The spatial limit for motion detection in noise depends on element size, not on spatial frequency

Area of Science:

• Visual perception
• Neuroscience
• Computational vision

Background:

• Perceiving motion direction from random patterns is crucial for navigation and interaction.
• The displacement limit (Dmax) for coherent motion perception in random noise patterns is not fully understood.

Purpose of the Study:

• To investigate the factors influencing the maximum displacement limit (Dmax) for motion perception in random noise patterns.
• To test a model of low-level motion detection based on pattern statistics and spatial filtering.

Main Methods:

• Measured Dmax using a two-frame sequence of random spatial noise patterns.
• Manipulated pattern element size and contrast.
• Applied spatial band-pass filtering to remove low spatial frequencies.

Main Results:

• Dmax was dependent on the size of pattern elements, even after spatial filtering.
• Contrast significantly affected Dmax for filtered patterns but had less impact on unaltered patterns.
• The findings suggest Dmax is determined by pattern element separation after low-pass filtering.

Conclusions:

• Low-level motion detection is influenced by the statistical properties of the visual stimulus, specifically element separation.
• The maximum detectable spatial displacement is primarily governed by pattern statistics, not detector size in the visual system.