JP4385197B2 - Eye movement data calibration coefficient calculation device and eye movement data calibration coefficient calculation program - Google Patents

Description

The present invention, eye movement data obtained by measuring the eye movement of the subject, about the eye movement data calibration factor calculation equipment Contact and eye movement data calibration factor calculation program for calculating a calibration factor to calibrate.

  In general, the human eye is called a window of mind, and is considered a means of exploring the inside of a person from the appearance. For example, what kind of vision is obtained by examining where the eye is looking (what) It is said that it is possible to know objectively and non-invasively (non-intrusive) whether information is needed and processed in the head or what a person is thinking.

  From this point of view, it is assumed that it can be used for various studies by measuring and analyzing eye movements when looking at something (something), that is, eye movements. For example, research on the role of gazing points in assessing the image quality of video, research on the difference between the eyes of a so-called professional who studied something and an amateur who just started something, The research (for example, refer nonpatent literature 1) etc. regarding the application to the initial diagnosis of Alzheimer type dementia etc. are mentioned.

  In addition, research using these eye movements has been energetically performed by many researchers outside of Japan in various fields such as psychology, human interface, sports engineering, and medical care.

  For example, when an eye movement measuring device is taken out of a laboratory (medical site, sports gym, etc.) other than a laboratory that measures eye movement, advanced knowledge is required for complicated connection and setting between the devices constituting the device. Yes, it requires a lot of labor to handle.

  In other words, it is still difficult to say that measurement of eye movement can be performed easily and immediately by anyone. That is, until an experimenter who measures eye movement purchases a device (eye movement measurement device) that measures eye movement and obtains valid data (eye movement data), the experimenter handles the device. Because it takes time and effort to measure the eye movement (subject to labor), it is necessary to request devoted cooperation.

  Conventionally, when measuring the eye movement of a subject, in order to perform the initial setting of the eye movement measuring device and to increase the accuracy of the measured eye movement, the measurement is performed in advance by the eye movement measuring device. A close calibration work was required to correlate (associate) the eye position (or device output value, usually just a digital value) of the subject (detected) with the viewing position actually viewed by the subject. . This association is referred to as “conventional calibration work”.

  Here, “conventional calibration work” performed before the conventional eye movement measurement will be described with reference to FIGS. 14 to 16.

  FIG. 14 schematically shows a calibration board (calibration board) for calibrating eye movement data called “5-point method”. As shown in FIG. 14, in the calibration board 50, five LEDs 52 (52a, 52b, 52c, 52d, 52e) are incorporated in the center of the top, bottom, left, and right, and the distance from the calibration board 50 to the subject is determined. By measuring in advance and sequentially turning on these LEDs 52 during calibration, the viewing angle viewed from the subject is determined for each LED 52. Then, the calibration coefficient can be obtained based on the measurement output (eye movement data) of the eye movement obtained when the subject gazes at the sequentially lit LED 52 and the viewing angle between the LEDs 52 gazed by the subject. A thing seen by the subject employed when obtaining the calibration coefficient, such as the LED 52, is referred to as a “target”.

  Note that the number of LEDs 52 incorporated in the calibration board is not limited to five, but the “13-point method” employing 13 LEDs as shown in FIG. 15 (the “9-point method employing nine LEDs”). ”) (LEDs are not shown in FIG. 15). In FIG. 15, the target of “5-point method” is indicated only by “▲”, the target of “9-point method” is indicated by “▲” and “●”, and “13-point method” Targets are indicated by “▲”, “●”, and “■”.

  As the number of targets such as the LED 52 increases, the accuracy of the calibration coefficient and thus the accuracy of the eye movement data can be improved.

  FIG. 16 shows an example of an evaluation result (calibration result) for a measurement error described later by the “5-point method”, “9-point method”, and “13-point method”. In FIG. 16, “◯” indicates the position of the visual target (gazing target) that the subject gazes at, and “Δ”, “●”, and “□” indicate the actual measured positions. In FIG. 16, in the region (field of view) above the coordinate point (0, 0) of the xy coordinates in the y-axis direction, the calibration result by the 9-point method and the calibration result by the 13-point method overlap, In the lower region (field of view) in the y-axis direction, the portion that has not been calibrated very well in the nine-point method is improved well in the thirteen-point method (it is linear as in the gaze target). .

  Here, the evaluation result for the measurement error shown in FIG. 16 will be described. In the five-point method indicated by “Δ”, the coordinate point (0, 0) of the xy coordinates is used as a reference in the y-axis direction. The upper visual field has a small amplitude and is slightly inclined, and the lower visual field in the y-axis direction spreads in a trapezoidal shape and is slightly distorted in the lower right direction. When the number of calibration points is increased to the nine-point method indicated by “●”, the accuracy of the upper visual field in the y-axis direction is extremely improved, and the accuracy of the lower visual field in the y-axis direction is also improved. When the number of calibration points is further increased to the 13-point method, as shown by “□”, in the upper field of view in the y-axis direction, the 9-point method has already obtained sufficient accuracy, but it does not change much. It can be seen that a dramatic improvement is obtained in the lower visual field in the y-axis direction, which almost coincides with the position of the gaze target.

  As shown in FIG. 16, as the number of targets is increased (5-point method <9-point method <13-point method), the accuracy of the eye movement data is greatly improved. The restraint time to restrain becomes longer, the fatigue of the subject increases with the increase of the restraint time, and it becomes difficult for the subject to gaze all the targets correctly in one experiment of measuring eye movements, There is a drawback that the number of cases where the gaze fails and the experiment is repeated increases.

  Further, in the conventional calibration work, it is necessary to appropriately adjust the positional relationship between the eye of the subject and the eye movement sensor (a part of the eye movement measuring device) that detects the eye movement. Skill improvement is required, and the subject is required patience during adjustment, and it takes a lot of effort to measure eye movement.

  Therefore, in order to reduce such labor, a one-piece portable wireless eye movement analysis device and an eye movement data automatic calibration algorithm have been developed (for example, see Non-Patent Document 2). Even with the analyzer disclosed in Non-Patent Document 2, calibration work may be difficult. Further, in the automatic calibration algorithm of eye movement data, the subject is instructed verbally to perform the calibration work, for example, five LEDs are arranged in the center of the top, bottom, left, and right on a board installed almost in front of the subject, These LEDs are turned on sequentially, and the subject gazes at them, and by observing the turned-on LEDs and the movement of the eyeball of the subject, the eyeball position of the subject where the eye movement is measured, and the subject actually seeing The viewing position is associated (associated) (see, for example, Patent Document 1).

M.M. Fujii, M .; Yamada et. al. : Visuo-Constructional Disturbance in Patents with alzheimer's Disease, Society for Neuroscience Abstracts, 306.18, Vol. 18, 1992 M.M. Yamada et. al. : Development of an Eye Movement Analyzer Positioning Functions for Wireless Transmission and AutoCalibration, Medical and Biological Engineering, 28. 317-324 1990 JP 2000-005130 A (paragraphs 0026 to 0030, FIG. 1)

  However, it is difficult to perform conventional calibration work on subjects who have advanced symptoms such as Alzheimer's disease and dementia, infants, and other subjects who are difficult to give verbal instructions. There is a problem that fine contrivance is required, such as calling attention while pointing to the LED that is being operated.

  In addition, in the “eye movement measurement method and apparatus” disclosed in Patent Document 1, a general image is presented to a subject who is difficult to gaze at the lit LEDs in order. The eye movement (line of sight) of the subject is recorded, and the position of the subject's gazing point and the shape of the eye movement (the shape of the line of sight) and the preset area (the area that is likely to be gazed from a general image) ) And the eye movement data measured based on the correspondence result is calibrated.

  However, if there is a possibility that the subject does not necessarily see the gaze area obtained from a general image in advance, or if the order of viewing the areas is different, the eyeball position of the subject to be measured for eye movement and the subject's eyeball position are measured. There is a problem that it is difficult to obtain the correspondence with the actual position of the gaze point (viewing position).

