JP2000067225A - Eye position detector - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide an eye position detector with which the positions of eyes can be accurately detected or chased even in the case of a driver who wears eyeglasses. SOLUTION: An input image photographed by a camera 21 as an image input means CL1 is stored through an A/D converter 22 into an image memory 23. An image data arithmetic circuit 24 composed of a density detecting means CL2 for a pixel stream, point extracting means CL3, extracting means CL4 for a curve group, detecting means CL5 for the curve group at the top end, setting means CL6 for a prescribed area 2, lateral symmetric state discriminating means CL7 for curve group, setting means CL8 for a prescribed area 3, reducing means CL9 for a prescribed area 1 and presence/absence discriminating means CL10 for nostril data and an eye position detecting circuit 25 composed of an eye position specifying means CL11 with the nostril data as a reference and an eye position specifying means CL12 with the laterally symmetric curve group as a reference are connected to the image memory 23.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an eye position detecting device, more specifically, to automatically control a mirror and a sheet.
The present invention relates to an eye position detection device that can be used to identify the position of an eye required to detect a dozing driving state from an open / closed state of a driver's eyes.

[0002]

2. Description of the Related Art A conventional eye position detecting device for a driver or the like has been disclosed in, for example, Japanese Patent Application No. 8-101904. This eye position detection device detects the density of pixels along a pixel row in the vertical direction of the face, determines one pixel for each local increase in density in the pixel row, sets the pixel as an extraction point, and sets the pixel as an extraction point. The configuration is such that the extraction points adjacent to each other in the pixel column direction of the pixel column are connected to detect the position of the eye from a group of curves extending in the horizontal direction of the face.

[0003]

However, in such a conventional eye position detecting device, in the case of a driver wearing spectacles, reflection may occur on the lens portion of the spectacles, At this time, if the position of the eyes is detected, there is a problem that the eyes cannot be correctly captured. In addition, when the eye is correctly caught, and the eye data is tracked and the judgment of the open / closed eye is performed, if the reflection occurs, the eye data is lost, so that the open / closed eye detection cannot be continuously performed. There was also a problem.

The present invention has been made in view of such conventional problems, and an object of the present invention is to provide an apparatus which can maintain the eye position detection accuracy even for a driver wearing eyeglasses. .

[0005]

According to the present invention, in order to solve the above-mentioned problems, as shown in FIG. 1, an image input means CL1 for inputting a face image and a face image input from the image input means CL1 are used. Density detection means CL2 for a pixel row for detecting the density of a pixel row in the vertical direction, point extraction means CL3 for extracting a point based on a rise in the density of the pixel row and its change state, and a pixel row direction of an adjacent pixel row. A curve group extraction means CL4 for connecting adjacent points to extract a curve group in the horizontal direction of the face, and a curve group appearing at the uppermost end in a predetermined area 1 (an area which can cover the entire image, see FIG. 6) Curve group detecting means CL5 at the uppermost end for detecting
And a predetermined area 2 based on the uppermost curve group (an area set to determine whether or not the corresponding continuous data is a frame of eyebrows, eyes, or glasses based on a symmetric state, see FIG. 10) Setting means CL for the predetermined area 2 for setting
6; a left-right symmetric state determination unit CL7 for determining a left-right symmetric state of the curve group appearing in the predetermined area 2;
When a symmetrical curve group is detected in the predetermined area 2, a predetermined area 3 based on the position of the curve group (an area set for determining whether there is nostril data, FIG.
3) setting means CL8 for setting a predetermined area 3; and a predetermined means for reducing the predetermined area 1 based on the uppermost curve group when a symmetric curve group cannot be detected in the predetermined area 2. The reduction means CL9 of the area 1, the presence / absence determination means CL10 of the nostril data which determines presence / absence of the nostril data appearing in the predetermined area 3, and the nostril data could be detected in the predetermined area 3 In the case, the eye position specifying means CL11 for specifying the position of the eye based on the position of the nostril hole, and the eye position based on the symmetric curve group when no nostril data is detected in the predetermined area 3. Eye position specifying means CL12 for specifying the position is provided.

The operation of the present invention will be described. According to the present invention, the eyebrow or eye candidate data is specified in the predetermined area 2 and the presence or absence of nostril data is confirmed based on the candidate data, thereby confirming that the detection target is the eye. Thus, it is possible to provide a device that can maintain the eye position detection accuracy even for a driver wearing eyeglasses.

