This chapter describes an identification task and associated test methodology. Image quality is quantified by the ability to identify objects in a test set. Chapter 7 describes an imager resolution metric that predicts the ability of observers to perform the identification task. The resolution metric predicts the impact of imager blur, noise, and sampling on target identification.

This chapter also discusses the use of diverse target sets to establish operational utility. In a target acquisition sense, the focus is on the average probability of identifying many objects in the scene. The ability to identify a specific target is not addressed. The focus on diverse target sets greatly simplifies the analysis presented in Chapter 7. Also, if the goal of imager analysis is to select a good imager for a particular type of task, target sets provide a better way to quantify performance than specific target analysis.

Using standard target sets to check visual acuity is quite common. In Fig. 6.1, the patient is reading an eye chart in the optometrist's office. The doctor tells her to read the smallest line she can. Many others have tried to read this or similar charts. The doctor knows the letter size that represents average vision. Reading the chart provides a quantitative method of comparing vision among observers.

Using high-contrast letters is not the best way to check vision, however. A scene consists of many luminance levels. The eye achieves an integrated view of objects by connecting lines and surfaces. These lines and surfaces do not share a particular brightness throughout their extent. For example, the background immediately behind a target might not be uniform, and yet the eye sees the target silhouette. Perspective is gained from converging lines even if they vary in luminance with increasing range. Slight changes in hue or texture can provide an excellent cue as to the distance and orientation of an object. Acute vision requires not only the ability to discriminate small details that happen to have good contrast, but also the ability to discriminate small differences in shades of gray.