Therefore, in the present invention, the above-described problem is solved, and the correspondence between the eyeball position of the subject whose eye movement is measured and the position of the gaze point (viewing position) at which the subject is actually gazing is accurate. In addition, eye movement data calibration coefficient calculation that can perform calibration work when measuring eye movements for subjects with advanced symptoms such as Alzheimer's disease, dementia, infants, etc. an object of the present invention is to provide the equipment you and eye movement data calibration coefficient calculation program.

  In order to solve the above-mentioned problem, the eye movement data calibration coefficient calculation device according to claim 1 is an eye movement data calibration coefficient for calculating a calibration coefficient for calibrating eye movement data output from an eye movement measurement device for measuring eye movement. Computation device, eye movement data calibration graphic data storage means, eye movement data calibration graphic pattern generation output means, eye movement data acquisition means, eye movement data graphic pattern association means, correspondence degree calculation means And a calibration coefficient calculation means.

According to such a configuration, the eye movement data calibration coefficient computing device uses the eye movement data calibration graphic data from the eye movement data calibration graphic data accumulated in the eye movement data calibration graphic data accumulation means by the eye movement data calibration graphic pattern generation output means. Generates the movement data calibration graphic pattern and divides the display screen of the display device into a matrix based on the position data indicating the display position when displaying the eye movement data calibration graphic pattern on the display screen of the display device among areas formed by, by Outputs sequential multiple times into a plurality of regions designated Me pre, in the display screen, moving the eye movement data calibration graphic pattern.

  The eye movement data calibration graphic pattern that is generated based on the eye movement data calibration graphic data and displayed on the display device is a graphic pattern that is assumed to have high attractiveness. This attractiveness refers to a characteristic in which the subject looks at the object without being conscious as a result of examining the movement of the eye in a semi-unconscious state by a psychological experiment.

  In addition, the reason why the eye movement data calibration graphic pattern is generated by accumulating the eye movement data calibration graphic data without sequentially accumulating the eye movement data calibration graphic pattern is as follows. This is because it is necessary to display data calibration graphic patterns with different colors, brightness, or sizes (similar shapes), and display them on a display device rather than storing them as eye movement data calibration graphic patterns. This is because it is assumed that it is possible to flexibly cope with this, and it is possible to output eye movement data calibration graphic patterns in various display modes. That is, the eye movement data calibration graphic data is a data group in which the coordinates, color, and luminance of the generated graphic pattern are parameterized.

  Further, the plurality of areas designated in advance are divided portions obtained by partitioning the display screen in a matrix, and the eye movement data calibration graphic pattern is output (displayed) to these areas. For example, the eye movement data calibration graphic pattern is continuously displayed so as to draw an inverted S-shape from the upper left area of the display screen to the lower right area. Then, display of the eye movement data calibration graphic pattern is performed a plurality of times. By displaying the graphic pattern continuously in adjacent areas, the eye movement can be seen by the subject as if the eye movement data calibration graphic pattern moves (draws a trajectory).

  However, when a well-understood subject who understands the purpose of calibrating eye movement data carefully and carefully watches the eye movement data calibration graphic pattern, the graphic pattern is output a plurality of times (for example, three times) ( It is not necessary to display), and the calibration can be completed with one output (display).

  The eye movement data calibration coefficient calculation device acquires the eye movement data output from the eye movement measurement device when the eye movement data calibration graphic pattern is displayed on the display device by the eye movement data acquisition means. The eye movement data and the eye movement data calibration based on the eye movement data and the position data when the subject gazes at the eye movement data calibration figure pattern displayed on the display device by the means for associating the eye movement data figure pattern. Correlate with the graphic pattern. That is, in the eye movement data acquisition means, the eye movement data in which the subject is gazing at the eye movement data calibration graphic pattern that is output a plurality of times in a plurality of areas specified in advance is acquired. The data graphic pattern association means associates the position data of the eye movement data calibration graphic pattern, that is, the data for specifying the region and the eye movement data.

Thereafter, the eye movement data calibration coefficient calculation unit, by corresponding calculating means, for computation of the degree of correspondence between the eye movement data and the eye movement data calibration graphic pattern associated with eye movement data graphic pattern correlation means . Here, the correspondence calculation means includes data comparison means and data determination means. The eye movement data calibration coefficient computing device did not match the number of times that the area and the line of sight indicated by the eye movement data matched in the area where the eye movement data calibration graphic pattern was output by the data comparison means. If the number of matches is large, that is, if the number of eyes that are gazing at the figure pattern for calibrating eye movement data exceeds the number that is not gazing, the data judging means will use eye movement. It is assumed that eye movement data corresponding to the data calibration graphic data is obtained.
The eye movement data calibration factor calculation unit, the Calibration coefficient calculation means, based on the calculated degree of correspondence, calculates a calibration factor to calibrate the eye movement data to eye movement data calibration graphic pattern. If this calibration coefficient is used, it is possible to calibrate (correct) the eye movement data measured by the eye movement measuring apparatus.

The eye movement data calibration coefficient calculation device according to claim 2 is the eye movement data calibration coefficient calculation device according to claim 1 , wherein the figure data output means for eye movement data calibration includes a figure pattern display change means. Features.

  According to such a configuration, the eye movement data calibration coefficient computing device is configured to display the eye movement data calibration figure by the data judgment means in the area where the figure pattern display changing means outputs, for example, the eye movement data calibration figure pattern. The display method of the eye movement data calibration graphic pattern to be displayed in an area where it is not determined that the eye movement data corresponding to the pattern has been obtained is changed. The change of the display method by the graphic pattern display changing means is such that the shape of the eye movement data calibration graphic pattern is left as it is, and the eye movement data is not displayed in the area where it is determined that the eye movement data is obtained. Change to display graphic patterns only in areas where it is not determined that data has been obtained.

The eye movement data calibration coefficient calculation device according to claim 3 is the eye movement data calibration coefficient calculation device according to claim 1 or 2 , wherein the eye movement data calibration figure is located in a specified area in the area. It is characterized by comprising a calling means for calling the test subject that the eye movement data calibration graphic pattern is output when the pattern is output.

  According to such a configuration, the eye movement data calibration coefficient calculation device outputs the eye movement data calibration graphic pattern to the subject when the eye movement data calibration graphic pattern is output to the specific area by the arousing means. Awaken (notify) what has been done. For example, even if the eye movement data calibration graphic pattern has a high attractiveness, in the case where the subject is a dementia patient or an infant, it may not be possible to continue watching the graphic pattern. To prompt attention to the graphic pattern. In the conventional method, the LED lamp that is lit as an artificial arousing means is struck with a finger to urge attention. The specific region corresponds to a region effective for calibration work indicated by “Δ”, “●”, and “□” in FIG. For example, when the eye movement data calibration graphic pattern is output to the area, the subject is particularly encouraged to watch the graphic pattern.

5. The eye movement data calibration coefficient calculation program according to claim 4 , wherein the eye movement data calibration figure pattern generation output is obtained from an apparatus for calculating a calibration coefficient for calibrating eye movement data output from an eye movement measurement apparatus for measuring eye movement. Means, eye movement data acquisition means, eye movement data graphic pattern association means, correspondence degree calculation means, and calibration coefficient calculation means.

  According to this configuration, the eye movement data calibration coefficient calculation program generates the eye movement data calibration figure pattern based on the eye movement data calibration figure data by the eye movement data calibration figure pattern generation output unit, and the display device. Are sequentially output to a plurality of predetermined areas on the display screen. The eye movement data calibration coefficient calculation program acquires the eye movement data by the eye movement data acquisition means, and the eye movement data acquired by the eye movement data graphic pattern association means and the eye movement displayed on the display device. Based on the position data of the data calibration graphic pattern, the eye movement data is associated with the eye movement data calibration graphic pattern. Thereafter, the eye movement data calibration coefficient calculation program calculates the degree of correspondence between the eye movement data associated with the eye movement data graphic pattern association means and the figure pattern for eye movement data calibration by the correspondence degree calculation means, and performs calibration. The coefficient calculation means calculates a calibration coefficient for calibrating the eye movement data with respect to the eye movement data calibration graphic pattern based on the correspondence calculated by the correspondence degree calculation means.