[0007]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS An embodiment of an eye position detecting device according to the present invention will be described below in detail with reference to the accompanying drawings.
FIG. 2 is a diagram showing an embodiment of the eye position detecting device according to the present invention.

First, the configuration will be described. Image input means C installed to install the instrument and photograph the driver's face from the front
There is a camera 21 as L1. In this embodiment, an input image of the camera 21 is composed of 512 pixels in the horizontal (X) direction and 480 pixels in the vertical (Y) direction, as shown in FIG. An input image captured by the camera 21 is converted into a digital amount of input image data via an AD converter 22 in an image memory 2.
3 is stored. The image memory 23 includes a pixel line density detecting unit CL2 for detecting the density of a vertical pixel line of the face based on the input image data stored in the image memory, and a density increase in the pixel line and a change state thereof. A point extraction unit CL3 for extracting points by the following equation; a curve group extraction unit CL4 for connecting points adjacent to each other in the pixel column direction of an adjacent pixel column to extract a curve group in the horizontal direction of the face;
Predetermined area 1 (area that can cover the entire image, see FIG. 6)
In the uppermost curve group detecting means CL5 for detecting a curve group appearing at the uppermost end in a predetermined area 2 (continuous data corresponding to eyebrows or eyes, or eyeglasses from a left-right symmetric state) based on the uppermost curve group The setting means CL6 of a predetermined area 2 for setting an area to be set for determining whether or not the frame is a frame, and a curve for determining a left-right symmetry state of a curve group appearing in the predetermined area 2 The left-right symmetry determination unit CL7 determines whether a predetermined symmetric curve group is detected in the predetermined area 2 and a predetermined area 3 based on the curve group position (e.g., whether there is nostril data). The setting means CL8 of a predetermined area 3 for setting an area to be set for determination, see FIG. 13) and the uppermost curve group when no symmetrical curve group is detected in the predetermined area 2 As the predetermined area 1 Reduction means CL of a predetermined area 1 for small
9, an image data calculation circuit 24 comprising: a nostril hole data presence / absence determining means CL10 for determining the presence / absence of nostril hole data appearing in the predetermined area 3; Eye position specifying means CL11 for specifying the position of the eye with reference to the position of the nostril when the can be detected,
An eye position detecting circuit 25 comprising eye position specifying means CL12 for specifying an eye position based on the symmetrical curve group when nostril data cannot be detected in the predetermined area 3 is connected. I have.

Next, the flow of the operation in the above configuration will be described with reference to the flow charts of FIGS. 3 to 5, 15 and 18, and FIGS.
This will be described with reference to the explanatory diagrams of FIGS. 14, 16, 17, and 19 to 24.

First, in step S101 of FIG.
The predetermined area 1 is set. As shown in FIG. 6, the predetermined area 1 is an area where the entire image can be covered. This is to allow a change in the eye position due to a difference in physique or the like.

In step S102, pixel data in the predetermined area 1 is read, and in step S103, candidate eye data to be a curve group (in the following steps, the curve group is referred to as continuous data). Step S10
2, the processing in S103 will be described with reference to the flowchart of FIG. 5 and the explanatory diagrams of FIGS.

First, in step S201 of FIG.
A driver's face is photographed by a camera, and an input image for one frame is converted into a digital signal and stored in an image memory.

In step S202, as shown in FIG.
After the end of the line, move to the processing of the next line,
It is determined whether or not point extraction has been completed for all lines in a predetermined direction.

If it is determined in step S202 that point extraction has not been performed on all lines, the process proceeds to step S203. This processing is intended to eliminate small variations in the change in density value at the time of capturing image data, and to capture a global change in density value. FIG. 7A shows a processing result of the arithmetic averaging operation of the line data of Xa in FIG.

In step S204, step S203
The differential operation is performed on the arithmetic mean value that is the result of the operation.
This processing result is shown in FIG.

