Term Paper: Perception Theory

Pages: 13 (3545 words)  ·  Bibliography Sources: 1+  ·  Level: College Senior  ·  Topic: Psychology  ·  Buy This Paper


[. . .] There are however also shortcomings to Gibson's approach. The discovery process relating to affordances and invariants in the environment appears to be very complex. He also provides no distinction between "seeing" and "seeing as." Furthermore the complete elimination of an internal representation process combined with the physical act of perception appears extreme and incorrect. It seems that the brain does have a function beyond orienting the perception organs for an optimal perception experience. This is also signified by the cognitive paradigm, and particularly the computational approach proposed by Marr.

Marr's Computational Approach

The combined shortcomings of the ecological approach and the constructivist approaches to vision seemed to favor a synthesis of both theories. In this way the shortcomings of each could be cancelled out while the best qualities of each theory could be optimized. The bottom-up approach thus works well in conditions that are favorable for vision, and the top-down theory helps to explain visual perception while viewing conditions are bad, or where there is a lack of clarity in stimulus.

Further important areas of study in the field include eye movement, reaching and grasping, time to contact, and optic flow pattern. Reaching and grasping is dependent on spatial location information, and judgments regarding direction can be made with regard to optic flow pattern. Elements that substantiate Gibson's view of direct vision include decisions regarding time to contact, detection of biological motion, apparent movement and perceived causality. Apparent motion on the other hand does appear to depend on cognitive processes, whether these be low- or high level.

Marr's theory of perception is seen from the neuropsychological level. Marr in fact compares the human visual system to a computer system. Various pathways exist by which the human eye is allowed to perceive. Specialized cells are devoted to conducting the retinal image to the visual centers of the brain. Here the image is processed, and perception occurs. This indicates the primary function of the brain as a system by which perception occurs, in contrast to the approach by Gibson, which gives the brain only minor functions of movement and orientation.

In order to reach a better understanding of the cell system involved in the process of perception, Marr (1982) conducts further investigation of visual perception on a computational level. Marr's investigation leads to three levels of theory in order to understand visual perception in its capacity as a computational device. Firstly, a computational theory should describe the phenomenon being evaluated and the reason for this evaluation. Secondly, algorithms need to be established in order to achieve the computation, as well as models for the input to and output from the algorithms. Thirdly, the implementation of the algorithms should be explained.

In contrast to Gibson, Marr studies the deep recesses of the body in order to come to a conclusion about perception. This is a very scientific approach, whereas Gibson relied on data gathered from observers and from himself in a more philosophical paradigm. For Marr then a thorough study going beyond the anatomy and physiology of the retinal ganglion cells and lateral geniculate neurons is essential in order to understand the reasons for their receptive fields. Furthermore the connections and interactions of these cells are useful in understanding their behavior. However in order to understand the reasons for the nature of the receptive fields, it is necessary to study the phenomena of differential operators, band-pass channels, and the mathematics of the uncertainty principle.

According to Marr's theory of vision then, the visual image is composed of a wide array of intensity. This could be compared to Gibson's more physical presentation of the optical array, made up of variants. The array of intensity as observed by Marr is created light reflected from observed objects.

When visual perception occurs in its early stages, according to Marr, early visual processing within the brain attempts a description of the perceived object. This is done by the construction of various representations created from the intensity values inherent in the image. This is what Marr refers to as the primal sketch. This sketch includes the description of surfaces and objects, as well as their orientations and distances away from the viewer. This first stage of vision makes local changes in light intensity explicit. Discontinuities in light intensity are thus located first, and this coincides with important boundaries in the visual scene. Thus the primal sketch is a collection of statements about edges and blobs. Information is given about their whereabouts and orientation, as well as other information for the purposes of visual definition.

The following stage then constructs boundaries and regions from the first interpretation by means of application and grouping procedures. The description is then refined, and Marr refers to this as the full primal sketch. Many of the contours and textures of an image are now perceived, but still only one aspect of early visual processing has developed.