According to the first and fourth aspects of the present invention, the eye movement data and the eye movement data are based on the eye movement data and the position data, which are eye movements when the subject is gazing at the eye movement data calibration graphic pattern. The calibration graphic pattern is associated with each other, the correspondence between the eye movement data and the eye movement data calibration graphic pattern is calculated, and a calibration coefficient for calibrating the eye movement data is calculated based on the correspondence. Therefore, the eye movement data (eye position) of the subject whose eye movement is measured and the position (area, viewing position) of the gazing point that the subject is actually gazing, that is, the eye movement data calibration graphic pattern Since the eye movement data can be calibrated based on the calibration coefficient, it can be used for subjects with advanced symptoms such as Alzheimer's disease and dementia, infants, and other subjects who are difficult to give verbal instructions. On the other hand, there is no need to give verbal instructions, and calibration work can be performed when measuring eye movements.

Number According to the invention of 請 Motomeko 1,4 described, in the area where the eye movement data calibration graphic pattern is output, the number of times the line of sight indicated by the area and the eye movement data match does not match Since the eye movement data corresponding to the eye movement data calibration graphic data is obtained, the correspondence between the eye movement data and the eye movement data calibration graphic pattern can be made accurate.

According to the invention described in claim 2 , for example, in the region where the eye movement data calibration graphic pattern is output, it is displayed in the region where it is not determined that the eye movement data corresponding to the eye movement data calibration graphic pattern has been obtained. The display of the eye movement data calibration graphic pattern to be changed is changed.For example, the graphic pattern is not displayed in the area where it is determined that the eye movement data is obtained, and the graphic pattern is displayed only in the area where it is not determined that the eye movement data is obtained. Change to display. For this reason, if the method (display method) of displaying the eye movement data calibration graphic pattern is changed for an area for which eye movement data has not been acquired, the eye movement data can be acquired efficiently.

According to the third aspect of the present invention, when the eye movement data calibration graphic pattern is output in a specific area, the subject is alerted (notified) that the eye movement data calibration graphic pattern has been output. Therefore, for example, even if the eye movement data calibration graphic pattern has a high attractiveness, it may not be possible to continue watching the graphic pattern when the subject is a dementia patient or an infant. Then, it is possible to urge the attention means to pay attention to the graphic pattern.

Next, embodiments of the present invention will be described in detail with reference to the drawings as appropriate.
<Outline of eye movement data calibration coefficient calculation system>
FIG. 1 is a block diagram of an eye movement data calibration coefficient calculation system. As shown in FIG. 1, an eye movement data calibration coefficient calculation system A calculates a calibration coefficient for calibrating eye movement data obtained by measuring the eye movement of a subject, and includes an eye movement data calibration coefficient calculation device 1 and an eyeball. The motion measuring device 2 and the display device 4 are included. Prior to the description of the eye movement data calibration coefficient calculation device 1, the eye movement measurement device 2 and the display device 4 will be described.

  The eye movement measuring device 2 measures (detects) the eye movement of the subject, and outputs the eye movement data as a result of the measurement to the eye movement data calibration coefficient calculation device 1. This eye movement measuring device 2 is a method for observing eye movement data of a subject (horizontal eye movement data, vertical eye movement) measured when gazing at the eye movement data calibration graphic pattern displayed on the display screen of the display device 4. Motion data). In this embodiment, the eye movement measuring device 2 is installed at a position where the eye movement of the subject can be measured, that is, at a position facing the face of the subject, for example, on the display surface side of the display device 4.

  The eye movement measurement method (measurement method) includes various conventional general methods (for example, corneal reflection method, scleral reflection method ["Recent Trends in Eye Movement Research", Mitsutoshi Yamada, Shingaku) Information MBE 95 • 132, NC 95.90 (1995.12)]) can be used. Here, one of the corneal reflection method and the scleral reflection method is adopted.

Here, the principles of the corneal reflection method and the scleral reflection method will be described with reference to FIGS. FIG. 2 is an explanatory diagram for explaining the principle of the corneal reflection method.
As shown in FIG. 2, in the corneal reflection method, the eyeball E is irradiated with infrared light from a spotlight 30 that is a light source using an infrared light emitting LED, and a virtual image VI by the spotlight 30 is displayed as the eyeball. Projected on E. Then, since the curvature O ′ of the cornea C, which is the black eye portion of the eyeball E, does not correspond to the rotation center O of the eyeball E, the position of the virtual image VI relatively moves (shifts) according to the rotational motion of the eyeball E. Thus, in the corneal reflection method, the amount of movement (movement amount D) is detected as the amount of movement of the eyeball E.

  FIG. 3 is an explanatory diagram for explaining the principle of the scleral reflection method. 3A shows the positional relationship among the eyeball E, the infrared light emitting LED 40, and the photodiodes 42 and 44. FIG. 3B shows the detection principle of the eyeball movement in the vertical direction of the eyeball E. FIG. 3C shows the principle of detecting the eye movement in the horizontal direction of the eyeball E. FIG.

  As shown in FIG. 3, in the scleral reflection method, the eyeball E is irradiated with weak infrared light by the infrared light emitting LED 40, and the reflected light reflected by the eyeball E is reflected by the photodiode 42, 44. Due to the difference in reflectance between the cornea (black eye) C and the sclera (white eye) S, the photodiodes 42 and 44 detect a light amount value (a light amount value of reflected light) according to the movement direction of the eyeball E.

  That is, in the scleral reflection method, as shown in FIG. 3B, the eye movement in the vertical direction can be found by adding the light quantity values detected by the photodiodes 42 and 44. That is, if the value is greater than or equal to a certain value, the sclera (white eye) S portion has increased, indicating that the eyeball E is facing upward, and if the value is less than a certain value, This is a case where the portion of the cornea (black eye) C is increased, which indicates that the eyeball E faces downward.

  Further, as shown in FIG. 3C, the eye movement in the horizontal direction can be determined by subtracting the light amount value detected by the photodiodes 42 and 44 (taking the difference). For example, the value of the amount of light detected by the photodiode 44 arranged on the right side is subtracted from the value of the amount of light detected by the photodiode 42 arranged on the left side toward the eyeball E. Black eye) C is directed to the right, and if it is negative, the cornea (black eye) is directed to the left.

  Here, “offset”, “gain”, “crosstalk”, and “linearity”, which are typical measurement errors that occur when the scleral reflection method is used, will be described with reference to FIG. 4A shows “offset”, FIG. 4B shows “gain”, FIG. 4C shows “crosstalk”, and FIG. 4D shows “linearity”. is there. 4A to 4D, the white circle “◯” indicates the actual (accurate) gaze point of the subject.

  As shown in FIG. 4A, the “offset” indicates a translational displacement, and when there is no measurement error called “gain”, “crosstalk”, and “linearity”, the pattern is a broken line. On the other hand, it is a position shift in the vertical and horizontal directions, or a position shift in the vertical or horizontal direction (horizontal offset, vertical offset).

  As shown in FIG. 4B, the “gain” includes the measured (detected) horizontal and vertical eye movement data components and the eye movement data components to be actually output when correctly calibrated. When the two components are different, the shift is reduced or enlarged (horizontal gain, vertical gain).

  As shown in FIG. 4 (c), “crosstalk” is a measurement error (frequently seen in the scleral reflection method that measures (detects) eye movement data using the difference in the amount of reflected light of infrared light). A change in the amount of reflected light caused by eye movement in the horizontal direction leaks to the sensors (photodiodes 42 and 44) that detect the amount of reflected light in the vertical direction, or the amount of reflected light generated by the eye movement in the vertical direction. This change is caused by leaking to the sensors (photodiodes 42 and 44) that detect the amount of reflected light in the horizontal direction and receiving light (horizontal crosstalk and vertical crosstalk).