In step S205, step S204
The point is extracted based on the differential value that is the calculation result of. This point extraction method extracts points (p1, p2) at which the differential value changes from negative to positive, and in FIG. 7A, points where the graph is convex downward. Next, it is determined whether or not the change (q1, q2) of the density value until the point is reached is equal to or less than a predetermined value, based on whether or not it is equal to or less than the reference value for the differential value in FIG. , A Y coordinate value (A1, A1
Extract 2). After this processing is completed for one line, the processing is switched to the processing for the next line in step S206.

If it is determined in step S202 that point extraction has been completed for all lines, points as shown in FIG. 8 are extracted. That is, on the Xa line in FIG.
This means that two points A1 and A2 have been extracted.

Thereafter, the process proceeds to step S207, where the Y coordinate values of the extraction points (A1, A2, A3,...) Of the adjacent lines are compared, and if the Y coordinate value is within a predetermined value, the The data group number, continuous start line number, and continuous data number are stored in memory.

In this case, since the detection target is an eye, the feature amount thereof can be data that continues relatively long in the horizontal direction. Therefore, continuous data is extracted on the condition that the data continues for a predetermined value or more in the horizontal direction. I do. Therefore, step S
As a processing result in 207, G1 to G as shown in FIG.
4 continuous data can be recognized as a curve group.

After extracting the continuous data in step S103 of FIG. 3, the process proceeds to step S104. Step S104
Then, the judgment value of each continuous data is read.
The judgment value of each continuous data is the representative X coordinate of the continuous data by calculating the median value of the continuous start line and the end line of the continuous data, and the continuous data by calculating the average value of the Y coordinate value of each extraction point of the continuous data. Means the representative Y coordinate.

In step S105, it is determined whether or not a predetermined area 3 for detecting nostril data is set. Since the predetermined area 3 has not been set immediately after the start, the process proceeds to step S106.

In step S106, it is determined whether or not a predetermined area 2 for detecting symmetric continuous data is set. In step S106, similarly to step S105, since the predetermined area 2 is not set immediately after the start, the process proceeds to step S107.

In step S107, step S10
5. Since it is determined in S106 that neither the predetermined area 3 nor 2, the continuous data is detected in the predetermined area 1. In this step, as shown in FIG. 10, the uppermost continuous data appearing in the predetermined area 1 is detected. In this process, instead of suddenly patterning a large number of continuous data appearing in the entire image, the continuous data appearing at the uppermost end is first focused on, and the corresponding continuous data is used as a face feature amount (continuous eyebrow and eye data). ), The eye search method can be performed as easily as possible.

In step S108, as shown in FIG. 10, a predetermined area 2 is set based on the representative coordinate value of the uppermost continuous data. The predetermined area 2 is an area set in order to determine whether the continuous data corresponding to the left-right symmetry state is a frame of eyebrows, eyes, or glasses.

After setting the predetermined area 2 in step S108,
Returning to step S102, the image data of the next frame is read out in the predetermined area 2. Then, step S
The same processing and determination as the above-described contents of 103 to S105 are performed, and the process proceeds to step S106.

In step S106, since the predetermined area 2 has already been set, the flow shifts to step S112 in FIG. In step S112, two consecutive data appearing in the predetermined area 2 are detected from the upper part.

In step S113, as shown in FIG. 11, it is determined whether or not the corresponding two continuous data satisfy the left-right symmetry condition. At this time, the left-right symmetry condition of the two continuous data is set such that the continuous data appearing from the upper part of the predetermined area 2 has a length equal to or more than a predetermined value, taking into account the wearer of glasses as shown in FIG. The number of occurrences may be one. Even if there are two or more continuous data, one continuous data having a length equal to or more than a predetermined value is determined. If the continuous data appearing in step S113 satisfies the left-right symmetry condition, step S114
Move to

In step S114, FIG.
As shown in (b), a predetermined area 3 is set on the basis of two symmetrical continuous data. The predetermined area 3 is an area set to determine whether or not nostril data exists.

If the left-right symmetry condition is not satisfied in step S113, the flow shifts to step S115, and FIG.
As shown in (a) and (b), a predetermined area 1 which is reduced so as not to include the coordinate values of the continuous data used as a reference for setting the predetermined area 2 is set, and the process proceeds to step S116.

