Eye tracking is the process of measuring the point of view (where one sees) or the eye movements relative to the head. The eye tracker is a tool for measuring eye position and eye movements. Eye trackers are used in research on visual systems, in psychology, in psycholinguistics, marketing, as input devices for human-computer interaction, and in product design. There are a number of methods to measure eye movement. The most popular variants use video images from which the position of the eye is extracted. Another method uses a search roll or based on an electrooculogram.
Video Eye tracking
Histori
In the 1800s, the study of eye movement was done using direct observation.
In 1879 in Paris, Louis ÃÆ'â € mile Javal observed that the reading did not involve a delicate sweep of the eye along the text, as assumed before, but a series of short stops (called fixations) and fast saccade. This observation raises important questions about reading, the questions explored during the 1900s: Which word halts the eye? For how long? When do they retreat to the already visible words?
Edmund Huey built an early eye-tracker, using a kind of contact lens with a hole for the pupil. The lenses are connected to a moving aluminum pointer in response to eye movement. Kwiet studied and quantified the regression (only a small part of the saccade was regression), and he pointed out that some words in the sentence are not fixated.
The first non-intrusive eye tracker was built by Guy Thomas Buswell in Chicago, using a beam of light reflected in the eye and then recording it in the film. Buswell made a systematic study to read and view pictures.
In the 1950s, Alfred L. Yarbus conducted important eye-track research and his 1967 book is often quoted. He points out that the task assigned to the subject has a huge influence on the subject's eye movements. He also wrote about the relationship between fixation and interest:
- "All records... show conclusively that the character of the eye movement is completely independent or depends only slightly on the drawing material and how it is made, as long as it is flat or almost flat." The cycle pattern in the image examination "is not only depending on what is shown in the drawing, but also the problem facing the observer and the information he or she wishes to obtain from the image. "
- "The eye movement record shows that the attention of the observer is usually held only by certain elements of the image.... The eye movement reflects the process of human thought, so that the observer's mind can be followed to some extent from the record of eye movement accompanies the examination of a particular object.) It is very easy to determine from these records that the element draws the eye of the observer (and, consequently, his thinking), in what order, and how often. "
- "Observer attention is often drawn to elements that do not provide important information but which, in his view, can do it.Often an observer will focus his attention on unusual elements under certain circumstances, foreign, incomprehensible , and so on. "
- "... when changing the fixation points, the viewer's eye repeatedly returns to the same element of the image.Additional time spent on perception is not used to check the secondary element, but to double check the most important element. "
In the 1970s, eye search research developed rapidly, especially reading research. A good description of the research in this period was given by Rayner.
In 1980 Just and Carpenter formulated the hypotheses of strong minds , that "there is no meaningful lag between what is stuck and what is processed". If this hypothesis is true, then when the subject sees a word or object, he also thinks about it (cognitive process), and for exactly as long as the fixation is recorded. Hypotheses are often taken for granted by researchers who use eye tracking. However, the contingent-gaze technique offers an attractive option for separating blatant and confidential attention, to distinguish what is glued and what is processed.
During the 1980s, hypotheses of minds were often questioned in the light of disguised attention, attention to something that people do not see, which people often do. If hidden attention is common during eye tracking recording, the resulting scanning traces and fixation patterns often indicate not where our attention, but only where the eye has seen, fail to show cognitive processes.
The 1980s also saw birth using eye trackers to answer questions related to human-computer interaction. Specifically, investigators investigate how users search for commands on the computer menu. In addition, computers allow researchers to use eye-tracking results in real-time, especially to help disabled users.
Recently, there has been a growth in using eye tracking to learn how users interact with different computer interfaces. Specific questions asked by researchers related to how easy the different interfaces for users. The results of eye-tracking research may lead to changes in interface design. Yet another field of research has recently focused on Web development. This can include how users react to drop-down menus or where they focus their attention on websites so developers know where to place ads.
According to Hoffman, the current consensus is that visual attention is always small (100 to 250 ms) in sight. But once the attention moves to a new position, the eye will want to follow.
We still can not deduce certain cognitive processes directly from fixations on certain objects in a scene. For example, the fixation on the face in the picture may indicate confession, likes, dislikes, confusion etc. Therefore, eye tracking is often combined with other methodologies, such as introspective verbal protocols.
