Visual perception

In physiology, visual perception is the recording and processing of visual stimuli, in which takes place over the eye and brain extraction of relevant information, detection of elements and their interpretation by comparison with memories. Thus, the visual perception goes far beyond simply recording information.

Physiological Details of the perceptual apparatus can be found under visual system.

• 3.1 Perception of Space
• 3.2 Object Extraction
• 3.3 Object Detection 3.3.1 Balance of features
• 3.3.2 Detection of complex objects

History

In ancient times there were different theories of visual perception:

The perception theory of Euclid ( 365-300 BC) grappled with problems of space perception (eg, perspective and size constancy ). The lines of sight are based on this theory from the eye and determine the perception. This theory of the " line of sight " seems to be rather absurd in the light of physical considerations, but acquired the modern findings of eye movement analysis, a kind of late rehabilitation ( see foveal perception).

The perception theory of Empedocles ( 492-432 BC) said seems to be the opposite. The of the pores (today we would say pigments) of things outgoing influences, a type of radiation that affected by the light (today we would say reflected), invade the senses, and be perceived, if there an equivalent find ( like is known by like ). This theory, which deals with recognizing things of the external world, is modern in its approach as it seems. It applies from today's perspective on peripheral perception, so also acts on the viewer, even if it does not looking.

Alhazen or Ibn al-Haytham, the "father of optics " (965-1040), proved first that the visual perception related to the light that strikes the eye. He placed first on the assumption that the visual perception in the brain and not in the eye occurs.

Through experiments, he demonstrated that the perception is influenced by the personal experience of a man. Al- Haytham conducted experiments on human perception and complemented the work of Ptolemy on stereoscopic vision.

Leonardo da Vinci (1452-1519) was the first to the visual feature of the human eye. He wrote: " The eye has a single central line, and all things which pass through this line to the eye, are well seen. To this line, there are an infinite number of other lines that come with the center line of contact and what is even more ineffective, the further they are away from said center line. "

Came to this view Leonardo with observations and with the help of optical experiments. "The eye, of which we experience so clearly reveals the function was described by an infinitely large number of authors in a particular way, but I find that it is very different. " He is the discoverer of the difference between fovealem and peripheral vision.

Hermann von Helmholtz is often regarded as the father of modern visual perception theory. He compared the eye with optical devices and found its design features very primitive. In theory, the eye could provide no useful Sehresultate. He concluded that the perception can only come through " unconscious inferences " about, made ​​possible by existing perceptual experiences.

Such perceptual experiences are, for example:

• Light usually comes from above
• Objects can not be seen from below
• Faces are detected in the upright position

The study of optical illusions, has shown the unconscious conclusions can be included in the process of perception.

Another type of unconscious inferences based on the probability sensation, which is determined by the frequency already made ​​a similar perception.

Perceptual process

Receiving stimuli

Through the dioptric apparatus of the eye a wrong and upside-down image on the retina. The light stimuli from the sensory cells of the retina, the rods (brightness ) and cones ( color vision ) is registered. The ratio of cell types differs depending on the location on the retina; in the fovea there are only cones. Rods and cones form a membrane potential, which is transmitted via bipolar cells to ganglion cells when exposed to light.

Forwarding and encoding

Each ganglion cell processes information from a receptive field ( a spatially limited area of ​​the retina ). There are two main types of ganglion cells, on-and off-center cells, which play an important role, especially in edge detection. On- center cells respond when light falls in the center of the receptive field, and reduce their firing rate when peripheral areas of the field are more exposed. Off-center cells behave the other way around and cheer enhanced when light stimuli are received in marginal areas of the receptive field.

The information of the ganglion cells are routed through the visual pathway to the left and right lateral geniculate body. The output of the on-and off-center cells is connected in the lateral geniculate bodies so that edges ( ie areas where a change in brightness occurs) or bar ( brightness changes and return to the initial intensity ) are extracted. The processed information is projected from the lateral geniculate to the visual cortex. This processing of the signals includes an increase of the contrast perception via feedback processes which ensure that the same high light sensitivity is maintained for contrast enhancement.

Interpretation

Perception of Space

The perception of space is based on several methods to create from the two-dimensional image on the retina of a representation of the three-dimensional world. Through the stereoscopic vision space information can be constructed from the slight differences between the recorded images from the pair of eyes. The viewer moves relative to objects in space, so the images on the retina move slower, the further the object is away from the viewer. In addition, spatial perception can take place over the texture gradients, ie of changes in the texture depending on the spatial distance.