  As shown in FIG. 4D, the “linearity” is correctly calibrated with the measured (detected) horizontal and vertical eye movement data components, for example, because the surface of the human eyeball is not a perfect sphere. This is a phenomenon (measurement error) in which the “gain”, which is the amplitude ratio with the eye movement data component to be actually output when it is applied, appears non-uniform due to the eyeball rotation angle. That is, this “linearity” is obtained when the gaze point detected (measured) coincides with the actual gaze point when the central part is viewed and the peripheral part is viewed as shown in FIG. In spite of this, when looking at the area between the center and the periphery, the detected (measured) gazing point does not match the actual gazing point ("shift due to linearity" occurs) ) And expressed as non-uniform gain difference due to the amplitude of eye movement (horizontal linearity, vertical linearity).

Returning to FIG. 1, the description of the configuration of the eye movement data calibration coefficient calculation system A will be continued.
The display device 4 is configured by a general liquid crystal display, a CRT display, or the like, and includes a display screen. The eye movement data calibration graphic pattern output from the eye movement data calibration coefficient calculation device 1 is displayed on the display screen. It is what is displayed. In FIG. 1, the broken line from the display device 4 to the eye movement measuring device 2 shows the eye movement data calibration graphic pattern for the subject whose eye movement is measured by the eye movement measuring device 2. This is shown to imply that.

<Configuration of eye movement data calibration coefficient calculation device>
The configuration of the eye movement data calibration coefficient calculation device 1 will now be described.
The eye movement data calibration coefficient calculation device 1 calculates (calculates) a calibration coefficient for calibrating the eye movement data output from the eye movement measurement device 2, and includes eye movement data calibration graphic data storage means 3, eye movements, and the like. Data calibration graphic pattern generation output means 5, eye movement data acquisition means 7, eye movement data graphic pattern association means 9, correspondence degree calculation means 11, and calibration coefficient calculation means 13 are provided.

  The eye movement data calibration graphic data storage means 3 is constituted by a hard disk or the like, and includes eye movement data calibration graphic data for drawing an eye movement data calibration graphic pattern displayed on the display screen of the display device 4. Position data indicating an area on the display screen (corresponding to a predesignated area) when the eye movement data calibration graphic pattern generated (drawn) from the eye movement data calibration graphic data is displayed on the display screen Are accumulated.

  This eye movement data calibration graphic data storage means 3 converts the stored eye movement data calibration graphic data into eye movement data based on an operation signal input from an operation means (keyboard, mouse, etc.) not shown. In addition to outputting to the calibration graphic pattern generation output means 5, the position data is output to the eye movement data graphic data association means 9. That is, the position data output from the eye movement data calibration graphic data storage means 3 is the current position of the eye movement data calibration graphic pattern displayed on the display screen of the display device 4.

  Further, the eye movement data calibration graphic data storage means 3 displays (moves) the eye movement data calibration graphic pattern in a predetermined area (usually a plurality of areas) of the display screen of the display device 4. A plurality of patterns are stored in the order (order). Hereinafter, the pattern of the display (moving) order will be referred to as a display order pattern. That is, this display order pattern can be said to be a series of a plurality of position data.

  The display order pattern can be changed every time the eye movement data calibration graphic pattern is output. That is, when the eye movement data calibration graphic pattern generated by the eye movement data calibration graphic pattern generation / output means 5 is sequentially output to a plurality of areas designated in advance for the first time and sequentially for the second time. Different display methods can be used for the output and the third output in order.

  In this embodiment, the number of times that the eye movement data calibration graphic pattern is displayed in the same area is three, but the number of times is not limited to this, and each area is considered due to processing in the correspondence calculation means 11 described later. It is sufficient if it is 3 times or more with a slight difference. The display order pattern is output to the eye movement data calibration graphic pattern generation output unit 5 based on an operation signal input from an operation unit (not shown).

  The eye movement data calibration graphic pattern generation / output unit 5 generates (draws) eye movement data calibration graphic patterns based on the eye movement data calibration graphic data output from the eye movement data calibration graphic data storage unit 3. In addition, according to the display order pattern output from the eye movement data calibration graphic data storage means 3, the eye movement data calibration graphic pattern is sequentially applied to a plurality of predetermined areas on the display screen of the display device 4. It outputs (displays) a plurality of times and is provided with graphic pattern display changing means 5a.

  The eye movement data calibration graphic pattern generation / output means 5 is based on the eye movement data calibration graphic data which is a data group in which the coordinates, color, and luminance of the eye movement data calibration graphic pattern are parameterized. 4 A graphic drawing that fits in each area on the display screen, such as a geometric pattern, or a line drawing of a pattern such as an animal or plant, is generated, and the range surrounded by the line drawing is colored, and the brightness of the colored color is adjusted. The eye movement data calibration graphic pattern is generated.

  In this embodiment, in consideration of various modes when displayed on the display device 4, the eye movement data calibration graphic pattern is sequentially generated from the eye movement data calibration graphic data. However, when the data amount of the generated eye movement data calibration graphic pattern itself may be small, the eye movement data calibration graphic pattern itself may be stored and output (displayed).

  The graphic pattern display changing means 5a changes the color or brightness of the whole or part of the eye movement data calibration graphic pattern generated (drawn) by the eye movement data calibration graphic pattern generation output means 5 and changes the size. Change (make similar figure). That is, the figure pattern display changing means 5a keeps the shape of the eye movement data calibration figure pattern as it is, and the eyeball so that the subject pays more attention to the eye movement data calibration figure pattern (so that it can be easily watched). This changes the way (how to display) the graphic pattern for calibrating the exercise data.

The eye movement data calibration graphic pattern generated by the eye movement data calibration graphic pattern generation / output unit 5 is a psychological experiment conducted in advance using highly attractive figures shown in FIGS. 5 (a) to (i). since also reflecting the results, a specific example in FIG. 6 (a) ~ (i) . That is, the eye movement data calibration graphic pattern has attractiveness, and this attractiveness refers to a characteristic that attracts the line of sight of a person (subject).

  FIGS. 6A to 6D show “simple circle”, “triangle”, “square”, and “star” figures, which are examples of eye movement data calibration figure patterns. When these figures are displayed when nothing is displayed on the display screen 4, the possibility that the subject gazes at these figures increases. Moreover, if the color and brightness | luminance of the part (center part) enclosed by these figures are changed with the figure pattern display change means 5a, a test subject will further increase possibility of paying attention to these figures.

  Also, the “triangle”, “square”, and “star” figures shown in FIGS. 6B to 6D are provided with an edge portion (end portion, pointed portion), and the color of the edge portion. By changing the brightness, the possibility that the subject gazes at these figures will be further increased (it will be easier to attract the subject's line of sight).

  FIG. 6E shows an example of a figure pattern for calibrating eye movement data, which is a “Vernier pattern” (here, a Vernier pattern is used because it is a Vernier pattern). This “Vernier pattern” has a delicate notch (recalls the image of four straight lines intersecting at the center), and subjects tend to be interested in the delicate notch. Since there is, there is a high possibility that the subject will pay attention to this figure.

  It should be noted that the “Vernier pattern” shown in FIG. 6 (e) needs to be intentionally viewed with a focus of attention as a result of a psychological experiment as compared to the figures shown in FIGS. 6 (a) to 6 (d). There is a subject who can measure (detect) the eye movement data by displaying (presenting) this “Vernier pattern” on the display device 4 as a figure pattern for eye movement data calibration. A healthy adult who fully understands the purpose of calibration.

  FIG. 6F shows “double circles (two concentric circles)” which is an example of a figure pattern for calibrating eye movement data. For example, if the color or brightness in the circle inside the “double circle” shown in FIG. 6 (f) is changed by the graphic pattern display changing means 5a, the possibility that the subject will pay attention to these figures is even higher. Become.

  6 (g) to 6 (i) show “fish pattern”, “butterfly pattern”, and “cake pattern”, which are examples of eye movement data calibration graphic patterns. The test subject is a figure suitable for not only healthy adults but also infants and dementia patients (a figure with high attractiveness).