In step S116, the reduction limit of the predetermined area 1 is determined. Until the reduction limit is reached, the process returns to step S102 to read the image data of the next frame in the reduction predetermined area 1 shown in FIG. Do. Thereafter, in step S105 and step S106, the setting of the predetermined area is 1, so the determination is NO, and the process proceeds to step S107. At this time, the continuous data at the uppermost end in the predetermined reduced area 1 is the continuous data B in FIG.
Becomes

If it is determined in step S116 that the predetermined area 1 has reached the reduction limit, step S10
1, the predetermined area 1 in the initial state is set again.

If the left-right symmetry condition of the continuous data is satisfied in step S113 and the predetermined area 3 is set in step S114, the process returns to step S102, and the image data of the next frame is read in the predetermined area 3. Thereafter, the same processing and determination as the above-described contents of steps S103 and S104 are performed, and the process proceeds to step S105.

In step S105, since the predetermined area 3 has already been set, the flow shifts to step S109. In step S109, it is determined whether or not the nostril data exists in the predetermined area 3.

The details of the nostril hole data detection process will now be described with reference to the flowcharts of FIGS. 15 and 18 and the explanatory diagrams of FIGS. 16, 17 and 19.

In step S301 in FIG. 15, FIG.
As shown in (a) and (b), the density values of the predetermined area 3 are read in the horizontal direction at predetermined intervals, and the process proceeds to step S302.

In step S302, it is determined whether or not all the horizontal lines in the predetermined area 3 have been searched.
In the initial state, the line search in the horizontal direction has not been performed, so the process proceeds to step S303.

In step S303, arithmetic operation of the density value of one horizontal line is performed. This processing is intended to eliminate small variations in the change in density value at the time of capturing image data, and to capture a global change in density value. FIG. 17A shows a processing result of the arithmetic averaging operation of the line data of Ya in FIG. 16B.

In step S304, step S303
The differential operation is performed on the arithmetic mean value that is the result of the operation.
This processing result is shown in FIG.

In step S305, step S304
The point is extracted based on the differential value that is the calculation result of. In this point extraction method, the points (p1, p2) where the differential value changes from negative to positive, that is, the points where the graph becomes convex downward in FIG. 17A, are extracted. Next, it is determined whether or not the change (q1, q2) in the density value until reaching the point is equal to or less than a predetermined value, based on whether or not the difference value is equal to or less than the reference value for the differential value in FIG. , X-coordinate values (A1,
A2) is stored.

After storing the coordinate values in one horizontal line, it is determined in step S306 whether or not the number of stored points is two. At this time, if one or three or more points are stored,
It is determined that there is no nostril data in the corresponding line, the process proceeds to step S308, returns to step S302 for switching to the next line, and repeats the same processing.

If it is determined in step S306 that the number of points is two, the flow shifts to step S307 to determine whether or not the horizontal interval between the points is within a predetermined range. The predetermined range is a nostril interval in consideration of individual differences, and if not within the range, it is determined that there is no nostril data in the corresponding line, and the process proceeds to step S308.
Returning to the next line switching step S302, the same processing is repeated.

If it is determined in step S307 that the interval between the corresponding points is within a predetermined range, the process proceeds to step S307.
The flow shifts to 309, where the reference coordinates of the nostrils are set as shown in FIG. 20A, and the nostrils data detection process ends.

This process is repeated, and when nostril data is not detected to the end, the process proceeds from step S302 to step S310, a no-detection flag of nostril data is set, and this process ends.

Next, with reference to the flowchart of FIG. 18 and the explanatory diagram of FIG. 19, a description will be given of the contents of detection processing which can cope with a case where the light environment is poor and two nostrils data do not appear clearly.

If nostril data does not appear clearly,
This occurs when the sun's light hits the right half of the face and the left half of the face becomes shadow. The density change of the nostril data at this time is as shown in FIG. A method of reliably recognizing even the nostril data having such a density change will be described with reference to the flowchart of FIG.

In the flow chart shown in FIG. 18, steps S301 to S304 are the same as the processing contents described with reference to the flow chart of FIG.

In step S321, the first point (A1 in FIG. 19) is detected under the same density change condition as in the above embodiment, and the flow advances to step S322.

In step S322, the extraction range of the second point is set based on the horizontal coordinate value of the first point. A method for setting the extraction range of the second point will be described with reference to FIGS.

