Enhancing the mirror image

增强镜像

• 批准号: 464850937
• 负责人: Professor Dr. Heiko Hecht
• 金额: --
• 依托单位: Abteilung Allgemeine Experimentelle Psychologie
• 依托单位国家: 德国
• 项目类别: Research Grants
• 资助国家: 德国
• 项目状态: 未结题
• 来源: https://gepris.dfg.de/gepris/projekt/464850937?language=en
• 关键词: Enhancing; mirror; image

We have not evolved to base our actions on mirror images. Our intuitions about mirror objects are often erroneous. However, in many situations we nonetheless rely critically on mirror images. For instance, when driving, the rear-view mirror plays an important role in time-critical actions. We have become accustomed to convex mirrors in cars, but modern camera-monitor systems are about to replace mirrors and pose new challenges. Little is known about the advantages and potential disruptions of perception when using such systems. On the one hand, camera-monitor systems can help us overcome errors associated with mirror use, provided they can be optimized for the human user. On the other hand, if they are not user-optimized, they could make perception-based actions even harder or outright dangerous. It is our goal to investigate the conditions that facilitate use and minimize human error when acting upon the visual information provided by camera-monitor systems. One cause for potential danger is that the eye point in these systems differs from the eye-point to which we have become accustomed when driving with classical mirrors. The perspective change and the eye-point dissociation may distort the perceived ego-position with respect to the outside world. We propose a number of experiments to investigate the absolute and relative placements of camera and monitor with regard to their effects on eye-movement behavior, time-to-contact estimation, lane change decisions, and hazard detection. We also propose to explore different methods to enhance the rear-view image in ways that are not possible with traditional mirrors. In the monitor, the image can be processed to enhance relevant objects or to remove irrelevant clutter. The proposed research is designed to provide theoretical insights into the processing of visual information with unusual vantage points. And it is designed to assess different methods of mirror image enhancement. From these insights, we will derive practical recommendations for the placement of camera and monitor as well as image-processing. In order to conduct the proposed experiments in our laboratory, we would need a mobile eye-tracker and a simple custom-made driving-simulator.

我们还没有进化到将我们的行为建立在镜像的基础上。我们对镜像物体的直觉常常是错误的。然而，在许多情况下，我们仍然严重依赖镜像。例如，在驾驶时，后视镜在时间紧迫的行动中发挥着重要作用。我们已经习惯了汽车中的凸面镜，但现代摄像监控系统即将取代镜子，并提出新的挑战。在使用这种系统时，人们对感知的优势和潜在干扰知之甚少。一方面，摄像机监视器系统可以帮助我们克服与镜子使用相关的错误，前提是它们可以针对人类用户进行优化。另一方面，如果它们不是用户优化的，它们可能会使基于感知的操作变得更加困难或完全危险。我们的目标是调查的条件，促进使用和最大限度地减少人为错误时，对摄像机监视器系统提供的视觉信息。潜在危险的一个原因是，这些系统中的视点不同于我们在驾驶时使用经典镜子时所习惯的视点。视角变化和视点分离可能会扭曲相对于外部世界的自我位置感知。我们提出了一些实验，调查的绝对和相对位置的摄像头和监视器的眼睛运动行为，时间接触估计，车道变换的决定，和危险检测方面的影响。我们还建议探索不同的方法来增强后视镜图像的方式是不可能与传统的镜子。在监视器中，可以处理图像以增强相关对象或去除不相关的杂波。拟议的研究旨在为具有不寻常Vantage的视觉信息处理提供理论见解。它的目的是评估不同的镜像增强方法。从这些见解中，我们将获得摄像机和监视器的位置以及图像处理的实用建议。为了在我们的实验室进行拟议的实验，我们需要一个移动的眼动跟踪器和一个简单的定制驾驶模拟器。