|This report is an archived publication and may contain dated technical, contact, and link information|
Publication Number: FHWA-HRT-13-018
Date: April 2013
The effects of retroreflective properties on the chromaticity and luminance of reflected light under natural daylight and simulated daylight (D65) conditions were explored. Chromaticity and luminance values were also compared with the daytime appearance of the materials as judged by a group of human observers. Understanding the effects of retroreflective properties on physical measurements and perceptual ratings is important to FHWA in defining the size and shape of the color areas used to specify colors for traffic control signs. The color of traffic control signs is an important part of the coding system for conveying consistent information to drivers. On one hand, if the color areas are too large, some of them may come close to functionally overlapping, especially when the variability of human color judgments is taken into account, resulting in possible color confusion and misidentification by drivers. On the other hand, if the color areas are too small, manufacturers can be unnecessarily restricted in the types of materials and processes used for the production of street and highway signs, possibly reducing alternatives and increasing costs. A balance must be struck between these and other competing concerns in defining the size and shape of the FHWA color areas. Information on how drivers perceive the colors of traffic control signs is important in this regard. Equally important is information on how to objectively measure these colors so manufacturers can efficiently and reliably meet FHWA color requirements.
In the present experiment, a white diffuse reflector coupled with appropriate color and neutral density filters was employed as a reference standard. CIELAB plots representing laboratory physical color measurements of the reference standard show well-defined color areas that are arranged in an orderly manner in the color space. Based on the sizes and orientations of the color areas in terms of chromaticity coordinates (see figure 2), the only major unexpected outcome in the physical measurements is the relatively small size of the red color area. The laboratory physical measurements show that retroreflective materials tend to reduce the saturation of all of the colors relative to the diffuse reference. Based solely on physical measurements, the reduced size and closer proximity of the red, orange, and yellow color areas for retroreflective materials might indicate a higher probability of color confusion between certain orange and yellow signs and between certain orange and red signs.
Perceptual color measurements, expressed in terms of UADs, also show reduced saturation for retroreflective materials relative to the diffuse reference. This might be expected, since retroreflective properties reduced the luminance of the samples. This outcome confirms earlier results found by Davis and Miller.(9) The present experiment revealed an additional consideration; for retroreflective materials, there was significant insensitivity in the perception of differences in yellow saturation for the orange and yellow color boxes. This compression along the yellow dimension indicates that the human observers were relatively insensitive to chromaticity changes along the yellow dimension for retroreflective materials. This relative human insensitivity could have implications for the weathering or fading of yellow traffic control signs. Yellow signs usually fade toward white, representing a reduction in saturation along the yellow color dimension. While not desirable, the fading of yellow sign materials may not be as noticeable to drivers as might be anticipated from instrument measurements. In any case, constant hue lines for perceptual measurements show a clearer separation between orange and yellow and between orange and red than the instrument measurements indicated. To a certain degree, human observers seem to be able to discriminate between different colors in this region of the spectrum better than might be inferred from physical measurements. The results for the white, green, and blue color areas show distinct color separations for all retroreflective sign materials.
Overall, the results of the experiment revealed that, under daytime lighting conditions, drivers with normal color vision should not experience confusion discriminating hue among the six FHWA color boxes that were evaluated. On average, across all retroreflective sheeting types, the separation among the perceptual color areas appeared sufficient to discriminate between any two colors of the six tested. However, when individual sheeting types were considered, there were two cases where the differences between combinations of color and sheeting type were not statistically significant. One of these cases was between orange and yellow, and the other case was between orange and red. While this result may not directly imply that drivers would have difficulty discriminating between traffic control signs representing these combinations of colors and sheeting types, the result may indicate potential for color confusion at these two color boundaries.
It is likely that the participants in the experiment could have identified differences between these two pairs of sign colors and materials if the participants had viewed the pairs simultaneously side-by-side. However, recognition of traffic control signs should not require simultaneous presentation but should be unambiguous at a distance at which drivers may acquire textual information. In this sense, the method of direct perceptual hue and saturation rating employed in the experiment is closer to the absolute identification of the colors of individual signs required of most driving tasks. The results of the experiment suggest that caution should be applied in any future deliberations to alter the shapes and sizes of the red, orange, and yellow color areas specified for traffic control signs. The results also indicate that there is no pressing need to modify the FHWA color areas at the present time, but improvements may be possible so as to enhance separation of those color boundaries that are more difficult to discriminate.
As concerns perceptual measures of brightness, the diffuse white reflector with a color filter was always judged the brightest for any given color. Retroreflective properties tended to produce darker colors relative to this reference condition. Instrument measurements of luminance confirmed this trend. This overall reduction in luminance that accompanies retroreflective properties may partially explain the loss of saturation of these colors when compared to the diffuse reference condition.
The experiment employed the basic method of direct color scaling developed by Abramov et al.(5) This method does not require precise equipment and controlled viewing conditions and lends itself readily to field applications. The present experiment represents a validation of the basic technique using actual samples of highway traffic sign materials under variable field viewing conditions (differing degrees of cloud cover, passing clouds, differing sun angles, etc.). The technique performed well under such field conditions, producing color areas that were consistent and orderly in the UAD color space.
Gordon et al. expanded on this technique, and the present experiment incorporated some of the enhancements.(6) However, although they noted that about 5 percent of the research participants did not use the percentage scales appropriately, Gordon et al. claimed that research participants did not need special training to use the method. To the contrary, the present researchers found that substantial training was necessary beyond simply administering the formal verbal instructions (see appendix A). A partial explanation of this discrepancy may lie in the differences between the two samples of research participants. Most of the participants in the original studies were somewhat experienced, having participated in several experiments, and rather homogenous, coming from an academic environment.(6) The present experiment employed a broader sample of naïve research participants recruited from the general driving public.
With the caveat that additional training may be necessary, the present experiment may be regarded as a validation of the direct color scaling technique applied to a broader sample of research participants and testing conditions. The results of the experiment indicate that the basic color scaling technique can be employed successfully in an outdoor environment to provide technical answers to practical problems in highway and traffic engineering. The problems investigated in the experiment revolve around determining the size and shape of the FHWA color areas used to specify traffic control signs. Future implementations of this technique might be used to answer specific engineering questions about particular combinations of color and retroreflective properties. The experiment also added an analogous brightness scaling technique, which proved equally successful. The researchers recommend that these methods be considered in future research to address questions concerning driver perception of the color properties of traffic control signs.
Topics: research, safety
Keywords: research, safety, Daytime color appearance, Retroreflective signs, Hue scaling, Saturation scaling, Brightness scaling, Color boxes, Chromaticity, Luminance
TRT Terms: research, Safety and security, Safety, Transportation safety