También disponible en Español

Inf@Vis!

The digital magazine of InfoVis.net

Movement in Visualisation
by Juan C. Dürsteler [message nº 144]

The correct perception of movement has been an important routine of everyday life since the beginning of humanity. It is also an important resource in visualisation.
KinDep.gif (5219 bytes)
Kinetic Depth, Just by looking at the shadow on a screen of a twisted, rotating wire hanger you can capture its 3D structure. Whe the wire hanger is still this is impossible.
Source: Adapted by the author from the drawing in Colin Ware's book  "Information Visualization"

Motion is perceived strongly by the brain, which evolved by surviving different threats, many of them related to interpreting static patterns like poisonous fruits or dangerous fauna and dynamic patterns like moving animals or objects that could collide (or do something more concerning) with the brain’s owner. 

An important outcome of motion perception is that it facilitates the detection of small or barely discernible objects and it allows the brain to ascertain the three dimensional shape of an object (the so called kinetic depth). 

The perception of dynamic patterns is not so well understood compared with the perception of the static ones, but three effects appear to be prominent in deriving structure and properties of the object from the perception of its movement: Kinetic depth, Optic flow and Motion parallax.

Kinetic depth

Wallace and O’Connell (1953) showed this phenomenon in a simple way. If a wire hanger is twisted into a random three-dimensional figure, it’s impossible just by looking at the shadow of it on a piece of white paper to infer what the 3D figure is. But if the hanger rotates, its 3D shape is quickly perceived just by looking at the shadow. 

Recent experiments by Todd and Norman (1991) have shown that just by presenting alternately to the user two projected surfaces of random dots taken from a 3D object suffices to perceive the shape of the object. See for example the  rotating sphere (requires Java) or the  moving cylinder (requires PowerPoint or PowerPoint viewer)

Optic flow

Optic flow occurs when we are moving in a particular direction. If we look towards the point to which we are heading (the focus of expansion) this doesn’t show movement, whereas by looking at the surrounding space, the visual field appears to be expanding. Apparently this effect can be perceived with great accuracy by the human brain, contributing to the control of locomotion and helping to continue heading towards the specified location. 

OptFlow.gif (21880 bytes) DifMotion.gif (15950 bytes)
Optical Flow 
Source: Adapted by the author from the drawing in Colin Ware's book  "Information Visualization"
Differential motion parallax 
Source: Adapted by the author from the drawing in Colin Ware's book  "Information Visualization"

This effect gives the sense of movement that occurs when we see the famous star field screensaver, which produces the effect of navigating through a field of stars, heading towards a particular point on the screen.

Differential motion parallax

Differential motion parallax is the perceived difference in speed of nearby objects compared to far away ones. Motion parallax is easily perceived when we look through a car window, the velocity of nearby objects appears to be greater than that of distant objects.

On the other hand It’s also worth considering several phenomena relevant for the perception of movement.

Apparent motion 

Apparent motion is achieved whenever a series of still images compatible with the individual stages of motion are presented in rapid succession to the visual system. This perceptual principle lies in the foundations of cinema, television and animation.

Apparent motion was first shown by Wertheimer in 1912 who was one of the founders of the Gestalt group. He presented to users two vertical lines alternately in different spatial locations. When the delay between two successive appearances was under 1/20th of a second the observers reported that there was a line moving smoothly from one location to another.

MovAparent1.gif (2360 bytes) MovAparent2.gif (2562 bytes)
Apparent motion. If the displacement of the shapes is less than ð/2 coherence is maintained and the illusion of movement in a particular direction appears. Otherwise it could appear that the movement is the inverse than the intended one.  Apparent motion. Assigning different colours, shapes or sizes allows you to follow the movement disambiguating it when the pattern moves more than ð/2.

For apparent motion to occur there has to be a correspondence between the patterns included in successive frames. If motion is represented using a sequence of identical shapes spaced at regular quantity d, there is a strict limitation given by the fact that between frames the displacement of the pattern has to be less than ð/2 in order to avoid giving the impression of a movement in the inverse direction. 

This is what occurs in some movies where the wheels appear in some moments to rotate in the opposite direction. This limit can be overcome by making the elements different in shape, colour or size, so that the eye con focus on them avoiding the illusion.

Perception of causality

In 1946 Michotte carried out a series of studies that showed that causality can be as easily and immediately perceived as colour or shape, under certain circumstances. In one if these studies the observers were presented with a square moving towards a second one. When the first reached the second it stopped moving and the second began to move. The observers perceived that if the movement of the second square began less than 70 milliseconds after the “collision”, it was the first square causing the movement of the second. 

The perception of causality disappeared if the elapsed time was greater or if the direction of the second square’s movement was different than that of the first one. Other researchers have found different results but the important thing is that movement and event timing under certain circumstances can directly trigger the perception of causality, owing little to inference or elaborated experiences. 

Biological movement 

Most people are able to interpret the movements of other people or more generally speaking, the movement that has a biological origin. Experiences carried out by Johansson in 1975 attaching lights to actor’s limb joints dressed in black and then filming their movements in the dark showed that, despite the result being just an apparent cloud of white dots moving, observers were immediately aware that they were looking at human movement. Even when the movement was filmed with only six lights for a very short segment the biological movement was clearly perceived. See the revealing example of biological movement by Alos Feher e Ilona Kovacs

Pelea.gif (9230 bytes)
Biological movement. Even with an extremely elementary sketch like this, movement can express emotions and actions. 
Source: Adapted by the author from Rita Carter's book "Exploring Conciousness"
Biological movement. Heider & Semmel showed that moving appropriately three simple figures without any biological resemblance can create  the perception of human behaviour. 
Fuente: Adaptado por el autor del libro de Rita Carter "Exploring Conciousness"

Other experiences show that observers can make very precise discriminations, like identifying postures, gestures and the position of the limbs. They are also able to judge successfully the gender of the actor in 85% of the times. 

Moreover Heider and Semmel in 1944 devised a movie showing two triangles and a circle moving. When presented to observers the movements were consistently interpreted as conveying human emotions like being angry or that some shapes were chasing the others. 

Other researchers have found in cross-cultural studies that motion can express concepts such as kindness, fear, and aggression, in a way that could be considered universal.

Movement is clearly under-utilised as an attribute to convey information. Apparently it has compelling properties to do so. A great field of research and of trials of the expressive opportunities of motion to communicate certain kinds of relationships among data lies ahead. 


Links of this issue:

http://www.infovis.net/printRec.php?rec=llibre&lang=2#InfoVisWare   Colin Ware's Information Visualization
http://aris.ss.uci.edu/cogsci/personnel/hoffman/Sphere3.html   Rotating sphere
http://www.artsci.wustl.edu/~rabrams/11.Movement_files/frame.htm   Moving cylinder
http://zeus.rutgers.edu/~feher/kutya_e/example1.html   Example of biological movement
http://www.amazon.com/exec/obidos/tg/detail/-/0520237374/infovisnet/   Rita Carter's Exploring Conciousness
© Copyright InfoVis.net 2000-2014