Effects of real-world backgrounds on user interface color naming and matching in automotive AR HUDs

This study investigated which optical see-through AR user interface colors are most robust across different real-world driving backgrounds. We consider AR colors to be robust if they (1) support effective visual search task performance with low variability, and (2) retain their color semantics independent of background. Twelve experienced drivers performed two types of visual tasks; a structured (text) and a semi-structured (symbol) visual search task; using a head-up display. Drivers experienced eight user interface colors against three backgrounds (brick, grass, pavement) with repetition. We measured user task performance (response time & errors) as well as color recognition (color matching and naming). Our results showed no significant differences in response time across backgrounds, but identified significant differences in response time between colors within each background. For both text and symbol tasks, participants consistently matched perceived AR user interface colors to World Color Survey stimulus palette locations that were brighter as compared to the source AR color locations. Blue, green, and yellow AR colors showed very little hue shift, and participants verbally identified these colors by name with a high degree of accuracy as compared to other colors. Our results are generally consistent with other findings, in that we observed washout effects when using AR technology. Despite the washout, we found blue, green, and yellow AR colors to be fairly robust across all background and task conditions.