The effect of various shapes of micropatterns on the strength of the Pinna illusion

Reese Manikowski, Kalif Mathieu, Alex Cartwright
Alma College
          The Pinna illusion (Pinna & Brelstaff 2000) is comprised of two concentric circles of quadratic micropatters. Each local facet has black and white sides and lie on a gray background. When the stimulus is moved toward and away from the observer, moving it across the retina, rotary motion is perceived. The colored sides of the shapes create a luminance gradient within the shape that is oriented at a 45 degrees angle from the flow of the illusion. Pinna and Brelstaff (2000) explained the illusionary motion of the image to be due the orientation of the boarder created by the blurred black and white sides of the quadrilaterals in the periphery. They asserted that orientation sensitive cells are stimulated which are tuned to the perpendicular motion of the orientation.
          It is also known that the illusion is dependent upon the micropattern within the image and not its global structure (Gurnsey & Page 2005). Both the intensity of the gradients and the orientation were found to affect the intensity of the illusion. The illusion was maximized with an orientation angle of 56.25 degrees. Changes to the most global structure of the illusion yield minimal effects (Gurnsey & Page 2005).
          Lastly, it is known that consistency within the direction of the slanted components is required (Ichikawa, Masakura & Munechika 2006). This implies that motion signals must be received from the entire structure. It is also the case that the presence of the secondary circle exaggerates the motion of the illusion when the two circles have opposing orientations. The strength of the illusion is based on the gradient orientation of the local structures. It may be the case that different shapes, other than quadrilaterals produce a stronger illusion based on this principle. In this experiment, participants will be exposed to Pinna illusions comprised of concentric circle made of different shapes to determine which produce not illusionary motion.

Methods:
Subjects
          The participants of this study were 20 (10 male and 10 female) students of Alma college varying in age from 18 to 22.

Apparatus and Stimuli
           The stimuli were presented to each participant on a sheet of paper allowing each to adjust the distance of the image manually. Three of the four images were created using Microsoft Paint (figures 1, 2 & 3). The control stimulus was a quadrilateral Pinna illusion (figure 4) similar to the original which was obtained via the internet (Van den Burgh 2006). Each stimulus contains micropatterns drawn in black and white on a gray background and orientated at a near 45 degrees (135 degrees) angle.


Figure 1 Circular Pinna Illusion


Figure 2 Triangular PinnaIllusion


Figure 3 Linear Pinna Illusion


Figure 4 Quadrilateral Pinna Illusion

Procedure
          Participants were seated and then randomly assigned to the order in which the stimuli were presented. The control was presented first followed by one of the three altered designs. The participant then rated the comparative strength of the altered illusion on a five point scale: one is much weaker, two is somewhat weaker, three is neutral, four is somewhat stronger, and five is much stronger. The same steps were taken for the remaining illusions; the control was presented followed by another illusion then was compared against the control.

Results:
          The dependant measure was the subjective strength of the three illusions compared to the original. Scores ranged from 1 (much weaker) to 5 (much stronger). The average apparent strength of the circular Pinna was 1.4, the linear Pinna was rated 2.55, and the triangular Pinna was 2.1 (see figure 5). The variability among participants was high for both the linear design and the triangular design; both illusions received scores across the entire scale, much weaker to much stronger. The circular design showed much less variability and was almost unanimously weak receiving only one 3.


Figure 5 Results of altered Pinna illusions

Discussion:
          The results of the experiment show the perceived illusion strength of the altered Pinna designs are relatively weak compared to the original illusion. The designs of the three treatments all had characteristics different than the original due to their shapes: the circular design lacks straight lines, the triangles are asymmetrical, and the lines are one dimensional. Each of these characteristics most likely provides insight into their relative weakness to the original. Gurnsey et al. (2002) cite that the micropatterns in the periphery are blurred displaying a dominant orientation created between the color differentiation between black and white. The blurred patterns in both the triangular and linear designs are different than the original. The triangular pattern is dominated by the top half which is all black while the white consists of only one line. This may have skewed the orientation of the blurred micropatterns causing weaker apparent rotation. The linear design consisted of a black and white line coupled together. This allowed the orientation to stay the same and most likely caused it to be the most effective design. However, it may have been that since the line lacked a body, the blurred image may not have been as pronounced as the control accounting for its slightly weaker rotation. The circular design was somewhat anomalous. The only difference between the circles and quadrilaterals of the two illusions was the lack of straight lines in the circular Pinna. Since the lines are blurred by the eye this does not seem like it should impact the illusion. The orientation of the two blurred images was the same. The diameter of the circle however was slightly larger than the diagonal of the quadrilateral; this may have diluted the blurred image. The quality of the printed illusion is the most plausible reason the circular design was weaker. When the circular design was on a computer monitor it showed a much stronger illusion than on paper. A noticeable difference on the paper was that the white portion of the illusion was not as pronounced. This defiantly influences the participants' ability to visualize the blurred version of the circles; if the white is too weak then the dominant orientation is lost. Theoretically this design should have scored much higher, but was stifled by its thinner lines and weak contrast to the background.

References:
Gurnsey, R. & Page, G., (2005). Effects of local and global factors in the Pinna illusion. Vision Research, 46, 1823-1837.
Gurnsey, R., Mancini, S., Potechin, C. & Sally S. L. (2002). Last but not least. Perception, 31, 1275-1280.
Ichikawa, M., Masakura, Y. & Munechika, K. (2006). Dependence of illusory motion on directional consistency in oblique components.
          Perception, 35, 933-946.
Pinna , B., & Brelstaff, G. (2000). A new illusion of relative motion. Vision Research, 40, 2091-2096.
Van den Bergh, Stefan. "Illusions." Planet Perplex. 2006