Marr saw this model as an observer-orientated representation. He refers to it as the 21/2D sketch, produced by means of motion, depth and shading analysis, as well as a full analysis of the primal sketch. Any action requires the 21/2D sketch as guidance. This is parallel to Gibson's view of the environment and the organism interacting. The organism observes the environment, and is able to act and move because of this perception.

According to Marr a third representational level is necessary for recognition purposes. The observer needs to recognize the object to which a particular shape corresponds. This level is focused on the object, rather than the observer. Gibson's view also is centered on the object of observation, but without applying the complex processes of observation within the body. Marr terms his third level of observation the 3D-model representation.

A number of levels is thus involved in this complex theory of vision. Gibson's theory appears to be one-dimensional, with a variety of components on the same assumed level of vision. Each level in Marr's theory is an increasingly complex and refined symbolic representation of information presented in the retinal image. Each symbolic representation then builds increasing detail into the image towards final recognition and response.

This theory therefore proposes vision by means of explicit computation in terms of symbolic description of the image. The object is recognized when the reconstructed description matches a representation already stored in the brain. The original acquisitions of such stored images are however still questioned. Visual centers in the embryonic brain for example show only instinctive development patterns. This matches a computer before any programs are loaded into it.

Marr however does not intend his early processing model to solve visual perception theory problems such as initial image storage. The goal of his theory is also not to recover the actual object, but rather the initial description of its surfaces, which would then trigger the same stored image already in the mind.

This is more or less the same as Gibson's theory, where texture rather than the image itself is used in the construction of an image on the brain. There is thus a relationship for Marr between the light intensity of an image and its surface composition. Thus changes in intensity provides the observer with important information regarding the geometrical distribution and organization of the object. These physical properties of an object may include depth in visible surfaces, as well as contours and discontinuities. All of these intensity changes then provides the observer with information to decode and identify the object being observed.

Changes in intensity are however not only the result of the physical properties of an object. Changes also occur as a result of elements outside of the object. Shadows passing over the object are an example of this. There must then be a distinction between the features of a scene and the actual image. Visual perception should therefore recreate a representation of the image, and not elements from the scene.

Marr also suggests that surfaces can give information through components such as depth cues, luminance contours, texture, stereo and motion details. These are all integrated in the representation of an image. Also, the representation will be recreated in such a way that ambiguity from one cue will result in a supply of extra information from another.

Marr's theory involves the construction and manipulation of abstract symbolic descriptions of the environment. It is unnecessary to initially know or hypothesize the object of visual perception in order to know some aspects of it or to describe it partially.

This is done by the previously mentioned algorithms that are applied to the retinal image. This application results in a description closely related to a written description including the code of the image. This in turn can be likened to a programming code on a computer that describes the formation of an icon on the screen. This role is undertaken by neurons. The intensity of activity is related to inputs these neurons receive. Neurons that are stimulated by… [END OF PREVIEW]

Four Different Ordering Options:

Which Option Should I Choose?

1.  Buy the full, 13-page paper:  $28.88


2.  Buy + remove from all search engines
(Google, Yahoo, Bing) for 30 days:  $38.88


3.  Access all 175,000+ papers:  $41.97/mo

(Already a member?  Click to download the paper!)


4.  Let us write a NEW paper for you!

Ask Us to Write a New Paper
Most popular!

Abnormal Psychology: Theories, Issues, Diagnosis Thesis

Perception Research Into Aspects Research Proposal

Perceptions and Points-Of-View Who Do You See Assessment

Theory -- Horotwitz & Bartholomew Attachment Chapter Writing

Application of Nursing Theory Term Paper

View 1,000+ other related papers  >>

Cite This Term Paper:

APA Format

Perception Theory.  (2003, April 15).  Retrieved June 16, 2019, from https://www.essaytown.com/subjects/paper/perception-theory-issue/9193080

MLA Format

"Perception Theory."  15 April 2003.  Web.  16 June 2019. <https://www.essaytown.com/subjects/paper/perception-theory-issue/9193080>.

Chicago Format

"Perception Theory."  Essaytown.com.  April 15, 2003.  Accessed June 16, 2019.