JP3422183B2 - Eye detection method for face images - Google Patents

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a method for detecting eyes in a face image obtained by photographing a human face.

[0002]

2. Description of the Related Art Recently, a device for preventing a driver of a vehicle such as an automobile from falling asleep has been developed, and a part of the device is put to practical use. As an example of this drowsy driving prevention device, there is a device that determines a drowsy state based on the movement of the driver's eyes, for example, the time the eyes are closed. In this case, in order to capture the movement of the driver's eyes, it is necessary to photograph the driver's face and identify the eyes from this face.

[0003] JP 7-18 1 012 JP, a face region is extracted by binarizing an image including the face, relative positions and each other a plurality of clusters of black pixels within this face area A technique for identifying an eye from the aspect ratio of each block of black pixels is disclosed. A block of black pixels is an eyebrow, an eye, a nostril, etc., and there are no nostrils outside the width of the left and right eyebrows and eyes, the eye is under the eyebrow, etc. Then, the eyes are determined from the aspect ratio of this candidate.

[0004]

According to the above publication, in order to detect the eyes, it is necessary to confirm the relative positions of the blocks of black pixels, which causes a problem of complicated arithmetic processing. . Further, if any one of the elements such as eyebrows and eyes is missing, it becomes impossible to detect the eyes, and also if the eyeglass frame or shadow exists as a block of black pixels, the eyes cannot be detected. There was a problem.

The present invention has been made to solve the above-mentioned problems, and an object of the present invention is to provide a method for detecting eyes in a face image that can detect the eyes in a short time.

[0006]

In order to achieve the above-mentioned object, a method for detecting eyes in a face image according to the present invention comprises:
An imaging step of capturing a face of a person to obtain a face image, a nostril detection step of detecting a nostril from the face image, and a search area setting for setting an area for eye search based on the position of the nostril And an eye detecting step of detecting an eye in the search area. Then, the search area setting
In the constant step, the nose in the nostril detection step
The size of the search area changes depending on the continuation of hole detection.
Be changed.

According to this structure, since the eye search area is set with reference to the nostrils that are easy to detect, the eye search area, that is, the area where the eye is likely to exist is quickly and surely specified. be able to. It can also detect nostrils
When it disappears, the position or orientation of the driver's face is
This is the case when it changes drastically.
By increasing the size, you can reliably detect the eyes.
it can.

Further, in the eye detection, an edge in which the brightness changes sharply in the vertical direction can be extracted, and a portion where the edge is continuous in a substantially elliptical shape can be detected as an eye. Alternatively, it is possible to extract an edge pair in which a negative edge whose brightness sharply decreases from the top to the bottom and a positive edge whose brightness sharply increases from the top to the bottom are vertically adjacent to each other.

By detecting the eyes with the edges of the substantially elliptical shape, elements other than the eyes such as eyebrows and frames of eyeglasses can be excluded, and the eyes can be surely detected. Also,
When a pair of a negative edge and a positive edge that are vertically adjacent to each other is detected as an eye, the detection can be performed more quickly.

Furthermore, in the search area setting step, the position of the search area may be changed according to the change in the orientation of the two nostrils. According to this, even when the driver tilts his / her neck, the eyes can be surely detected.

[0011]

In order to achieve the above-mentioned object, the present invention
Another way to detect eyes in other facial images is to capture a person's face.
To obtain a face image, and a nostril from the face image.
Nostril detection step to detect and the position of the nostril as a reference
Area setting step to set the area to search the eyes
And an eye detection step of performing eye detection in the search area
And have. In the nostril detecting step, the photographed face image is binarized by a predetermined threshold to detect a nostril position, and eyes are searched based on the nostril position. When the eye is not detected in the eye detecting step, the threshold value for the binarization process is changed and the nostril can be detected again. According to this, the threshold value used for the binarization process can be set to an appropriate value.

[0013]

BEST MODE FOR CARRYING OUT THE INVENTION Preferred embodiments (hereinafter referred to as embodiments) of a method for detecting eyes in a face image according to the present invention will be described below with reference to the drawings.

