Phenomenology of Motion

In this section I will present an assortment of phenomenological motion effects and observations.

Speed and the Retinal Stimulus

Consider an object moving smoothly from point A to point B as in the figure above. In both (a) and (b) the object is moving at the same velocity however in (a), because the object is closer to the eye, the velocity of the object's projection on the retina is greater than the velocity of its projection in (b). The perceived speed does not correspond with the speed of the image on the retina. We must take into account the distance of the object to accurately judge speed.

Velocity Transposition

The context (the size of the objects and the size of the reference frame through which they move) influences the perception of velocity. “A cat in a large cage must move much faster than a mouse in a small cage if they are to appear to move at the same speed.” Hopefully, by looking at the movie below you will get the idea. The cat on the right is moving at the speed of the mouse in the middle, yet the cat on the left, moving twice as fast, appears to move at the same speed as the mouse.

Five Ways to make a spot of light move (and it only really moves in one)

I should make flashy examples of these, but I am running out of time. Please go through this chart. Some comments will follow below.

Comments:

In apparent motion, the distance between the stimuli, as well as the interstimulus interval will have an effect on the perceived motion. Apparent motion is what we see in marquis lights as well as every imaging display I can think of (TV, CRT, film, etc.).

Induced motion is influenced by the frame of reference. Usually, the larger object appears stationary and the smaller object appears to move. An example of this is looking at the moon on a partially cloudy, windy night. It appears that the moon is moving through the clouds, although the opposite is the case. Sometimes when I go into the lab and the lights are all out, all I can see is the screen saver running on a monitor on the other side of the room. The screen saver displays white word on a completely dark background and the words weave up and down, left and right across the screen. In the complete darkness, my visual system assumes the words are the stable frame of reference and a weird illusion of the whole room bobbing around takes place. (Or maybe it's just me)

Related to this effect is the induced movement of your location caused by movement around you. Have you ever stopped at a red light and feel like you are moving backwards if the cars on both sides of you begin to inch up? You can also get this induced movement when watching 1st person roller coaster movies, or moving camerawork in those super wide field cinemas.

Perhaps you may have noticed the autokinetic effect in a dark room where there was some sort of small LED or a light from an electronic device still on. The small spot of light appears to wander. If you fatigue your eye muscles by looking all the way to the side for a while, and then look at such a stimulus, you may notice that the object may appear to drift to one side until you correct the direction of gaze. I've even heard that the autokinetic effect has been used in psychotherapy. That is, the patient is told that the small spot is spelling out words and the Analyst asks the patient to tell him what words are being written out by the light.

We will cover the motion aftereffect on a later page, with an example.

Apparent Motion

The appearance of apparent motion is affected by the time interval between the offset of the first light and the onset of the second.

It has been hypothesize that optimal movement (at interstimulus intervals (ISIs) of 60 ms) and the phi movement at longer ISIs are due to separate motion mechanisms.

In phi motion, the spot appears to move from the first location to the second location without perception of a change in the intermediate positions.

I've tried to give examples of these different cases. The ISIs on the top are short and increase as you go down.

The bottom two figures are a demonstration of the effects of shape and color described below.

Move the cursor over the red dots to hide the other figures. (Interestingly, note how the coincidental occurrences of onsets and offsets causes all sorts of motion when all the figures are showing.)

 

 

 

The Motion Flow Field

As you move through the world, the objects out in the world change in predictable ways. By examining the way that objects move, the flow field, one can create algorithms to calculate the positions of objects relative to the observer.

There are two components related to the flow field: rotational and translational. When the viewpoint changes due to rotation, (rotating you head for example), all the components in the flow fields rotate the same amount around the axis of rotation regardless of distance.

For translational movement. Closer points shift more than further points. Try moving your head from side to side to see this. When the viewpoint translates towards the fixation, central points move less than peripheral points in the field. Therefore, distance affects both the speed and the direction of the translational component of the motion-flow-field.

This figure shows motion parallax. When your movement is lateral to the direction of gaze, the motion of objects depends on their distance and the fixation. Objects closer than fixation appear to move in the direction opposite to your direction of movement. Objects further appear to move in the same direction as you are moving.

 

Motion Thresholds

Here is an experiment in which the threshold for detecting motion was measured as a function of retinal eccentricity. The graph shows the data collected with and without correcting lenses that correct for defocus in the peripheral visual field. The thresholds for detecting motion increase in the periphery. This increase is smaller when the optical correction is used.

Thresholds for motion and the perception of velocity is also influenced by context in the scene. Thresholds for velocity are higher in blank, context-free fields than they are when objects, such as reference lines are added to the scene. The movie below is an attempt to demonstrate this. Does the motion seem faster in context? Does the dot seem to speed up near the lines?

Contrast and Color

The perception of motion and velocity is affected by factors in the stimulus unrelated to the motion itself. Perceived velocity is diminished as the contrast is reduced. A similar effect is seen with motion stimuli defined only by chromatic content. (This has suggested an M-pathway specialization.)

The graph below shows the psychometric functions from an experiment to determine the threshold for determining the speed at which a test grating is seen as moving faster than a standard grating. The center curve is a fit to the data when the test grating is of the same contrast as the standard. The curve on the right is the fit to the data for a grating that is 1/7th the contrast of the test. You can see that the threshold is higher meaning that it has to move faster to be perceived as the same speed as the test. The curve on the left is the fit to the data for a grating that is 7 times the contrast of the test. Now the threshold is lower indicating that the slower, higher contrast grating appears to move at the same speed as the test.

Here is a movie showing a high contrast luminance pattern and a lower contrast chromatic pattern both moving at the same speed. The chromatic pattern appears to move slower.

 

 

Pulfrich's Pendulum

As mentioned above, a low contrast object is perceived to move slower than a higher contrast object. Therefore the low contrast object appears to lag behind a higher contrast object moving at the same speed. A consequence of this is the phenomonon of Pulfrich's Pendulum.

Place a dark filter (such as one lens from a pair of sun glasses) over one eye. Observe an object moving back and forth in a plane perpendicular to your line of sight, such as a swinging pendulum.

The subjective speed of the object in the attenuated eye is slower and the position appears to lag. This creates a disparity between the position of the object in the two eyes which is ccompatable with an elliptical motion of the object as shown on the left.

If you have an old TV you can try an alternative version of this. Set the TV to a vacant channel so that you see snow. Now cover one eye with sunglasses. You may see the snow appear to swirl horizontally into and out of the plane of the screen. Switch the sun glasses over the other eye and the swirling should reverse.

 

 

Continue on to
Structure from Motion