Maps Eye tracking
Tracker type
Eye trackers measure the rotation of the eye in one of several ways, but basically they are divided into three categories: (i) measurement of movement of an object (usually, special contact lenses) attached to the eye; (ii) optical tracking without direct contact with the eye; and (iii) measuring the electrical potential using electrodes placed around the eye.
Tracking attached to the eye
The first type uses an attachment to the eye, such as a special contact lens with embedded mirror or magnetic field sensor, and the movement of the attachment is measured with the assumption that it does not slip significantly when the eyes rotate. Measurements with tight contact lenses have provided highly sensitive eye movement records, and magnetic search coils are the method of choice for researchers who study the underlying dynamics and physiology of eye movements. This method allows measurement of eye movements in horizontal, vertical and torsional directions.
Optical tracking
The second broad category uses several non-contact optical methods to measure eye movement. Light, usually infrared, is reflected from the eyes and felt by video cameras or other specially designed optical sensors. The information is then analyzed to extract the eye rotation from changes in reflection. Video-based eye trackers typically use corneal reflections (first Purkinje images) and pupil center as a feature to track over time. A more sensitive eye tracer, Purkinje double-eye tracker, uses reflections from the front of the cornea (first Purkinje image) and the back of the lens (fourth Purkinje picture) as a feature to track. A more sensitive tracking method is an inner image feature, like a retinal vein, and follows this feature when the eyes rotate. Optical methods, especially those based on video recording, are widely used for sight tracking and are favored because they are not invasive and inexpensive.
Electrical potential measurement
The third category uses electrical potentials measured with electrodes placed around the eye. The eye is the origin of a stable electric potential field that can also be detected in total darkness and if the eyes are closed. It can be modeled to be generated by a dipole with its positive pole on the cornea and its negative pole on the retina. Electrical signals that can be derived using two pairs of contact electrodes placed on the skin around one eye are called Electrooculogram (EOG). If the eye moves from its central position toward the periphery, the retina approaches one electrode while the cornea approaches the opposite. This dipole orientation changes and consequently the electric potential field produces a change in the measured EOG signal. Reverse, by analyzing the changes in eye movement it can be traced. Because of the discretization provided by the general electrode arrangement, two separate motion components - horizontal and vertical - can be identified. The third EOG component is a radial EOG channel, which is the average EOG channel referenced to some posterior scalp electrodes. This radial EOG channel is sensitive to the potential of saccadic spikes derived from extra-ocular muscles in the onset of saccade, and allows reliable detection even for miniature saccades.
Due to the potential drifts and variable relationships between the EOG signal amplitude and saccade size, it is challenging to use EOG to measure slow eye movement and detect gaze direction. The EOG, however, is a very powerful technique for measuring saccadic eye movements associated with shifting gazes and detecting flickers. Contrary to video-based eye trackers, EOG allows recording of eye movements even with closed eyes, and thus can be used in sleep studies. This is a very light approach that, in contrast to current video-based eye trackers, requires only very low computing power; work under different lighting conditions; and can be implemented as a self-embedded and embedded system. Thus the method of choice to measure eye movement in everyday life and the REM phase during sleep. The main disadvantage of the EOG is its relatively poor directional gaze accuracy compared to the video tracker. That is, it is difficult to use EOG to determine with good accuracy exactly where the subject searched, even though the time of eye movement can be determined.
Technology and techniques
The most widely used design currently is video-based eye tracking. The camera focuses on one or both eyes and recording eye movement when viewers see some kind of stimulus. Most modern eye-trackers use pupil center and infrared/near-infrared non-collimated light to create corneal reflections (CR). The vector between the pupillary center and the reflection of the cornea can be used to calculate the point of sight on the surface or direction of the view. A simple calibration procedure of an individual is usually required before using an eye tracker.
Two common types of near infrared/infrared (also known as active light) eye-tracking techniques are used: bright pupils and dark pupils. Their differences are based on the location of the source of illumination with respect to optics. If the coaxial illumination with the optical path, then the eye acts as a retroreflector when light bounces the retina that creates a bright pupil effect similar to the red eye. If the source of illumination is offset from the optical pathway, the pupil appears dark because retoreflection of the retina is directed away from the camera.