  In the “fish pattern” shown in FIG. 6G, when the “fish pattern” is displayed on the display screen of the display device 4, for example, the displayed area transitions (moves). The head of the fish drawn in the “fish pattern” is pointed in the direction of the head and the color and brightness of the fish eyes are changed by the graphic pattern display changing means 5a. It becomes higher (it becomes easier to attract the subject's eyes).

  In the “butterfly pattern” shown in FIG. 6H, for example, when the “butterfly pattern” is displayed on the display screen of the display device 4, for example, the displayed region transitions (moves). The direction of the head of the butterfly drawn in the “butterfly pattern” is turned and the color and brightness of the butterfly wings are changed by the graphic pattern display changing means 5a so that the subject gazes at the figure. The possibility is even higher (it is easier to attract the subject's eyes).

  In the “cake pattern” shown in FIG. 6 (i), when the “cake pattern” is displayed on the display screen of the display device 4, for example, the strawberry of the cake depicted in the “cake pattern” By changing the color and brightness of the pattern by the graphic pattern display changing means 5a, the possibility that the subject gazes at the graphic is further increased (it becomes easier to attract the subject's line of sight).

In addition, eye movement data calibration graphic pattern is, (the higher the attractiveness of) material as long as the material can easily attract the eye of a subject, not good in any thing.

  Further, at the same time as changing the color and brightness of the eye movement data calibration graphic pattern, a beep sound from a speaker (not shown) and vibration from a tactile device (not shown) are given to the subject via the arousing means 15 described later. Thus, the possibility that the subject gazes at the figure is further increased (it becomes easier to attract the subject's line of sight).

  An example of the display order pattern is shown in FIG. As shown in FIG. 7, in one example of the display order pattern, the display screen of the display device 4 is divided into 7 horizontal parts and 5 vertical parts to obtain 35 divided areas (divided display areas, area numbers 11 to 17, 21 to 27, 31 to 37, 41 to 47, 51 to 57), and the eye movement data calibration graphic pattern is sequentially (continuously or discontinuously) along the solid line in FIG. Display). Note that the number of divisions for dividing the display screen of the display device 4 may be any number. If the number of divisions is increased, the accuracy of calibration coefficients obtained later will be improved. In FIG. 7, the eye movement data calibration graphic pattern is indicated by a “simple circle”.

  An example of this display order pattern is eye movement data calibration so that an inverted S-shape is drawn in the divided display areas from area number 11 to area number 57 (11 to 17, 27, 37 to 31, 41, 51 to 57). A graphic pattern is displayed. When the eye movement data calibration graphic pattern is displayed up to the divided display area of area number 57, the eye movement data calibration graphic pattern is displayed again in the divided display area of area number 11, and the eyeballs in the same order. The movement data calibration graphic pattern is displayed (moves). In this embodiment, the eye movement data calibration graphic pattern is displayed (moved) three times (three times) by the same display order pattern. The number of times (frequency) needs to be a plurality of times because of the processing in the correspondence degree calculation means 11 described later, but the number of times is an arbitrary number.

  Here, the relationship between the display speed and the display time of the eye movement data calibration graphic pattern in the display order pattern will be described with reference to FIG. In FIG. 8, the eye movement data calibration graphic pattern is simply displayed as a pattern, and the region (divided display area) is displayed as a target position.

  FIG. 8A is a diagram for explaining an example of the relationship between the display speed and the display time when the eye movement data calibration graphic pattern is smoothly moved (displayed), and FIG. It is a figure explaining an example of the relationship between the display speed and the display time when moving (displaying) while stopping in the area | region (division display area) which displays the data pattern for data calibration.

  When smoothly moving (displaying) the eye movement data calibration graphic pattern shown in FIG. 8A, the graphic pattern can be accurately followed without delaying the movement of the eye movement data calibration graphic pattern. Since the movement speed of the eyeball is around 5 degrees / second due to the nature of the compliant eyeball movement, the movement (display) speed for moving the eye movement data calibration graphic pattern is preferably less than 5 degrees / second. However, as in the divided display areas of area numbers 17, 37, 31, and 51 shown in the example of the display order pattern in FIG. 7, the division in which the direction of the graphic pattern for moving (displaying) the eye movement data calibration changes is changed. In the display area, it is necessary to temporarily stop the figure pattern due to the nature of the eye movement, and it is necessary to stop the figure pattern for at least 200 ms or more due to the nature of the jumping eye movement that occurs at this time. It is preferable to stop the graphic pattern to some extent.

  When moving (displaying) the eye movement data calibration graphic pattern shown in FIG. 8B while stopping in the area (division display area) for displaying the eye movement data calibration graphic pattern, for example, the area for displaying the eye movement data calibration graphic pattern (division) When it is desired to calibrate the eye movement data acquired in the display area), from the nature of the jumping eye movement, at least 200 ms or more in the area (divided display area) for displaying the eye movement data calibration graphic pattern, for example, about 1 second, It is preferable to stop the graphic pattern.

  In any of the cases shown in FIGS. 8A and 8B, when the eye movement data calibration graphic pattern is displayed in the region (division display area) where the eye movement data is desired to be calibrated, Attention to the subject by changing the brightness or presenting auditory information such as a beep sound by a speaker (not shown) and tactile information such as vibration by a tactile device (not shown) such as vibration by the arousing means 15 And the accuracy of calibration of the eye movement data can be improved.

Returning to FIG. 1, the description of the configuration of the eye movement data calibration coefficient calculation device 1 will be continued.
The eye movement data acquisition means 7 acquires eye movement data (horizontal eye movement data and vertical eye movement data) measured (detected and measured) by the eye movement measuring device 2.

  Here, FIG. 9 shows an example of the acquired eye movement data (so-called raw data before calibration [raw data before calibration]). The eye movement data example shown in FIG. 9 is measured by the eye movement measuring device 2 when the eye movement data calibration graphic pattern is displayed (moves) in the display order pattern shown in FIG. It is a thing.

  An example of the eye movement data is an eyeball that the subject moves faithfully without making a mistake in visual recognition of the eye movement data calibration graphic pattern (without losing line of sight to the eye movement data calibration graphic pattern). The figure follows the movement data calibration graphic pattern (target).

  Usually, when the eye movement of the subject is measured by the eye movement measuring device 2, even if the subject is an adult healthy person, the eye movement data is lost several times (some places) by blinking or the like. Occurs. Further, the movement of the subject's line of sight reproduced from the eye movement data is displayed as a thick solid line in FIG. As indicated by this bold solid line, the gain (amplitude) is insufficient in both the horizontal and vertical directions, and the movement of the line of sight is irregular, which is different from the movement of the figure pattern for eye movement data calibration. It has become.

Incidentally, in FIG. _{9, (x 11, y 11} ), (x 34, y 34), (x 57, y 57) and the coordinates of the subscript position (point) that are filled (11, 34, 57) corresponds to the area number of the divided display area where the eye movement data calibration graphic pattern is actually displayed, and the line of sight of the subject indicated by the measured eye movement data is originally divided by this area number. This indicates that it should match the display area.

  That is, with this eye movement data (raw data before calibration), when a certain divided display area is observed, it cannot be accurately determined (determined) whether or not the subject has actually seen the divided display area. That is, it can be assumed that the subject may have seen another divided display area where the eye movement data calibration graphic pattern is not displayed, and that blinking or forgetfulness (caused or carelessness) occurred. .

  For this reason, in the eye movement data calibration coefficient computing device 1, the eye movement data calibration graphic pattern is generated by the eye movement data calibration graphic pattern generation output means 5 a plurality of times (several times) in the same divided display area (region). By displaying, the number of eyes | visual_axis shown by the eye movement data which fits in the said division | segmentation display area is increased, and the precision of the measured eye movement data is improved.

Returning to FIG. 1, the description of the configuration of the eye movement data calibration coefficient calculation device 1 will be continued.
The eye movement data graphic pattern association means 9 is based on the eye movement data acquired by the eye movement data acquisition means 7 and the eye movement data calibration graphic data stored in the eye movement data calibration graphic data storage means 3. The eye movement data and the eye movement data calibration graphic pattern are associated with each other based on position data indicating a region (divided display area) when the movement data calibration graphic pattern is generated and output (displayed).