FIG. 1 is a block diagram showing a schematic configuration of a drowsy driving prevention apparatus using the present invention. Infrared projector 1
0 emits infrared rays to the area including the driver's face, and the charge coupled device (CCD) camera 12 captures an image of the area including the driver's face. In front of the lens of the CCD camera 12, an optical filter that transmits mainly infrared wavelengths (near 870 nm) is installed. The image taken by the CCD camera 12 is sent to the blink recognition section 14, and the time of one blink, that is, the time during which the eyes are continuously closed (eye closing time) is measured. The dozing determination unit 16 determines the dozing state based on the length of time the eyes are closed, and when the dozing starts, the warning unit 18 issues an alarm to alert the driver.

Basic Flow FIG. 2 is a block diagram of the drowsiness prevention device, particularly the blink recognition unit 14.
A basic flow chart of the above is shown. When the device is activated, it is detected where the driver's eyes are (S100). In this initial search step, the position of the driver's face changes greatly depending on the physique of the driver, the position of the driver's seat, and the like, so a relatively wide search area is set and the nostrils are first detected. Since the positional relationship between the nostrils and the eyes does not change so much depending on the driver, an area in which the eyes are estimated to be present is determined from the position of the nostrils, and the eyes are detected in this area. When the eyes are detected, the position of the nostrils at this time,
The positional relationship between the nostrils and the eyes is stored.

When the eyes are detected, the eyes are tracked (S200). Once you know the location of the nostrils,
Since the position of the nostril does not change so much unless the driver performs a very large motion, the nostril search area is significantly narrowed and the nostril search is performed. When a nostril is detected, the eye search area is determined from the positional relationship between the nostril and the eye, and the eye is detected. The eye search region at this time can be made smaller than the region at the time of the initial search.

When the eyes are detected (S300),
It is determined whether the eyes are closed or open (S
400). On the other hand, when the eyes are not detected, the re-search control for rediscovering the eyes is started (S500).
In this re-search, several search areas that are different according to the elapsed time after the nostril is lost are set. When the elapsed time after losing sight of the nostril is short, the search area is set narrower, and when the elapsed time is longer, the search area is set wider. If you haven't lost sight of your nostrils for a while, it is unlikely that your nostrils or eye positions have changed significantly. It can be shortened.

Then, from step S200 to step S
500 is repeated, and each time it is determined whether the eyes are closed. The time during which the closed state continues is measured by the doze determination unit 16 and the doze state is determined based on the continuous time. Hereinafter, the initial search step S
100, tracking step S200, blink determination step S4
00 and the re-search step S500 will be described in detail.

Initial Search Routine FIG. 3 shows an initial search step S immediately after the present apparatus is started.
A detailed control flow chart of 100 is shown. First, the nostril search area is set (S102). In the initial search, the position of the driver's nostrils greatly varies depending on the driver's physique, seat position, etc., so that the nostril search area 20a has the entire driver's face as shown in FIG. 4 (a). Set to be included. Next, the first timer is set to a predetermined value, for example, 2 minutes (S104). As will be described later, this first timer is
It is a timer that measures the time to search the same area while changing the binarization threshold. Further, the second timer is set to a predetermined value, for example 10 minutes (S106). Second
The timer is a timer for measuring the time of the first search,
If no eyes are found after 0 minutes, there is some abnormality and the driver is warned. At the time of the first search, it is considered that the driver is not yet in the dozing state, so the eyes are searched for with sufficient time.

Next, the threshold value for binarizing the face image is set (S108). On the other hand, looking at the match between the typical template of the shape of the nostrils and the face image, a candidate whose match exceeds a predetermined value is selected (S110). And
Binarization processing is performed on the area including each selected candidate using the threshold value set in step S108 to extract a black area (S112). Then, the two black areas arranged substantially horizontally are detected as nostrils (S114). At this time, if the nostril is not found, the process returns to step S108 (S11
6) The threshold for binarization is set again (changed) and the nostril is searched again.

When the nostril is detected in step S116, the eye search area 22 is set based on the position of the nostril (S118). If the position of the nostril is known, the area where the eyes are present can be almost specified although there are individual differences. Therefore, as the eye search area 22, a slightly wider area 22a is set as shown in FIG. Further, the first search area 22a is set slightly downward so that the eyebrows are not mistakenly detected as eyes, so that the search areas 22a do not touch the eyebrows.