Light-pupil tracking creates greater iris/pupil contrast, enabling stronger eye tracking with all iris pigmentation, and greatly reduces the disruption caused by eyelashes and other obscure features. It also allows tracking in lighting conditions ranging from total darkness to very bright. However, the bright pupil technique is not effective for outdoor tracking, because a foreign IR source interferes with monitoring.
Another less used method, known as passive light. It uses visible light to illuminate, something that can cause some annoyance for the user. Another challenge with this method is that the contrast of the pupil is less than the active light method, therefore, the center of the iris is used to calculate the vector instead. This calculation needs to detect the iris boundary and white sclera (limbus tracking). This presents another challenge for vertical eye movement because of the obstruction of the eyelids.
Eye-tracking settings vary widely: some are mounted on the head, some require the head to be stable (for example, with the chin remaining), and some functions remotely and automatically track the head during movement. Most use a minimum sampling rate of 30 Hz. Although 50/60 Hz is more common, today many video-based eye trackers run at 240, 350 or even 1000/1250 Hz, the speed required to capture eye fixation or correctly measure the saccade dynamics.
The eye movement is usually divided into fixation and saccade - when the eye view stops at a certain position, and when it moves to another position, respectively. The resulting fixation and saccade series is called scanpath. The subtle pursuit describes the eye after the object moves. Eye fixation movements include micro saccade: small, saccade involunter that occurs during fixation effort. Most of the eye information is available during fine fixation or pursuit, but not during saccade. Central one or two degrees from the visual angle (the visual field area falling on the fovea) provides most of the visual information; inputs from a larger eccentricity (fringe) have fewer resolutions and less color, although contrast and motion are detected better in edge vision. Therefore, the location of fixation or subtle pursuits along the scan path indicates what locus information on the stimuli are processed during the eye tracking session. On average, the final fixation is about 200 ms during linguistic text reading, and 350 ms during viewing. Setting up a saccade to a new destination takes about 200 ms.
Scanpaths are useful for analyzing intentions, interests, and cognitive meanings. Other biological factors (some as simple as gender) may affect the path scan as well. Eye tracking in human-computer interaction (HCI) usually investigates the scan path for usability purposes, or as an input method in the contingent-gaze view, also known as gaze-based interface.
Presentation of data
Interpretation of data recorded by different types of eye trackers using various software that animates or visualizes the visual, so that the visual behavior of one or more users can be graphically resumed. Videos are generally encoded manually to identify AOI (Interest Fields) or recently use artificial intelligence. Graphic presentation is rarely the basis of research results, as they are limited in terms of what can be analyzed - research that relies on eye tracking, for example, typically requires a quantitative measure of eye movement events and their parameters. The following visualizations are most commonly used:
An animation representation of a point on the interface This method is used when visual behavior is checked individually showing where users focus their views on each moment, equipped with a small path that shows the previous saccade movement, as shown in the figure.
The static representation of the saccade path This is quite similar to the one described above, with the difference that this is a static method. A higher skill level than the animation is required to interpret this.
Heat maps Alternative static representations, used primarily for agglomeration analysis of visual exploration patterns within user groups, are different from the two previously described methods. In this representation, 'hot' zones or higher density zones are set where users focus their view (not their attention) with higher frequencies. The hot map is the best known visualization technique for eye observation studies.
Blind zone maps, or focus maps This method is a simplified version of the Heat map where zones that are visually unnoticed by the user are clearly displayed, allowing for easier understanding of the most relevant information, which that is, we are being told which zone is not seen by the user.
Eye-tracking vs. gaze
Eye trackers always measure eye rotation with respect to some frame of reference. This is usually associated with a measurement system. Thus, if the measurement system is mounted on the head, just as the EOG or video-based system is mounted to the helmet, then the eye-to-head viewing angle is measured. To conclude the line of sight in world coordinates, the head must be kept in a constant position or its motion must be traced as well. In this case, the head direction is added to the inner-in-head to determine the direction of the view.
If the measurement system is mounted on the table, such as on a scleral search coil or a camera system mounted on the table, the angular view is measured directly in the world coordinates. Usually, in this situation head movement is prohibited. For example, the head position is fixed using a bite rod or forehead support. Then the head-centered reference frame is identical to the world-centered reference framework. Or everyday language, eye-in-head position directly determines the direction of the view.
Some results are available on human eye movements under natural conditions where head movements are also allowed. The relative position of the eyes and head, even with the direction of constant gaze, affects neuronal activity in the higher visual area.
