JPH0556736B2 - - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION [Field of Industrial Application] The present invention relates to an eye movement detection device, and particularly to an eye movement detection device that can accurately detect eye movement without causing pain to a subject.

[Conventional technology]

By observing the movement of a person's eyeballs, it is possible to know the psychology and health condition of that person, so eye movement detection devices are used in medicine, psychology, education,
It is used in a wide range of fields, from sports to industry.

Conventional devices for detecting eyeball movements can be roughly divided into: eyemark recorders that use the movement of eyemarks, devices that use changes in the amount of reflected light from the cornea, and EOG (electro oculography) that uses the potential difference in the skin around the eyeballs. There are three known types: , and, but the eye mark recorder is currently considered the most suitable for accurately detecting eye movements without causing pain to the subject.

The eye mark recorder includes means for forming an image of the eye mark on the eyeball, a shutter, an imaging section, a third register for storing an image signal of the eye mark converted by the imaging section, and a signal stored in the third register. The image forming apparatus includes an extracting means for extracting the coordinate points of the eye mark based on the image signal, and a calculating means for calculating the rotation angle and rotation direction of the eye mark based on the coordinate points of the eye mark.

Next, a conventional eye movement detection device will be described in detail with reference to the drawings.

FIG. 6 is a block diagram showing a conventional example.

The eye movement detection device shown in FIG. 6 includes a means 9 for forming an image of an eye mark on an eyeball, a shutter 1, an imaging section 2, and a third register for storing an image signal of the eye mark converted by the imaging section 2. 3C and the third
an extraction means 10 for extracting the coordinate points of the eye mark based on the image signal stored in the register 3C; and a calculation means 11 for calculating the rotation angle and rotation direction of the eye mark based on the coordinate points of the eye mark. I'm here.

Here, means 9 for forming an image of the eye mark on the eyeball.
An eye mark g is formed on the eyeball by the incident light f on the eyeball from
When the image is formed, the reflected light h from the eyeball is reflected as if the eye mark g were the light source, and passes through the shutter 1 and enters the imaging section 2. Next, after converting the reflected light h into an eye mark image signal i in the imaging unit 2,
It is stored in the third register 3C.

Next, after extracting the coordinate point j of the rear eye mark g by the extracting means 10 for extracting the coordinate point of the eye mark g based on the image signal i of the eye mark g stored in the third register 3C, the extracted eye mark As a means 11 for calculating the rotation angle and rotation direction of the eye mark g based on the coordinate point j of the eye mark g,
The rotation direction β 1 is calculated from the equation (2) using the equation ( ₁ ) that is established from the geometrical relationship between the coordinate point j and the rotation angle θ ₁ .

θ ₁ ≒sin ^-1 (√ ₅ ² + ₅ ² /m) ...(1) β ₁ = sin ^-1 (X ₅ /√ ₅ ² + ₅ ² ) ...(2) [The problem that the invention attempts to solve Problem] The conventional eye movement detection device described above calculates the rotation angle θ ₁ of the eye mark g instead of the eyeball rotation angle θ, and there is a geometric proportional relationship between the two only up to about 15°. does not exist. In general, the maximum eyeball rotation angle for humans is approximately 30° left and right, approximately 30° upward, and approximately 75° downward.
Conventional detection devices had the disadvantage that the detection range was extremely narrow.

Furthermore, since the dimensions of the eyeball differ depending on race, gender, and age, when formula (1) is used, the distance m from the center of eyeball rotation to the center of corneal curvature for each subject is There was a drawback that the method had to be determined in advance.

In addition, the distance m from the center of rotation of the eyeball to the center of corneal curvature cannot be directly measured like the size of a fistula, but is calculated indirectly using various devices, so there is a large error. , the formula using this value
In (1), an error occurs between the actual rotation angle θ ₁ of the eye mark g and the value of equation (1) even within the proportional range of the rotation angle θ ₁ of the eye mark g and the rotation angle θ of the eyeball. There was a drawback that an error occurred between the angle of rotation θ of the eyeball.

[Means for solving problems]

The eye movement detection device of the present invention includes: a shutter, an imaging section, a first register that stores an image signal converted by the imaging section, and extracts contour data based on the image signal stored in the first register. extraction means; first calculation means for calculating center coordinates of the contour based on the contour data; second calculation means for calculating the area within the contour based on the contour data; and the second calculation means. a switch for selecting an area to be stored in a register from among the areas calculated by; a second register for storing the contour area selected by the switch; and a second register for storing the contour area stored in the second register and the second
and third calculation means for calculating the rotation angle and rotation direction of the eyeball based on the contour area that is not stored in the register.