Object extraction

Law of the unity and continuity of the law

Law of proximity

Before objects can be recognized and interpreted, must first be extracted from the information about where objects are located and which of the detected lines belong to an object. This form of heuristics (also shape principles, although the historical term Gestalt laws should be avoided ) all work on the principle of finding a succinct as possible solution ( Gestalt psychology ):

• Conciseness: A figure is perceived that it corresponds to the simplest possible structure
• Nearby pixels are perceived as belonging together if they are close together
• Similarity: image parts with the same shape or color to be seen as a whole
• Symmetry: symmetric structures are assigned to the same object
• Joint movement (common fate ): the same movements and simultaneous appearance or disappearance of picture elements creates a togetherness
• Continuity: image elements that seem to be a continuation of previous or broken items, be regarded as belonging together
• Unity: lines that enclose an area, be easier as a unit under the same circumstances be regarded as those who are not close together
• Common region: Elements in demarcated areas are perceived as belonging together
• Related items: Related elements are perceived as an object

Object recognition

In the object recognition, the extracted objects are interpreted and can be matched with memories.

Balance of features

An object can be matched with an existing memory by feature analysis. In this model, it is assumed that a shape or an object is represented in a set of abstract features. The letter K has, for example, the features " long vertical bar ", "short stroke with 30 ° inclination " and " shorter stroke with approx 75 ° tilt ." This model is superior to that of a template matching by also modified patterns are detected. When "K" can be changed size, location, rotation, or font, so a "template " would no longer fit. However, under all these transformations, the features remain the same. In addition, the feature matching of encoding by ganglion cells (extraction of lines) corresponds.

Recognition of complex objects

For the detection of complex objects there is a " theory of volumetric detection"

• The complex object is divided into simpler components (three-dimensional ), and each component of a " Geon " (of " Geometrical icon" ) is assigned. Geons are a set of simple, three-dimensional objects, from which you can create more complex by combining objects.
• If the child objects ( = geons ) identified and their arrangement determines each other, the ( total ) are allocated object.

When assigning the actual recognition takes place. The object is classified and then represents an object of a category (such as assignment as " dog " or "Phone" )

Face perception

The recognition of faces is one of the most important social perception performance of humans. Faces have to be in contrast to many everyday objects identified singular and therefore are a special kind of schemes of perception. For face recognition but serves not only the imprint of specific characteristics ( eg size of the nose ), but also on their relationships (eg: eye relief ).

How important is the configuration of features that could prove Leather and Bruce ( 2000). Subjects were asked to memorize different faces and were checked afterwards. Here them the pictures were shown both upright and upside down. In this inverted position, it was primarily the relations between the features, which allowed the subjects to recognize the faces.

As is already known from brain research, injury or damage to certain cortical areas, the identification of familiar faces is no longer possible ( prosopagnosia ). This suggests the theory that the perception of faces is well supported by specialized cortical fields.

Examples of investigation of visual perception

In the years after 1960, eye movements have been increasingly recorded and analyzed, such as when reading a text in the image viewing and later also in solving visual problems and when driving. The picture on the left shows what can happen in the first 2 seconds of viewing an image. The background is blurred by peripheral vision. Nevertheless, one can realize that there is a scene in a room with people. The first eye fixation shows a couple of men's shoes, perhaps because they have a strong contrast and also are very close to the initial position of the eye fixations. All of the following fixations jump from face to face.

One can conclude that the human face receives the most attention in the normal case, because it allows an identification or assessment of a person on the basis of biometric similarities and so already allowed a first assessment of an interpersonal situation.

An essential fact: The human perception is heuristic and non-linear, that is, there are those parts of an image considered, which contain additional information, while subjectively unimportant or already well -known picture elements are not fixed ( additional examples see eye tracking ).

On computational level a theory of processing of the visual impressions of visual perception in the brain was created by David Marr in the 1980s, which finds its application in the field of artificial intelligence. This theory has never been tested experimentally.

Neurophysiological Theory

A neurophysiological theory of visual perception is exemplified in Article perception.

Related areas

• Psychophysics
• Neuroscience
• Cognitive Psychology
• Optometry
• Ophthalmology
• Unaufmerksamkeitsblindheit
• Change blindness
• Imagination
• Response priming