Thresholds for detecting the direction of motion of drifting (8-Hz) vertical gratings [of spatial frequencies 0.1, 1.0, and 10.0 cycles per degree (c/deg)] were measured in the presence of masks that varied in both spatial frequency and orientation. Masks with different temporal properties were used. The specificity of masking was also measured for a stationary test grating of spatial frequency 3.0 c/deg. After suitable scaling and transformation, the masking data gave an estimate of the two-dimensional spatial-frequency tuning surface of cortical detector units in human vision. With the assumption of small-signal linearity and zero phase, the tuning surfaces were inverse Fourier transformed to give an indication of the size and structure of the psychophysical receptive fields of detector units. The results obtained with drifting test gratings and jittering (random phase) mask gratings indicate that motion-detector receptive fields increase in size (in cycles) with increasing spatial frequency but, at all spatial scales, have a length–width ratio of 1. These results are in close agreement with the summation results reported in J. Opt. Soc. Am. A 8, 1330 ( 1991). Using the same jittering mask stimuli and stationary test gratings, we confirm reports by Daugman [ Vision Res. 24, 891 ( 1984)] and Harvey and Doan [ J. Opt. Soc. Am. A 7, 116 ( 1990)] that motion-independent units have elongated receptive fields with a length–width ratio near 1.8. We conclude that the receptive fields of motion-dependent and -independent mechanisms in human vision are fundamentally different. The possibility that the orientation selectivity of a motion unit is sharpened by its selectivity for direction of motion is discussed.
© 1991 Optical Society of AmericaFull Article | PDF Article