Next, in the above-mentioned eye search area 22a,
An edge whose brightness changes sharply in the vertical direction is extracted (S
120). There are two types of edges, a negative edge where the brightness decreases sharply from top to bottom, and a positive edge where the brightness rises sharply from top to bottom. In the case of eyes, the negative edge is on the top and the positive edge is on the bottom. A pair is formed. This pair of positive and negative edges is selected as an eye candidate (S122). Then, it is determined whether the selected candidate has a substantially elliptical shape as a whole (S124). When it is determined that the eye shape is substantially elliptical, it is determined that the eye shape, the position with respect to the nostril is stored (S126), and this routine ends.

If the positive / negative edge pair is not substantially elliptical in step S124, the first and second timers are counted down (S128). If the first timer is not 0 (S130), the process returns to step S108 and the binary value is returned. Change the threshold of activation and start over from detecting the nostril. First
When the timer is 0, that is, when the eyes cannot be detected even after 2 minutes have elapsed (S130), it is further determined whether the second timer is 0 (S132). And the second timer is 0,
That is, if the eyes cannot be detected even after 10 minutes have passed, it is considered that the driver is wearing sunglasses or a mask, or there is an obstacle in front of the camera. (S134), the flow ends.

If the second timer is not 0 (S1
32), eye search region 22 set based on the nostril position
The process of shifting a upward is performed (S136), and the first timer is set to 2 minutes again (S138). As described above, the eye search region 22a set in step S118
Is set to the lower side so as not to detect the eyebrows, so there is a possibility that the eyes may be higher up depending on the driver. Therefore, in step S136, the search area is shifted upward. By gradually shifting it upward, only the eyes are extracted so that the eyebrows are not mistakenly detected as eyes.

In the above initial search routine, the characteristic nostrils are detected in the binarized image, and the eyes are detected with reference to this position, so that the eyes can be surely detected. The nostrils are arranged substantially horizontally and can be detected as two substantially round black areas. Also, since there are few individual differences, it is easy to detect. Also, the positional relationship between the nostrils and the eyes
To some extent, it is possible to specify that the left eye is located diagonally above and to the right of the nostril, so that the eye search region can be specified accurately with this positional relationship.

Further, eye detection is performed by edge detection without performing binarization processing. This is because the edge remains even if it is affected by illumination or light from the outside of the vehicle compartment, so that detection can be performed reliably. Although this edge detection itself takes a little longer processing time, it does not significantly affect the processing time as a whole because it limits the eye search area from the nostril position as described above.

Further, 2 set in step S108.
The threshold used for the quantification process is the
It is determined whether the value was appropriate. That is,
If no eye is detected for a nostril candidate, then the candidate is determined not to be a nostril. Therefore, an appropriate threshold value can be set by changing the threshold value, performing the initial search, and adopting the threshold value when the eyes are detected. Therefore, it can be finally prevented that the threshold value is improper and the edges or eyes of the glasses are mistaken as nostrils.

Tracking Routine FIG. 6 shows a detailed flowchart of the tracking step S200 that follows the initial search step S100. When nostrils and eyes are detected in the initial search, a routine for tracking them is started. Based on the position of the nostril detected in the previous search, a nostril search region 20 of a predetermined size is set around this position (S202). This area is a horizontally long rectangular area including the nose, as shown in FIG. The area 20c is set to be smaller than the area 20a in the above-described first search and smaller than the area 20b in the second mode at the time of re-search described later. This is if the nostrils were detected last time,
This time, it is based on the premise that nostrils can be detected without searching for such a large range. Further, the search region 20c is set to be wide in the horizontal direction because the action of shaking the head sideways is faster than the action of swinging the head vertically in the operation during driving. This is because it corresponds to the lateral movement of the hole.

Then, the inside of the nostril search area 20c is binarized to extract a black area (S204). Then, the two black areas arranged substantially horizontally are detected as nostrils (S2).
06). If no nostril is not detected (S208),
End the tracking routine. When a nostril is detected, the eye search area 22b is set based on the previous positional relationship between the nostril and the eye (S210). This search area 22b is shown in FIG.
As shown in, the size is slightly larger than the size of the eyes and slightly smaller than the area 22a at the time of the initial search. This is because in the previous search, the positional relationship between the nostrils and the eyes was known, and since the previous nostril positions were known, the eye positions could be specified quite accurately. Then, in the search area 22b, the edge is extracted similarly to the initial search (S212). Further, an edge pair in which positive and negative edges are adjacent in the vertical direction is searched (S214), and if there is this edge pair, this position and the position of the nostril are stored (S2).
18). If there is no edge pair (S216), the tracking routine ends.

