What we taste is affected by what we see, and that includes the colour, opacity, and shape of the food we consume. We report two experiments designed to investigate how the standard deviation (SD) of the luminance distribution of food images influences the perceived visual texture and the taste/flavour experience by using the latest Augmented Reality (AR) technology. We developed a novel AR system capable of modifying the luminance distribution of foods in real-time using dynamic image processing for simulating actual eating situations. Importantly, this form of dynamic image manipulation does not change the colour on the food (which has been studied extensively previously). Instead, the approach outlined here was used to change the SD of the luminance distribution of the food while keeping the chromaticity, the average luminance, and the skewness constant. We investigated the effects of changing the luminance SD distribution of Baumkuchen (a German baked cake) and tomato ketchup on visual perception, flavour expectations, and the ensuing taste experience. Participants looked at a piece of Baumkuchen (Experiment 1) or a spoonful of tomato ketchup (Experiment 2) having different luminance distributions and evaluated the taste on sampling the food. Manipulating the SD of the luminance distribution affected not only the expected taste/flavour of the food (e.g. expected moistness, wateriness and deliciousness), but also the actual taste properties on sampling the food itself. The novel food modification method and system outlined here can therefore potentially be used to control the taste/flavour of different foods crossmodally by means of modifying their appearance properties (specifically the SD of the luminance distribution while keeping other aspects of image statistics constant), and can do so in real time, without the need for food markers.

我们品尝到的东西会受到我们所见所闻的影响，其中包括我们食用的食物的颜色、不透明度和形状。我们报告了两个实验，旨在通过使用最新的增强现实 (AR) 技术研究食物图像亮度分布的标准偏差 (SD) 如何影响感知的视觉纹理和味道/风味体验。我们开发了一种新颖的 AR 系统，能够使用动态图像处理来实时修改食物的亮度分布，以模拟实际饮食情况。重要的是，这种动态图像处理形式不会改变食物的颜色（之前已对此进行了广泛研究）。相反，这里概述的方法用于改变食物亮度分布的 SD，同时保持色度、平均亮度、和偏度常数。我们研究了改变 Baumkuchen（一种德国烘焙蛋糕）和番茄酱的亮度 SD 分布对视觉感知、风味预期和随后的味觉体验的影响。参与者看着一块年轮蛋糕（实验 1）或一勺番茄酱（实验 2）具有不同的亮度分布，并评估对食物取样的味道。操纵亮度分布的 SD 不仅会影响食物的预期味道/风味（例如预期的湿度、水份和美味），还会影响对食物本身进行采样时的实际味道特性。