Eye tracking in practice
Many studies have studied the mechanics and dynamics of eye rotation, but the purpose of eye tracking is most often to estimate the direction of sight. Users may be interested in what image features are eye catching, for example. It is important to realize that eye tracers do not provide the direction of absolute gaze, but can measure only changes in direction of view. To know exactly what the subject sees, some calibration procedures are needed where the subject sees a point or set of points, while the eye tracker records the values ​​corresponding to each view position. (Even techniques that track retinal features can not provide the right gaze direction because there is no special anatomical feature that marks the exact point at which the visual axis meets the retina, if there is a single stable point.) Reliable calibration is essential to obtain valid and repetitive eye movement data, and this can be a significant challenge for non-verbal subjects or those with unstable views.
Each eye tracking method has its advantages and disadvantages, and the choice of eye tracking system depends on cost and application considerations. There are offline methods and online procedures like AttentionTracking. There is a trade-off between cost and sensitivity, with the most sensitive systems costing tens of thousands of dollars and requiring sufficient expertise to operate properly. Advances in computer and video technology have led to the development of low cost systems that are useful for many applications and quite easy to use. Interpretation of results still requires some level of expertise, however, because systems that are not aligned or not well calibrated can produce very erroneous data.
Eye tracking while driving a car in a difficult situation
Eye movements of two groups of drivers have been filmed with special head cameras by teams from the Swiss Federal Institute of Technology: Beginning and experienced riders record their eye movements as they approach a narrow road curve. The series of images has been condensed from the original film frame to show 2 eye fixations per image for a better understanding.
Each of these stills corresponds to about 0.5 seconds in realtime.
The series of pictures shows an example of eye fixation # 9 to # 14 from beginners and experienced drivers.
The comparison of the top images shows that experienced drivers check the curve and even have No Fixation. 9 left to look sideways while the novice driver needs to check the road and estimate the distance to the parked car.
In the middle image, experienced drivers are now fully concentrating on the location where the oncoming car can be seen. The novice driver focused his eyes on the parked car.
At the bottom of the beginner picture is busy estimating the distance between the left wall and parked car, while the experienced driver can use his peripheral vision for it and still focus his eyes on the dangerous point of the curve: If a car appears there, he should give way, me. e. stop on the right rather than pass the parked car.
Eye tracking for younger and older people while walking
As you walk, older subjects are more dependent on foveal visions than younger subjects. Their running speed decreases by a limited visual field, probably due to the deteriorating peripheral vision.
Younger subjects use their central vision and peripherals as they go. Their peripheral vision allows for faster control over the running process.
Apps
Different disciplines use eye-tracking techniques, including cognitive science; psychology (especially psycholinguistics, visual world paradigms); human-computer interaction (HCI); marketing research and medical research (neurological diagnosis). Special applications include eye movement trackers in reading language, reading music, human activity recognition, advertising perceptions, and sports games.
Aplikasi komersial
In recent years, increasing sophistication and accessibility of eye tracking technology has generated a lot of interest in the commercial sector. Applications include web usability, advertising, sponsorship, package design and automotive engineering. In general, commercial eye tracking studies function by displaying target stimuli to consumer samples while eye trackers are used to record eye activity. Examples of target stimuli may include websites; television programs; sporting events; movies and advertisements; magazines and newspapers; package; display shelves; consumer systems (ATMs, cashier systems, kiosks); and software. The resulting data can be analyzed statistically and graphically given to provide evidence of certain visual patterns. By checking for fixation, saccade, dilation pupils, blinking and various other behaviors, researchers can determine many things about the effectiveness of a particular media or product. While some companies resolve this type of research internally, there are many private companies offering eye tracking and analysis services.
One of the most prominent areas of commercial eye tracking research is web usability. Although traditional usability techniques are often quite powerful in providing information about clicking and scrolling patterns, eye tracking offers the ability to analyze user interactions between clicks and how much time a user spends between clicks, thus providing valuable insights into which features are most eye-catching , which features causing confusion and being ignored altogether. In particular, eye tracking can be used to assess search efficiency, branding, online advertising, navigation usability, overall design and many other site components. The analysis can target prototype or competitor sites other than the main client site.
Source of the article : Wikipedia
0 Comments