Maximal Acceptable Ghost Images for Designing a Legible Windshield-Type Vehicle Head-Up Display

Abstract

Windshield-type vehicle head-up displays (HUDs) are increasingly popular and stepping forward augmented reality; however, the windshield causes an annoying problem- ghost images. By now, the maximal extent of ghost images has not been determined for an HUD with an acceptable legibility. In this paper, to find the quantitative criterion of ghost images regarding subjective perceptions, we first design an HUD using a rotatable aspheric reflector and a wedge-glass windshield. Optical design of the HUD and experimental results of high-quality images are discussed. Next, eight different disparity angles between the primary and ghost images are controllably generated by rotating the aspheric reflector. Based on this HUD platform, human factor experiments utilizing a simulative automotive driving environment are conducted, in which the eight disparity angles are provided to 12 subjects under three ambient contrast ratios (ACRs). The human factor experiments demonstrate that the maximal disparity angles for an acceptable legibility under the ACRs of 3.90, 2.49, and 1.80 are 0.006 degrees , 0.017 degrees, and 0.024 degrees, respectively. These maximal acceptable disparity angles are important references for quantitatively and efficiently evaluating the optomechanical system of an HUD. In addition, they can help to make vehicle laws and regulations for HUDs.