〔Example〕

Next, embodiments of the present invention will be described in detail with reference to the drawings.

FIG. 1 is a block diagram showing one embodiment of the present invention.

The eye movement detection device shown in FIG. 1 includes: a shutter 1, an imaging section 2, a first register 3a that stores an image signal converted by the imaging section 2, and an image signal stored in the first register 3a. extraction means 4 for extracting contour data based on the contour data; first calculation means 5 for calculating the center coordinates of the contour based on the contour data; and second calculation means 5 for calculating the area within the contour based on the contour data. means 6; a switch 7 for selecting an area to be stored in the register from among the areas calculated by the second calculation means 6; and a second register 3b for storing the contour area selected by the switch 7. , the contour area stored in the second register and the second
The third step calculates the rotation angle and rotation direction of the eyeball based on the contour area that is not stored in the register 3b.
calculation means 8.

Next, the functions, operations, etc. of the configuration shown in FIG. 1 will be explained in order. The reflected light a on the eyeball is transmitted to the imaging section 2 through the shutter 1. The imaging unit 2 converts the reflected light a into an image signal b, and the first register 3 converts the reflected light a into an image signal b.
Store it in a. Next, the contour of the tunnel is extracted by the extraction means 4 which extracts contour data based on the image signal b stored in the first register 3a. As an extraction method, the contour of the tunnel is visualized on a monitor, and the contour data c of the tunnel is extracted from the difference in brightness. The extracted canal contour looks like the conceptual diagram shown in FIG. 2, which shows the relationship between the rotation of the eyeball and the flattening of the canal contour. That is, when observed from the emmetropic direction of the eyeball, the contour of the foramen gradually changes from a circular shape in emmetropic vision to a flattened shape as the eyeball rotates. The present invention determines the eyeball movement from the flatness of the foramen outline and the moved position.

Based on the extracted tunnel contour data c, the first
The calculation means 5 calculates the center coordinates of the canal contour. FIG. 3 is a conceptual diagram showing the first calculation means 5, in which the process of checking the color tone one dot at a time in parallel to the The coordinates p (X ₁ , y ₁ ) and q (X ₂ , y ₂ ) of the two points from which the dots were extracted are determined. These two points have the same slope and are tangent points of a straight line that touches the tunnel contour c, so the midpoint of the line segment pq (X ₃ , y ₃ ) is the center point d (X ₃ , y ₃ ) of the tunnel contour c.
matches. That is, it can be calculated using the following formula.

_{X 3 = (} X _{1 +}
However, the result is the same even if the point of contact with two straight lines parallel to the y-axis or having an arbitrary slope is found, and this data is input to the third calculation means 8.

On the other hand, the extracted canal contour data c is also input to the second calculation means 6, and as shown in the conceptual diagram of the canal area calculation means shown in FIG. The process of checking the color tone and calculating the total number of dots between the two points u and v where the color tone changes is carried out from y ₁ to y ₂ one dot at a time, and the total number of dots inside the hole surrounded by the outline of the hole is calculated. By calculating this, the canal area e is calculated, and the selection switch 7 selects the canal area S ₀ in normal vision and stores it in the second register 3b.

Next, the foramen area S in normal vision stored in the second register 3b ₀ The foramen area S of the eyeball in any state
(e=S+S ₀ ) is the rotation angle θ of the eyeball and the rotation direction β
is inputted to the third calculation means 8 which calculates.

The following equation, which is established from the conceptual diagram shown in FIG. 5 showing the geometrical relationship between the eyeball rotation angle and the amount of flattening of the canal contour, is used.

S ₀ = πr ² ...(4) S = πr ² cosθ ...(5) Therefore, θ=cos ^-1 S/S ₀ ...(6) Here, r is the hole radius at the hour. be.

From equation (5), it can be seen that the eyeball rotation angle θ is proportional to S/S ₀ , which is one parameter representing the amount of flattening of the foramen contour. Further, the rotation direction β can be calculated using the center coordinates d(X ₃ , y ₃ ) of the canal contour using the following equation.

