显著差异。驾驶期间得出的数据不到10%，基线测量期间的数据高达30%高度相关。得出的结论是，在行驶的汽车中使用热红外成像检测面部皮肤温度变化具有挑战性，结果很难与鼓膜温度相比较。未来的研究应针对皮肤和鼓膜温度的不同影响因素，同时通过减少乘客的动作或更谨慎地选择目标区域来实现增强ROI的跟踪或检测。
