Implications for contact analogue HUDs in the automotive application arise from this finding that the higher salience image is more readily perceived. Design exploitation of this would favor a high salience HUD image that would replace rather than supplement missing information from the external road scene.
This solution is practical under impoverished viewing conditions where a relatively bright high contrast HUD image can replace missing information. However, to accommodate situations where the background illumination is relatively high (e.g., in daytime fog, or passing through intermittent street lighting), control over the luminance of the HUD image would be required. Mulvanny ( 1993) suggested automatic brightness level control as a preferable solution.
Finally, the issue of HUD focal plane distance needs to be addressed in a less ambitious design, with a greater number of observations for each focal plane setting.
The presentation of stimulus materials in the present study was artificially manipulated so that images appeared in either the HUD or as elements in the background. This replicated two possible scenarios that might occur in a real application: An object is imaged in the HUD but not visible to the naked eye, or an object is visible but either not picked up by the VES sensor, or not imaged in the HUD in a recognizable or useful format. A third, and potentially most common scenario, is that an object is both imaged and simultaneously visible or partly visible to the naked eye.
The present study found that HUD target detection was superior to screen detection even though only a relatively slight salience discrepancy existed between the HUD image and screen image (in favor of the HUD image). This would imply that in the contest between a clear HUD image and a partial naked-eye view, as implied in the third scenario, the HUD image would emerge superior. For automotive contact analogue HUDs it implies that perfectly matching the HUD image to the external scene is of secondary importance to providing a high quality HUD image with higher information content than the unenhanced forward view. However, care should be observed in ensuring that any scenario 2 type objects are adequately accounted for. The issue of integrating the central HUD image with the unenhanced peripheral external scene must also be addressed.
This chapter forms part of the output of the vision enhancement strand of the BRIMMI Project, supported by UK Department of Trade and Industry, within the PROMETHEUS Program. The UK industrial partners