β＝sin ^-1 (X ₃ /√ ₃ ² + ₃ ² ) ...(7) Here, it should be noted that the eyeball rotation angle θ is calculated using the center of the foramen outline (X ₃ , y ₃ ), and This can also be determined from the formula.

θ≒sin ^-1 (√ ₃ ² + ₃ ² /m′) ...(8) Here, m′ is the distance from the center of rotation of the eyeball to the foramen. If the eyeball rotation angle θ is calculated using equation (8), there is no need to calculate the foramen area, but the eyeball size
The disadvantage that m′ must be determined in advance for each subject by another device cannot be overcome, and
The eyeball dimension m′ cannot be measured directly, but is calculated indirectly using various devices, so there is a large error. Therefore, in equation (8) using this value, there is a The disadvantage is that errors occur.

The major difference between Equation (8) and Equation (6) is that there are no dimensions that must be determined in advance for each subject using another device, and no eyeball dimensions that can be measured only indirectly, which can cause errors. It is.

〔Effect of the invention〕

The eye movement detection device of the present invention, instead of indirectly calculating the rotation angle θ of the eyeball from the movement of the eye mark imaged on the eyeball, uses Since the eyeball rotation angle θ and the rotation direction β can be calculated by directly measuring the amount and movement direction, there is an effect that the eyeball movement up to the maximum eyeball rotation angle can be accurately detected. moreover,
This has the effect that it is not necessary to calculate the eyeball dimension parameters for each subject in advance using a separate device, and in addition, there is the effect that the device can be simplified because there is no need to image the eye mark on the eyeball.

[Brief explanation of the drawing]

FIG. 1 is a block diagram showing one embodiment of the present invention;
FIG. 2 is a conceptual diagram showing the relationship between rotation of the eyeball and flattening of the contour of the foramen, and FIG. 3 is a diagram of the first calculation means 5.
FIG. 4 is a conceptual diagram showing the area calculating means 6, FIG. 5 is a conceptual diagram showing the geometric relationship between the eyeball rotation angle and the amount of foramen contour flattening, and FIG. 6 is a conceptual diagram showing the area calculating means 6. FIG. 2 is a block diagram showing an example. 1... Shutter, 2... Imaging section, 3a, 3b,
3c...Register, 4...Contour data extraction means,
5... center coordinate calculation means, 6... area calculation means,
7... Switch, 8... Eyeball rotation angle and rotation direction calculation means, 9... Means for forming an image of the eye mark on the eyeball, 10... Eye mark coordinate point extraction means, 11... Eye mark rotation angle and rotation direction calculation means, a... reflected light on the eyeball, b... image signal, c... fistula contour data, d... fistula center point, e... fistula area, f... to the eyeball. incident light, g...eye mark, h...reflected light from the eyeball, i...eye mark image signal, j...eye mark coordinate point, l...total number of dots between two points where the color tone changes , m...Distance from the center of rotation of the eyeball to the center of corneal curvature, m'...Distance from the center of rotation of the eyeball to the foramen, n...Iris, p, q...Contact between two parallel straight lines and the contour of the foramen Point, r...The radius of the canal when viewed straight, S ₀ ...The area of the canal when viewed straight, S...
Canal area at any rotation angle, u, v...
Both end points of line segment u, X ₁ , y ₁ ...X, y coordinates of point p,
X ₂ , y ₂ ...X and y coordinates of point q, X ₃ , y ₃ ...point d
X, y coordinates, y ₄ ...y coordinates of points u, v, X ₅ ,
y ₅ ...X and y coordinates of point j, θ ₁ ... rotation angle of the eye mark, β ₁ ... rotation direction of the eye mark, θ ... rotation angle of the eyeball, β ... rotation direction of the eyeball.

Claims (1)

[Scope of Claims] 1. A shutter, an imaging section, a first register for storing an image signal converted by the imaging section, and an extraction means for extracting contour data based on the image signal stored in the first register. a first calculation means for calculating center coordinates of the contour based on the contour data; a second calculation means for calculating an area within the contour based on the contour data; and a first calculation means for calculating the center coordinates of the contour based on the contour data; a switch for selecting an area to be stored in a register from among the selected areas; a second register for storing the contour area selected by the switch; and a second register for storing the contour area stored in the second register and a second
an eyeball movement detecting device comprising: a third calculating means for calculating a rotation angle and a rotation direction of the eyeball based on the contour area and the center coordinates which are not stored in the register;