  That is, in this eye movement data graphic pattern association means 9, the subject's line of sight (gaze position) indicated by the eye movement data and the current position of the region where the eye movement data calibration graphic pattern is displayed are obtained. It will be associated. The associated result is output to the correspondence degree calculation means 11.

  The degree-of-correspondence calculating means 11 calculates the degree of correspondence between the eye movement data associated with the eye movement data figure pattern association means 9 and the figure pattern for calibrating eye movement data. Means 11b. The degree of correspondence calculated by the degree-of-correspondence calculating means 11 is a match indicating how much the line of sight indicated by the eye movement data and the area where the eye movement data calibration graphic pattern is displayed match. It is a degree.

  The data comparison unit 11a compares the number of times that this area and the line of sight indicated by the eye movement data match with the number of times that does not match in the area (divided display area) where the eye movement data calibration graphic pattern is displayed. Is.

  As a result of the comparison by the data comparison unit 11a, the data judgment unit 11b determines that the number of times that this area matches the line of sight indicated by the eye movement data is greater than the number of times that the eye movement data does not match. It is determined that the eye movement data in the area where is displayed has been obtained.

  That is, when it is determined whether the eye movement data (line of sight) measured (detected) by the eye movement measuring device 2 is within each area (divided display area), the eye movement is moved to the area (divided display area). Assuming that the eye movement data is obtained when the data calibration graphic pattern is displayed (instantaneously), the display (presentation) of the eye movement data calibration graphic pattern is performed by the eye movement data calibration graphic pattern generation output means 5. Since it is repeated sequentially, when the eye movement data calibration graphic pattern is displayed in the area (divided display area) (instantaneous), a plurality of eye movement data showing almost the same value is collected (number of coincidence) > Number of times of mismatch.

  Note that the accuracy (calibration accuracy) of the calibration coefficient finally calculated by the calibration coefficient calculation means 13 is determined depending on which range of the eye movement data is regarded as substantially the same value. For example, if only the number of eye movement data showing exactly the same value is handled, the calibration accuracy will improve, but the central fovea, which is said to have the best visual acuity among the retinas of human eyes, ) Alone has a visual angle of 1 degree, and every time an eye movement data calibration graphic pattern is watched, there is a possibility that an error of ± 1 degree may occur, and the human visual physiology Due to its unique nature, it is unreasonable to handle only the number of eye movement data showing exactly the same value. For this reason, it can be said that it is practically preferable to observe whether the eye movement data matches or does not match (matching degree) with an error of about ± 2 degrees as an allowable range.

  FIG. 10 shows an example in which the degree of coincidence of the eye movement data of the subject when gazing at the eye movement data calibration graphic pattern displayed on the display screen of the display device 4 in this way is shown.

  As shown in FIG. 10, the eye movement data calibration graphic pattern is displayed three times in a predetermined area (divided display area) of the display screen of the display device 4 according to the same display order pattern.

  If the eye movement data calibration graphic pattern is displayed (moved) with the same display order pattern in the same region too many times, the subject may not see the eye movement data calibration graphic pattern. It can be said that about three times is preferable, and the display order pattern may be changed as appropriate. However, in order to calculate (calculate) the calibration coefficient of the eye movement data, it is necessary to display the eye movement data calibration graphic pattern multiple times in each area even if the display order pattern is changed.

  Further, in FIG. 10, the eye movement data calibration graphic pattern is displayed three times in the divided display area of the display screen of the display device 4, and the divided display area matches the eye movement data (line of sight). In this case, “○” is displayed, and when it does not match (when it does not match), “●” is displayed, and in each divided display area, two or more “○” are obtained. Indicates that the eye movement data effective for calculating the calibration coefficient is obtained.

  Incidentally, in the calibration by the conventional “5-point method” (calculation of the calibration coefficient), it is desirable to obtain effective eye movement data at five gaze positions at the center of the top, bottom, left, and right. However, since it can be complemented arithmetically by the calibration coefficient calculation means 13 to be described later, eye movement data matching the positions of the upper, lower, left and right five points is not necessarily required. In the example shown in FIG. 10, the area number 14, the area number 54, the area number 31, the area number 37, and the area number 34 are divided and displayed corresponding to the five gaze positions necessary for the calibration by the “5-point method”. Since the eye movement data of the area is acquired as valid eye movement data, the calibration coefficient can be calculated by the “5-point method” without complementation.

  Furthermore, the application to the calibration (calculation of the calibration coefficient) by the conventional “9-point method” in which the accuracy of the measured eye movement data is higher corresponds to the 9 gaze positions in the “9-point method”. In the divided display areas (area numbers 11, 14, 17, 31, 34, 37, 51, 54, 57) to be obtained, valid eye movement data is acquired in the divided display areas of area number 17 and area number 51. No (missing). Moreover, effective eye movement data is not acquired (is missing) in the (adjacent) divided display areas in the vicinity of the divided display areas of area number 17 and area number 51.

  In such a case, the display order pattern shown in FIG. 11, that is, the division of the area number 51 from which the effective eye movement data is not acquired from the divided display area of the area number 17 from which the effective eye movement data is not acquired. The eye movement data is measured by the eye movement measuring apparatus 2 so that the eye movement data calibration graphic pattern is displayed (moved) to the display area.

  Thus, even if the subject fails to gaze at the eye movement data calibration graphic pattern (as a result, effective eye movement data has not been acquired), the display order pattern is changed and effective eye movement data is obtained. It is possible to display the eye movement data calibration graphic pattern again only in the divided display areas for which the eye movement has not been acquired, reducing the burden on the subject, and easily calculating the calibration coefficient of the eye movement data (calibration work). Can be implemented.

Returning to FIG. 1 again, the description of the configuration of the eye movement data calibration coefficient computing device 1 will be continued.
The calibration coefficient calculation unit 13 calculates a calibration coefficient for calibrating the eye movement data based on the correspondence (matching degree) calculated by the correspondence calculation unit 11. The calibration coefficient calculated by the calibration coefficient calculating means 13 includes a horizontal gain calibration coefficient, a vertical gain calibration coefficient, a horizontal crosstalk calibration coefficient, a vertical crosstalk calibration coefficient, a horizontal offset calibration coefficient, and a vertical offset calibration. Coefficient, horizontal linearity calibration coefficient, and vertical linearity calibration coefficient.

  The calibration coefficient calculating means 13 employs a method using simultaneous equations and a method for sequentially obtaining these calibration coefficients.

  Here, for convenience, a description will be given in detail of the eye movement data measured (detected) by the eye movement measuring apparatus 2 with a calibration coefficient added (calibrated eye movement data), and then the calculation of the calibration coefficient. A method of (calculation) will be described.

The horizontal gain, which is a calibration coefficient calculated by the eye movement data calibration coefficient calculation apparatus 1 as horizontal eye movement data X _i and vertical eye movement data Y _i measured (detected) by the eye movement measurement apparatus 2. Calibration factor a ₁₁ , vertical gain calibration factor a ₂₂ , horizontal crosstalk calibration factor a ₁₂ , vertical crosstalk calibration factor a ₂₁ , horizontal offset calibration factor b ₁ , vertical offset calibration factor Is b ₂ , the eye movement data X ′ _{i in} the horizontal direction and the eye movement data Y in the vertical direction are calibrated (corrected) for the i th sample (graphic data for eye movement data calibration). ′ _I is expressed by the following formulas (1) and (2).

_{X 'i = a 11 · X} i + a 12 · Y i + b 1 ··· Equation (1)
Y ′ _i = a ₂₁ · X _i + a ₂₂ · Y _i + b ₂ (2)

At this time, the horizontal linearity calibration function for calibrating the horizontal linearity is f (X ′ _i ), the vertical linearity calibration function for calibrating the vertical linearity is g (Y ′ _i ), the horizontal linearity calibration coefficient is c _j , and the vertical linearity is Where d _j is the calibration coefficient of the horizontal eye movement data X ″ _i taking into account the horizontal linearity calibration, and the vertical eye movement data Y ″ _i taking into account the vertical linearity calibration is expressed by the following formula ( 3) and is expressed by Equation (4).