D1 shown in FIG. 19A assumes the minimum value of the interval between nostrils in consideration of individual differences. Also,
The maximum value in consideration of the individual difference of the nostril interval is set as D2. In this way, by specifying the section in which the other nostril data (the second point) appears, the point which becomes convex downward in the graph of the density change changes from negative to positive by the change of the differential value. Capture as point p2. With this nostril data detection method, even if one nostril data is in the shadow, it can be reliably detected.

In step S323, it is determined whether or not this second point (the point at which the differential value changes from negative to positive) can be extracted within the above-described range. If the second point exists, the flow proceeds to step S309. Set the reference coordinates of the nostrils.

In step S109 of FIG. 3, it is checked whether or not there is nostril data. If there is nostril data, the process proceeds to step S110.

In step S110, as shown in FIG. 20B, an eye position detection area is set based on the position of the nostrils. Then, continuous data appearing in the eye position detection area is detected. At this time, the specification of the eye data is determined in consideration of the relative positional relationship between the nostril data and the relative position when a plurality of continuous data appears.

According to the present invention, by associating the position of the eye with the position of the nose by the method described above, the probability that the finally specified continuous data is the eye is increased, and the eye detection accuracy is further improved. Can be done.

The effect of associating the position of the eye with the position of the nose may be as follows. In the case of tracking an eye once detected, as shown in FIG. 21, if reflection occurs temporarily in the spectacle lens unit, the tracking process is abandoned, and the eye position must be detected again. Sometimes. This situation has a problem that it takes time to detect the position of the eye. At this time, FIG.
As shown in the figure, the relationship between the position of the nostrils and the position of the eyes is represented by Xd, Y
By grasping with dimensions such as d, as shown in FIG. 22B, it is possible to track based on the position of the nostrils. If the presence of the nostril data can be confirmed by controlling the eye position detection area reduced based on the detected nostril position as a reference, even if the data is temporarily lost, for example, when the driver falls asleep, The open / closed eye determination used for detection or the like can be kept in a temporarily interrupted state, and when the reflection on the spectacle lens unit is eliminated, the open / closed eye determination processing and the like can be resumed immediately.

Finally, the processing in the case where nostril data cannot be detected in the predetermined area 3 in step S109 of FIG. 3 will be described.

If the nostril data cannot be detected in step S109, the flow shifts to step S111 to detect the position of the eye using information up to the symmetrical continuous data.
A specific eye position detection method will be described with reference to FIGS.

The symmetrical data in the face includes eyebrows, eyebrows, eyeglass frames, etc., in addition to eyes. FIG. 23 is an explanatory diagram when the eyebrows are captured as symmetrical continuous data.
In this manner, the detection area of the eye is set slightly widened downward with reference to the corresponding continuous data one by one in the left-right direction, and when two continuous data are detected in this area, the lower continuous data is set to the eye. When only one continuous data is detected, the continuous data is identified as an eye.

Further, as shown in FIG. 24, when frames of eyeglasses are considered as continuous data, and continuous data that should be symmetrical is connected, the determination is made based on the length of the continuous data. An eye position detection area is set, and the position of the eye is specified.

The processing in step S111 may be performed when the driver wears a mask, or when the angle of view of the camera has not been adjusted and only the upper half of the face has shifted. Step S111 is prepared to enable normal processing even for such an irregular driver.

[0060]

As described above, according to the present invention, the eyebrow or eye candidate data is specified in the predetermined area 2, and the presence or absence of nostril data is confirmed based on the candidate data. This increases the likelihood that the detection target is the eye, and enables a driver wearing eyeglasses to perform highly accurate eye position detection and tracking.

[Brief description of the drawings]

FIG. 1 is a block diagram illustrating a configuration of an eye position detection device according to the present invention.

FIG. 2 is a block diagram illustrating a configuration of an embodiment of an eye position detection device according to the present invention.

FIG. 3 is a flowchart for explaining the operation of the embodiment.

FIG. 4 is a flowchart for explaining the operation of the embodiment.

FIG. 5 is a flowchart for explaining a method of extracting points of a curve group.

FIG. 6 is an explanatory diagram relating to an eye position detection method.

FIG. 7 is an explanatory diagram relating to an eye position detection method.

FIG. 8 is an explanatory diagram relating to an eye position detection method.

FIG. 9 is an explanatory diagram relating to an eye position detection method.