As described above, in the tracking routine, since the previous nostril position, the eye position, and their relationship are known, the nostril and eye search regions 20c and 22b are set smaller than the regions 20a and 22a at the time of the initial search. can do. It was important to detect the nostrils and eyes reliably at the time of the first search as described above, but it is required to detect the eyes promptly at the time of tracking. In this device, since the doze determination is made with the continuous time in the eye-closed state, if the tracking takes a long time, the determination becomes inaccurate or impossible. Therefore, in order to quickly detect the eyes,
First, the most characteristic nostril is detected in the binarized image, and the eye search region 22b is narrowed down based on this nostril. The edge detection time, which tends to increase the processing time, is shortened.

Blink Determination Routine FIG. 7 shows a control flowchart of the blink determination step S400. The eye edge pairs detected in the tracking step S200 are read (S402), and the maximum value of the vertical distance of each pair is set as the eye opening at that time (S404). Then, this eye opening is compared with a predetermined threshold value (S406), and if it is less than the threshold value, it is determined that the eye is closed (S408), and if it is equal to or more than the threshold value, it is determined that the eye is open. (S410). The determination of whether the eyes are open or closed is sent to the above-mentioned drowsiness determination unit 16, and the drowsiness determination is performed here.

Re-Searching Routine FIG. 8 shows a control flowchart of the re-searching step S500. When the nostrils or eyes cannot be detected in the tracking step S200, the nostrils or eyes must be found promptly so as not to interfere with the snooze determination. Therefore, the nostrils and eyes cannot be detected again using the initial search routine. The initial search routine is intended to surely detect the nostrils and eyes even if it takes time as described above, and is not suitable for rediscovery in a short time. In other words, by providing the re-search routine separately, the nostrils and eyes are reliably searched over the initial search routine. The re-search routine has three types of search modes depending on the settings of the nostril search area and the eye search area. By selecting an appropriate mode, the search mode can be made faster and more accurately. Nostrils and eyes are discovered.

The first re-search mode is a mode corresponding to the case where nostrils are detected in the previous tracking mode but no eyes are detected. This mode is for when the movement of the face is not so great because the nostrils are detected, and it is mainly used when the eyes cannot be detected momentarily due to lighting conditions. . Therefore, the nostrils and eye search areas are also the narrow areas used in the tracking routine, that is, the area 20 in FIGS. 4C and 5B.
c, 22b are used.

The second re-search mode is a mode corresponding to the case where no nostrils were not detected in the previous tracking mode. At this time, there is a high possibility that the face is moving, and the search areas for nostrils and eyes are set larger than in the case of the tracking routine. That is, as shown in FIG. 4B, the nostril search region 20b in the second re-search mode is larger than the first re-search mode region 20c. However, it is smaller than the area 20a at the time of the initial search. Further, as shown in FIG. 5A, the eye search area 22a in the second re-search mode is larger than the area 22b in the first re-search mode and has the same size as the area in the first search. .

The third re-search mode is a mode corresponding to the case where the eye cannot be found even though the first re-search mode is searched for a predetermined time. In this case, the nostril search area is the same as the area 20c in the first re-search mode.
Is set, and only the eye search region is enlarged, and the same region 22a as in the second re-search mode is set.

When the re-search routine is started, the vertical acceleration is read from the acceleration sensor installed on the vehicle body (S502), and the vertical acceleration is compared with a predetermined threshold value (S504). When it is determined that the threshold value is not more than the threshold value, it is further determined whether or not the nostril has been detected in the previous tracking routine (S506). And
When there is a nostril, that is, when the nostril is detected in the previous tracking routine but the eyes (edge pair) cannot be detected in step S216, the first re-search mode is set (S508). . Then, the first re-search mode timer is set (S510). The set time at this time is the time to shift from the first re-search mode to the third search mode, and can be changed according to the driver's consciousness level. The driver's motion is fast in the awake state and becomes slower as the consciousness level is lowered. Therefore, if the consciousness level is lowered, it is unlikely to lose sight of the nostril even if the timer is set longer. The set time of the first re-search mode timer is
At a normal consciousness level, it is set to 200 msec, for example.