X ″ _i = f (X ′ _i ) = Σc _j · X ′ _i ^j (3)
Y ″ _i = g (Y ′ _i ) = Σd _j · Y ′ _i ^j (4)
J = 1 to n (where n = 3)

Here, n = 3 is a third-order equation that is often used as a linearity calibration function f (X ′ _i ), g (Y ′ _i ) for eye movement calibration (correction) in the scleral reflection method. Because.

Then, when the calibration coefficient is calculated by the calibration coefficient calculation means 13 using simultaneous equations, using this cubic calibration function, 12 variables (a ₁₁ , a ₁₂ , a ₂₁ , a _{22 are used.} , B ₁ , b ₂ , c _j (j = 1 to 3), d _j (j = 1 to 3)), two equations in the horizontal direction and the vertical direction can be obtained for one gaze point. Thus, these six variables can be calculated (calculated) by using six gaze points.

Further, in the method of sequentially obtaining by the calibration coefficient calculating means 13, all the variables (12 variables) are sequentially changed, and each calibration coefficient is obtained so that the least square error is minimized.
Even when the corneal reflection method is used for eye movement detection, each calibration is performed in substantially the same manner, except that there are no crosstalk items (calibration coefficient a ₁₂ for horizontal crosstalk, calibration coefficient a _{21 for} vertical crosstalk). Coefficients can be calculated (calculated).

  Furthermore, FIG. 12 shows an example of detection of the line of sight indicated by the eye movement data in which the calibration coefficient is added to the eye movement data, that is, the calibration work is completed.

  As shown on the left side of FIG. 12, in the example in which the subject looks at the jockey and the racehorse, the eye movement is looking at an important place from the jockey's face to the eyes, toes, and hips of the racehorse. It can be seen that (eye movement) is reproduced. In addition, as shown on the right side of FIG. 12, in the example where the subject looks at the face of the woman, the eye movement (when looking at the person's face) that is accurately viewed from the eye of the woman's face to the mouth. It can be seen that typical eye movements are reproduced.

Returning to FIG. 1 again, the description of the configuration of the eye movement data calibration coefficient computing device 1 will be continued.
When the eye movement data calibration graphic pattern is displayed in the specified region (divided display area), the awakening means 15 sends a control signal to the subject to the external device (notification) that the display is performed. Output to a speaker or a tactile device (not shown).

  In other words, this awakening means 15 provides the subject with a specified region (a region where eye movement data necessary for calibration is to be acquired in particular (in the case of the five-point method, area numbers 14, 31, 34, 37, 54 of the divided display areas). )) To notify that the figure pattern for eye movement data calibration is displayed, so that the specified area can be accurately observed, and senses other than human vision (auditory, skin sensation, olfaction, A control signal that activates an external device that affects the taste is output. In addition, although a speaker and a tactile device are illustrated here as external devices, as long as they can be incorporated into the eye movement data calibration coefficient calculation system A, they may act on olfaction and taste.

  According to the eye movement data calibration coefficient computing device 1, the eye movement data calibration graphic pattern generation output means 5 uses the eye movement data calibration graphic data accumulated in the eye movement data calibration graphic data accumulation means 3 to generate an eyeball. Based on the position data indicating the display position when the movement data calibration graphic pattern is generated and the eye movement data calibration graphic pattern is displayed on the display screen of the display device 4, the display screen of the display device 4 is designated in advance. Output multiple times to multiple areas. Then, when the eye movement data calibration graphic pattern is displayed on the display device 4 by the eye movement data acquisition means 7, the eye movement data output from the eye movement measurement device 2 is acquired, and the eye movement data graphic pattern correspondence is obtained. Based on the eye movement data and the position data when the subject gazes at the eye movement data calibration graphic pattern displayed on the display device 4 by the attaching means 9, the eye movement data and the eye movement data calibration graphic pattern are obtained. It is associated. Thereafter, the correspondence degree computing means 11 computes the correspondence degree (matching degree) between the eye movement data associated with the eye movement data figure pattern association means 9 and the eye movement data calibration figure pattern, and the calibration coefficient computing means. 13 calculates (calculates) a calibration coefficient for calibrating the eye movement data with respect to the figure pattern for eye movement data calibration based on the calculated correspondence. Therefore, the eye movement data (eye position) of the subject whose eye movement is measured and the position (area, viewing position) of the gazing point that the subject is actually gazing, that is, the eye movement data calibration graphic pattern Since the eye movement data can be calibrated based on the calibration coefficient, it can be used for subjects with advanced symptoms such as Alzheimer's disease, dementia, infants, and other subjects who are difficult to give verbal instructions. On the other hand, there is no need to give verbal instructions, and calibration work can be performed when measuring eye movements.

  Further, according to the eye movement data calibration coefficient calculation device 1, the eye movement data calibration graphic pattern is output to the display device 4 by the eye movement data calibration graphic data generation / output unit 5, and the display screen of the display device 4 is displayed. The eye movement data calibration graphic pattern is displayed in a predesignated area. In the region where the eye movement data calibration graphic pattern is output by the data comparison unit 11a, the number of times that the region and the line of sight (gaze point) indicated by the eye movement data match each other and the number of times that they do not match. When the number of coincidence is large, that is, when the number of gazing at the eye movement data calibration graphic pattern exceeds the number of not gazing, the data determination unit 11b causes the eye movement data to be compared. It is assumed that eye movement data (effective eye movement data) corresponding to the calibration graphic data is obtained. For this reason, the correspondence between the eye movement data and the eye movement data calibration graphic pattern can be made accurate.

  Furthermore, according to the eye movement data calibration coefficient calculation device 1, the eye movement data calibration is performed by the data judgment means 11b in the region where the figure pattern display changing means 5a outputs, for example, the eye movement data calibration figure pattern. The method (display method) of displaying the eye movement data calibration graphic pattern to be displayed in an area where it is not determined that the eye movement data corresponding to the graphic pattern for use has been obtained is changed. For example, the shape of the eye movement data calibration graphic pattern can be left as it is, and the color and brightness can be changed, or the calibration graphic pattern can be changed to be displayed only in the area where the eye movement data is not acquired. For this reason, if the display of the eye movement data calibration graphic pattern is changed for an area for which eye movement data has not been acquired, the eye movement data can be acquired efficiently.

  Furthermore, according to the eye movement data calibration coefficient computing device 1, when the eye movement data calibration graphic pattern is output to the specific area by the awakening means 15, the eye movement data calibration graphic pattern is given to the subject. Is output (notified), for example, to a subject such as a dementia patient or an infant who has difficulty in gazing even though the eye movement data calibration graphic pattern has high attractiveness. On the other hand, attention can be drawn so as to watch the eye movement data calibration graphic pattern.

Next, the operation of the eye movement data calibration coefficient computing device 1 will be described with reference to the flowchart shown in FIG. 13 (see FIG. 1 as appropriate).
First, the eye movement data calibration coefficient computing device 1 uses the eye movement data calibration graphic data storage means 3 as the eye movement data calibration graphic data used when measuring the eye movement based on the operation signal input by the experimenter. (Step S1). Subsequently, the eye movement data calibration coefficient calculation device 1 displays the eye movement data calibration graphic pattern generated from the eye movement data calibration graphic data on the display screen of the display device 4 based on the operation signal input by the experimenter. The order of display on top is specified, that is, the display method (display order pattern) is selected (step S2).

  Then, the eye movement data calibration coefficient calculation device 1 generates the eye movement data calibration graphic pattern from the eye movement data calibration graphic data by the eye movement data calibration graphic pattern generation output means 5 and generates it according to the display order pattern. The region (divided display area) in which the eye movement data calibration graphic pattern is output (displayed) is displayed while being moved (step S3).

  Then, the eye movement measurement device 2 measures the eye movement of the subject and outputs the eye movement data to the eye movement data calibration coefficient calculation device 1. The eye movement data calibration coefficient computing device 1 acquires the eye movement data by the eye movement data acquisition means 7 (step S4), and the eye movement data and eye movement data calibration means 9 by the eye movement data graphic data association means 9. The figure pattern position data is associated (step S5).

