Poirson A B, Wandell B A
Department of Psychology, Stanford University, California 94305.
J Opt Soc Am A. 1990 Apr;7(4):776-82. doi: 10.1364/josaa.7.000776.
When an observer's ability to discriminate colored objects is estimated from the variability in color matches, the observer inspects adjacent visual fields carefully and makes considered judgments. Color discrimination does not always take place under such viewing conditions. When color video displays are used in time-critical applications (e.g., head-up displays, video control panels), the observer must discriminate among briefly presented targets seen within a complex spatial scene. We compare color-discrimination thresholds by using two tasks. In one task the observer makes color matches between two halves of a continuously displayed bipartite field. In a second task the observer detects a color target in a set of briefly presented objects. The data from both tasks are well summarized by ellipsoidal isosensitivity contours. The fitted ellipsoids differ both in their size, which indicates an absolute sensitivity difference, and orientation, which indicates a relative sensitivity difference.
当根据颜色匹配的变异性来估计观察者辨别有色物体的能力时，观察者会仔细检查相邻视野并做出审慎判断。颜色辨别并非总是在这样的观察条件下进行。当彩色视频显示器用于时间紧迫的应用（例如平视显示器、视频控制面板）时，观察者必须在复杂空间场景中快速呈现的目标之间进行辨别。我们通过使用两项任务来比较颜色辨别阈值。在一项任务中，观察者在连续显示的二分视野的两半之间进行颜色匹配。在第二项任务中，观察者在一组快速呈现的物体中检测颜色目标。来自这两项任务的数据都可以通过椭球形等灵敏度轮廓很好地总结。拟合的椭球体在大小（表示绝对灵敏度差异）和方向（表示相对灵敏度差异）上都有所不同。