On the other hand, if the vertical acceleration of the vehicle body is larger than the predetermined threshold value in step S504, or if no nostril is not detected in the previous tracking routine, the second re-search mode is set. (S512).

When the first or second re-search mode is set in steps S510 and S512, it is determined whether the next re-search mode is the second search mode (S514). If not in the second re-search mode, the nostril search area 20c is set around the previous nostril position (S516). This area 20c is the same as the nostril search area of the tracking routine as described above. Then, binarization processing is performed in this area to extract a black area (S51).
8).

On the other hand, if it is determined in step S514 that the mode is the second re-search mode, the nostril search area 20b is set around the detected nostril position in the initial search (S520). This area 20b is, as described above,
It is set to be slightly larger than the area 20c in the tracking routine and the first re-search mode and smaller than that in the first search. Then, within this area 20b, nostril candidates are selected by template matching (S522), and binarization processing is performed on each candidate to extract a black area (S52).
4). As the threshold value during the binarization process, the value when the eye is found in the initial search routine is used, and the binarization process can be performed using an appropriate value.

Then, of the black areas extracted in steps S518 and S524, two black areas arranged substantially horizontally are selected as nostrils (S526). At this time, if the nostril cannot be selected (S528), the second search mode is set (S530), and the process returns to step S514. On the other hand, when it is determined in step S528 that there is a nostril, it is determined whether the first re-search mode is set (S532). If the first re-search mode is set,
The eye search area is set to the area 22b, that is, a small area (S534). If a mode other than the first search mode is set, the area 22a, that is, a large area is set (S536). Therefore, when the third re-search mode described later is set, the area 2 is set in step S536.
2a.

When the eye search area is set in steps S534 and S536, step S1 of the initial search routine is performed.
Edge extraction is performed in the same manner as 20 (S538). Then, in the case of the first search mode (S540), similar to steps S214 to S218 of the tracking routine, a positive / negative edge pair is searched (S542), and if there is an edge pair (S544), the position and nostril The position is stored (S546). On the other hand, if it is determined in step S540 that a mode other than the first re-search mode is set,
Similar to steps S122 to S126 of the initial search routine, positive and negative edge pairs are selected as eye candidates (S5).
48) If these candidates are arranged in an elliptical shape (S550), this candidate is detected as an eye, and its position and the position of the nostril are stored (S552). As described above, when the positions of the nostrils and the eyes are detected in steps S546 and S552, this routine ends.

On the other hand, if no eye is detected, the search is further continued. When it is determined in step S544 that there is no edge pair, the first re-search mode timer is counted down (S554), and it is determined whether this timer is 0 (S556). Then, when the timer is not 0, the process directly returns to step S514 and the search is again performed in the first re-search mode. Timer is 0
If so, the third search mode is set (S55).
8). The third re-search mode expands the eye search area when the eyes are not found within the predetermined time in the first search mode even though the nostrils are detected. The probability of detection increases. After setting the third re-search mode, the process returns to step S514 and the search is performed again. If the eyes cannot be detected in any mode other than the first re-search mode, steps S550 to S550 are performed.
Returning to 514, the rediscovery is performed.

As described above, in the re-search routine, the nostril and the nostril are detected depending on whether the nostril has been detected until the last time, and how long the eye is not detected even if the nostril has been detected. The size of the eye search area is changed. As a result, when it is considered that the face is not moving significantly, the search area is narrowed down so that rediscovery can be performed in a short time, while when it is considered that the face is moving significantly, the search area is expanded and the discovery probability is increased. Is increasing. Further, by providing a re-search routine independent of the initial search, it is possible to shorten the time required to find an eye during the re-search.

Estimation of eye position when face is tilted In this embodiment, the eye position is estimated based on the position of the nostril, and the eye search area is set around the estimated position. However, a method of estimating the position of the eyes when the driver tilts his / her face will be described below.

FIG. 9 shows the relationship between the position of the nostrils and the position of the eyes when the face is tilted. In the horizontal state (a), the eye position E ₀ (X _E0 , Y _E0 ) with respect to the midpoint N ₀ (X _N0 , Y _N0 ) of the two nostrils is at the distance r in the direction θ ₀ .
At this time, the distance r and the direction θ ₀ are expressed by the following equations.

