|
The contrast sensitivity model given in Chapter 3 was restricted to spatial contrast sensitivity. In this chapter the model will be extended to the temporal domain so that a spatiotemporal model is obtained that can also be used for purely temporal luminance variations. Temporal contrast sensitivity has already been intensively studied in the fifties by de Lange (1952, 1954, 1957, 1958a, 1958b, 1961) and in the sixties and seventies by Kelly (1960, 1961, 1971, 1972, 1979) and by Roufs (1972a, 1972b, 1973, 1974a, 1974b, 1974c). Kelly (1960) proposed to combine spatial and temporal contrast sensitivity measurements by using spatiotemporal stimuli to get more insight into spatiotemporal interactions. The two-dimensional contrast sensitivity function obtained with these types of stimuli is not simply the product of a spatial and a temporal response, but shows a much more complicated behavior (Robson, 1966; van Nes et al., 1967; Kelly, 1971, 1972, 1979; Koenderink & van Doom, 1979). To explain this behavior, Kulikowski & Tolhurst (1973) supposed the existence of a sustained and a transient channel in the human visual system, analogous to supposed spatial frequency channels in the spatial domain.
For the temporal contrast sensitivity, it may be assumed that it is determined by internal noise in the same way as for spatial contrast sensitivity. Besides a spatial character, noise generally also has a temporal character. For the spatiotemporal model that will be given here, no different channels will be assumed for a sustained and a transient response, but the complicated spatiotemporal behavior of the visual system will be explained in a much simpler way. In this model, it is only assumed that the inhibition signal undergoes, in the inhibition process, besides spatial filtering, also temporal filtering before being subtracted from the photo-receptor signal.
|