  Then, the eye movement data calibration coefficient computing device 1 uses the correspondence degree computing means 11 to associate a predetermined area (divided display area) of the display screen of the display device 4 with the line of sight (gaze point) indicated by the eye movement data. When calculating the degree (degree of coincidence), the data judging means 11b determines whether or not necessary eye movement data has been prepared (in each divided display area, the number of times that the area and the line of sight coincide with each other) It is determined whether or not (step S6).

  When it is not determined that the necessary eye movement data has been prepared (step S6, No), the display order pattern is changed (step S7), and the process returns to step S3. When it is determined that the necessary eye movement data has been prepared (step S6, Yes), the eye movement data calibration coefficient calculation device 1 calculates the calibration coefficient by the calibration coefficient calculation means 13 (step S8).

  As mentioned above, although embodiment of this invention was described, this invention is not limited to the said embodiment. For example, in the present embodiment, the eye movement data calibration coefficient calculation system A (including the eye movement data calibration coefficient calculation apparatus 1) has been described. However, the processing of each component of the eye movement data calibration coefficient calculation apparatus 1 is general-purpose or It can also be regarded as an eye movement data calibration coefficient calculation program described in a special computer language, and can also be regarded as an eye movement data calibration coefficient calculation method in which the processing of each component is regarded as one process. In these cases, the same effect as the eye movement data calibration coefficient calculation device 1 can be obtained.

1 is a block diagram of an eye movement data calibration coefficient calculation system (including an eye movement data calibration coefficient calculation device) according to an embodiment of the present invention. It is explanatory drawing explaining the principle of the cornea reflection system. It is explanatory drawing explaining the principle of the scleral reflection system. FIG. 4A shows a measurement error called “offset”. FIG. 4B is a diagram showing a measurement error called “gain”. FIG. 4C shows a measurement error called “crosstalk”. FIG. 4D is a diagram showing a measurement error called “linearity”. (A)-(i) It is the figure which showed the example of the figure with high attraction. (A)-(i) It is the figure which showed an example of the figure pattern for eye movement data calibration. It is the figure which showed an example of the display order pattern. It is the figure which showed the relationship between the display speed and display time of the figure pattern for eye movement data calibration in a display order pattern. It is the figure which showed an example of eye movement data (what is called raw data before calibration [raw data before calibration]). It is the figure which showed the example investigated about the coincidence degree of the test subject's eye movement data at the time of gazing at the figure pattern for eye movement data calibration displayed on the display screen of a display apparatus. It is the figure which showed an example of the display order pattern when effective eye movement data are not obtained. It is the figure which showed the example of a detection of the eyes | visual_axis shown by the eye movement data which calibration work was completed. It is the flowchart explaining operation | movement of the eye movement data calibration coefficient calculating apparatus shown in FIG. It is the figure which showed typically the board for calibration (calibration board) which calibrates the eye movement data called "5-point method". It is the figure which showed the target of "5-point method", "9-point method", and "13-point method". It is the figure which showed the evaluation result with respect to the measurement error mentioned later by "5-point method", "9-point method", and "13-point method".

Explanation of symbols

DESCRIPTION OF SYMBOLS 1 Eye movement data calibration coefficient calculating apparatus 2 Eye movement measuring apparatus 3 Graphic data storage means for eye movement data calibration 5 Graphic pattern generation output means for eye movement data calibration 5a Graphic pattern display change means 7 Eye movement data acquisition means 9 Eye movement data Graphic data correspondence means 11 Correspondence degree calculation means 11a Data comparison means 11b Data judgment means 13 Calibration coefficient calculation means 15 Arousing means

Claims (4)

An eye movement data calibration coefficient calculation device for calculating a calibration coefficient for calibrating eye movement data output from an eye movement measurement device for measuring eye movement,
The eye movement data calibration graphic data for displaying on the display device the eye movement data calibration graphic pattern shown to the subject whose eye movement data is measured by the eye movement measurement device, and the eye movement data calibration graphic pattern. Eye movement data calibration graphic data storage means for storing position data indicating a display position when displayed on the display screen of the display device;
The eye movement data calibration graphic pattern is generated from the eye movement data calibration graphic data stored in the eye movement data calibration graphic data storage means, and the display screen of the display device is changed based on the position data. among areas formed by dividing into a matrix, and eye movement data calibration figure pattern generation output means for outputs sequential multiple times in a plurality of said area designated Me pre,
When the eye movement data calibration graphic pattern generated by the eye movement data calibration graphic pattern generation / output unit and output to the region is displayed on the display device, eye movement data is acquired from the eye movement measurement device. Eye movement data acquisition means;
Based on the eye movement data acquired by the eye movement data acquisition means and the position data of the eye movement data calibration graphic pattern displayed on the display device, the eye movement data and the eye movement data calibration graphic pattern Eye movement data figure pattern association means for associating
Correspondence degree computing means for computing the degree of correspondence between the eye movement data associated with the eye movement data figure pattern association means and the eye movement data calibration figure pattern;
Calibration coefficient calculation means for calculating a calibration coefficient for calibrating the eye movement data for the eye movement data calibration graphic pattern based on the correspondence calculated by the correspondence degree calculation means;
Equipped with a,
The degree-of-correspondence calculation means compares the number of times that the region and the line of sight indicated by the eye movement data match each other and the number of times the eye movement data does not match in the region where the eye movement data calibration graphic pattern is output Data comparison means;
Data determination for determining that the eye movement data corresponding to the eye movement data calibration graphic pattern has been obtained from the region when the number of matches is greater than the number of mismatches as a result of comparison by the data comparison means Means ,
An eye movement data calibration coefficient calculation device comprising: 2. The eye movement data calibration according to claim 1, wherein the eye movement data calibration graphic pattern generation output means includes a graphic pattern display changing means for changing a way of displaying the eye movement data calibration graphic pattern. 3. Coefficient arithmetic unit. In the area, when the eye movement data calibration graphic pattern is output to the specified area, the subject is provided with an urging means for urging the subject to output the eye movement data calibration graphic pattern. The eye movement data calibration coefficient calculation device according to claim 1 or 2, characterized in that An apparatus for calculating a calibration coefficient for calibrating eye movement data output from an eye movement measuring apparatus for measuring eye movement;
The eye movement data calibration graphic pattern for displaying on the display device the eye movement data calibration graphic pattern shown to the subject whose eye movement data is measured by the eye movement measurement device. Based on the position data indicating the display position when the eye movement data calibration graphic pattern is displayed on the display screen of the display device, and the display screen of the display device is divided into a matrix among them, the eye movement data calibration figure pattern generation output means for the eye movement data calibration graphic pattern to a plurality of said area designated Me pre is sequential multiple outputs,
When the eye movement data calibration graphic pattern generated by the eye movement data calibration graphic pattern generation / output unit and output to the region is displayed on the display device, eye movement data is acquired from the eye movement measurement device. Eye movement data acquisition means,
Based on the eye movement data acquired by the eye movement data acquisition means and the position data of the eye movement data calibration graphic pattern displayed on the display device, the eye movement data and the eye movement data calibration graphic pattern Means for associating eye movement data with graphic patterns,
Correspondence degree computing means for computing the degree of correspondence between the eye movement data associated with the eye movement data figure pattern association means and the eye movement data calibration figure pattern;
Calibration coefficient computing means for computing a calibration coefficient for calibrating the eye movement data for the eye movement data calibration graphic pattern based on the degree of correspondence calculated by the correspondence degree calculating means;
To function as,
In the region where the eye movement data calibration graphic pattern is output, the correspondence degree calculating means compares the number of times that the region and the line of sight indicated by the eye movement data are matched with the number of times the eye movement data is not matched. as a result, when the number of times the matched is greater than the number that did not match, and wherein said child determines that eye movement data is obtained corresponding to the eye movement data calibration graphic pattern from the area eyeball Exercise data calibration coefficient calculation program.

Images