FIG. 10 is an explanatory diagram relating to an eye position detection method.

FIG. 11 is an explanatory diagram relating to an eye position detection method.

FIG. 12 is an explanatory diagram relating to an eye position detection method.

FIG. 13 is an explanatory diagram relating to an eye position detection method.

FIG. 14 is an explanatory diagram relating to an eye position detection method.

FIG. 15 is a flowchart illustrating a method for detecting nostril data.

FIG. 16 is an explanatory diagram related to detection of nostril data.

FIG. 17 is an explanatory diagram relating to detection of nostril data.

FIG. 18 is a flowchart illustrating a method for detecting nostril data according to another embodiment.

FIG. 19 is an explanatory diagram relating to detection of nostril data according to another embodiment.

FIG. 20 is an explanatory diagram relating to an eye position detection method.

FIG. 21 is an explanatory diagram relating to an eye position detection method.

FIG. 22 is an explanatory diagram relating to an eye position detection method.

FIG. 23 is an explanatory diagram relating to an eye position detection method.

FIG. 24 is an explanatory diagram related to an eye position detection method.

[Explanation of symbols]

CL1 Image input means CL2 Density detecting means for pixel rows in the vertical direction of the face CL3 Point extracting means for each pixel row CL4 Curve group extracting means CL5 Uppermost curve group detecting means within predetermined area 1 CL6 Setting means for predetermined area 2 CL7 Means for determining the symmetry of the curve group appearing in the predetermined area 2 CL8 Setting means for the predetermined area 3 CL9 Reduction means for the predetermined area 1 CL10 Means for determining the presence or absence of nostril data in the predetermined area 3 CL11 Nose hole data Eye position specifying means CL12 based on symmetrical eye position specifying means based on symmetrical curve group 21 Camera 22 A / D converter 23 Image memory 24 Image data calculation circuit 25 Eye position detection circuit

──────────────────────────────────────────────────続 き Continued on the front page (51) Int.Cl. ⁷ Identification symbol FI Theme coat ゛ (Reference) G08G 1/16 A61B 5/10 310Z // B60R 1/00 G06F 15/70 330Z

Claims (4)

[Claims] 1. An eye position detecting apparatus for processing an image data of a face to detect an eye position, wherein a density of pixels is read out along a pixel row in a vertical direction of the face, and one of density values on the pixel row is read out. A point extracting means for specifying a pixel of a one-way peak of the density value as an extraction point, provided that a differential value of the density change before the peak in the direction exceeds a predetermined value; and a pixel row of an adjacent pixel row. A curve group extraction means for connecting the extraction points close to each other in the direction to extract a curve group extending in the lateral direction of the face; and detecting a curve group appearing at the top end in a predetermined area 1 (an area which can cover the entire image). An uppermost curve group detecting means, a predetermined area 2 based on the uppermost curve group (set to determine whether or not the corresponding continuous data is an eyebrow, eye, or eyeglass frame based on a symmetrical state) Area to set) Setting means, and left-right symmetric state determining means for determining a left-right symmetric state of a curve group appearing in the predetermined area 2; and when a left-right symmetric curve group is detected in the predetermined area 2, the curve Setting means for setting a predetermined area 3 (area set for determining whether there is nostril data) based on the group; and a set of symmetrical curves in the predetermined area 2 When no is detected, a means for reducing the predetermined area 1 based on the curve group, a means for detecting the nostrils appearing in the predetermined area 3, a means for detecting nostrils, Eye position specifying means for specifying the position of the eye based on the position of the nostril data when the nostril data is detected in the predetermined region 3; and detecting the nostril data in the predetermined region 3 When there is no left-right symmetric curve group in the predetermined area 2 An eye position specifying device for specifying the position of the eye based on a reference. 2. The eye position detecting device according to claim 1, wherein the eye position specifying means based on the nostril hole data position detects that the eye position is detected by the eye position specifying means. An eye position detecting device, wherein a position of an eye is always detected using a nostril data as a tracking target as a trigger. 3. The eye position detecting device according to claim 1, wherein the nostril detecting means detects nostril data from a change in density in a lateral direction. 4. The eye position detection device according to claim 1, wherein the nostril detection means captures one nostril data by a change in density, and detects the other nostril data based on the position. An eye position detection device characterized by detecting a change in density.