Visual Expert Human Factors: Determining Visibility

Marc Green

"Contrast detection is the basic task from which all other visual behaviors are derived."
- Illuminating Engineering Handbook

Many cases hinge on the issue of visibility: was the pedestrian visible, should the construction worker have seen the signal, was the label legible, etc? Seeing is a function of two processes: sensation and perception. "Sensation" refers to the issue of whether the light from the object was sufficient to reach the physiological level needed for detection. Visibility is a measure of the sensation level caused by a target. The most important single visual attribute of an object's visibility is its contrast - the difference in brightness (or sometimes color or texture) between an object and its background. By "perception," I mean whether the viewer's attention, memory and other cognitive functions were operating on the sensory input. For example, an object might be visible but still not be seen because the viewer's attention was not engaged. Some term the distinction between sensation and perception as "visibility vs. conspicuity."

In this article, I address the methods used to measure contrast and to determine whether an object should have been detectable. Contrast is far more important than absolute light level, but it is not unusual for an "expert" to err by merely measuring light levels without determining contrast. If an object's contrast is too low to be seen, then other visual factors are irrelevant. Below, I outline the methods for determining visibility. See Green, et al. (2008) for a more detailed analysis.

There are a few basic ways to determine contrast, depending on circumstances:

You can accurately reconstruct the Scene

If you can recreate the accident - same objects, time of day, weather, etc, contrast can be determined with high accuracy. The general method is outlined below.

Light from some source, such as the sun, lamp, etc., falls on a surface. The surface reflects some of the light back to the viewer's eye, where it can be used for seeing. The amount of light falling on the surface is the illumination and is measured in units of lux. Normally, you measure illumination by putting a light sensor up against the surface and reading an "illuminance photometer". If there are many light sources, then the meter reads the total illumination coming from all sources.

However, illumination does not specify the amount of light reaching the eye. As stated succinctly in the National Transportation Library Document 97097:

"Illuminance criteria have been proven to be inadequate predictors of the effectiveness of lighting systems. Although the visibility of targets is typically directly proportional to illuminance (all other variables held constant), there are too many intervening variables that determine the visual stimulus and the efficiency with which that stimulus is processed by the visual system."

In short, merely measuring illumination, as is often done, is not enough. You need to measure the amount of light reflected to the eye from 1) the object and 2) its background. This requires measurement of a different quantity, "luminance," which gives the amount of light actually available for vision. The most straightforward measuring procedure is to use a specialized device called a "luminance photometer." Place it at the position of the viewer's eyes, aim at the object or surface which you wish to measure and read the meter in units called candela per meter squared (cd/m²). Of course, it is important to read the background, since it is just as important in determining visibility as the target object.

For outdoor measurements, the direct measurement is best performed at the same time, on the same date and in the same weather conditions as the accident. Naturally, the measurements will have to occur one or more years after the event. There are slight year to year lighting variations due to subtle changes in the positions of sun and moon. This adds some imprecision in extrapolating back to the accident scene, but if weather conditions are similar, then the variability is too small to affect any conclusions about visibility.

If this ideal condition cannot be met for some reason, then there is a series of possible alternatives, each introducing additional imprecision. Depending on the exact circumstances, however, these technique may be sufficient:

Use substitute objects

In some cases, the actual objects (vehicle, person's clothes, etc) may not be available. In this case, the best strategy is to find some substitutes which resemble the originals as closely as possible. Determine contrast as described above.

Estimate reflectance

Sometimes it is possible to determine an object's reflectivity from a set of tables. In this case, the illumination measurement can be used to estimate the luminance of the light reaching the eye.

Estimate illuminance and reflectance

The roughest estimate occurs when no measurements can be made at the accident site under the accident conditions. In this case, the only alternative is to use tables which provide illuminance and reflectance values. For example, tables can provide the average illumination at different times of day, twilight and so on. There are also standard tables for the reflectance of grass, asphalt, etc.

Conclusion

The most accurate contrast values will be obtained by making direct measurements. Using tables is less accurate since it depends on average, not actual, values. Moreover, it doesn't tell you exactly what the viewer saw from his eye position.

It is important to understand that measuring contrast is only the starting point of a visibility analysis. The next step is to determine whether the value was high enough for a viewer to see the object in question. This is a very tricky question to answer because perception depends on many interlocking variables. This why only someone highly trained in human vision can make a scientifically reasonable estimate: Here are just a few of the factors which must be considered:

Adapting luminance
Object size in visual angle
Object distance
Object orientation
Object shape
Visual field location
Duration
Motion/Flicker
Masking and camouflage
Color
Glare
Viewer drugs or alcohol
Viewer age
Viewer adaptation State
Viewer arousal level
Viewer uncertainty
Viewer expectation
Viewer optical correction
Viewer eye disease