## EQU1 ## r = {(X _E0 -X _N0 ) ² + (Y _E0 -Y _N0 ) ² } ^1/2 θ ₀ = tan ^-1 {(Y _N0 -Y _E0 ) / (X _E0 -X _N0 ). } When the face is tilted, the direction of the arrangement of the two nostrils is slightly shifted from the horizontal as shown in (b). In this case, the nostril position N
_{Even if 0} does not change, the eye position shifts. If the directions of the two nostrils deviate from the horizontal by an angle δ, the nostril position N
₁ (X _N1 , Y _N1 ) and the eye position E ₁ (X _E1 , Y _E1 ) have the relationship shown by the following equation.

[Equation 2] X _E1 = X _N1 + r · cos (θ ₀ −δ) Y _E1 = Y _N1 −r · sin (θ ₀ −δ) Estimate the eye position as in the above equation, and By setting the search area of, the eyes can be detected more reliably.

[Brief description of drawings]

FIG. 1 is a block diagram showing a schematic configuration of a drowsy driving prevention apparatus according to an embodiment of the present invention.

FIG. 2 is a basic control flowchart for blink determination and eye detection of the present embodiment.

FIG. 3 is a control flowchart of an initial eye search routine of this embodiment.

FIG. 4 is a diagram showing a search area for a nostril.

FIG. 5 is a diagram showing an eye search region.

FIG. 6 is a control flowchart of a tracking routine of this embodiment.

FIG. 7 is a control flowchart of a blink determination routine of this embodiment.

FIG. 8 is a control flowchart of a re-search routine of the present embodiment.

FIG. 9 is a diagram for explaining estimation of eye positions when the face is tilted in the present embodiment.

[Explanation of symbols]

10 infrared projector, 12 CCD camera, 14 blink recognition unit, 16 dozing determination unit, 18 alarm unit.

Continuation of the front page (51) Int.Cl. ⁷ identification code FI G06T 7/20 300 A61B 5/10 320B (56) Reference JP-A-8-300978 (JP, A) JP-A-61-45911 (JP, A) Japanese Patent Laid-Open No. 8-83344 (JP, A) Junichi Fukuda, Kunihiko Adachi, Makoto Nishida, Eisaku Akutsu, Development of technology for detecting unconsciousness, TOY OTA Technical Review, Japan, Toyota Motor Corporation Technology Management Department, 1995 May, Vol. 45 No. 1, PP. 34-39 (58) Fields investigated (Int.Cl. ⁷ , DB name) G06T ^7/ 00-7/60 A61B 5/107 A61B 5/117 B60K 28/06 G60T 1/00 280 JISST file (JOIS)

Claims (6)

(57) [Claims] 1. An imaging step of capturing a face of a person to obtain a face image, a nostril detection step of detecting a nostril from the face image, and an area for performing eye search based on the position of the nostril. a search region setting step of setting, the detection step of the eye for performing eye detection in a search area, have a, in the search area setting step, the nostrils detecting step
Of the search area according to the continuation situation of nostril detection in
A method for detecting eyes in a face image , the size of which is changed . 2. The method of detecting an eye in a face image according to claim 1, wherein in the eye detection, an edge whose brightness in the vertical direction changes sharply is extracted, and a portion where the edge is continuous in an approximately elliptical shape is detected. A method of detecting an eye in a face image, which is to be detected as an eye. 3. The method for detecting an eye in a face image according to claim 1, wherein in the eye position detection, a negative edge whose brightness sharply decreases from top to bottom and a brightness edge from top to bottom are detected. A method for detecting eyes in a face image, in which a positive edge that sharply rises is to extract edge pairs that are vertically adjacent. 4. The eye detection method for a face image according to claim 1, wherein in the search area setting step, the position of the search area is changed in accordance with a change in orientation of two nostrils. A method for detecting eyes in a face image, in which 5. The facial image according to any one of claims 1 to 4.
In the method of detecting an eye in an image, the nostril detection step
The captured face image according to a predetermined threshold.
Binarization is performed to detect the nostrils.
When the eye search is performed based on the position, the eye detection
If the eyes cannot be detected in the step, the binarization is performed.
The threshold for processing is changed and the nostril is detected again.
How to detect eyes in face images. 6. An imaging system for photographing a human face to obtain a face image.
And a nostril detection step for detecting a nostril from the face image.
And an area for eye search based on the position of the nostrils.
Setting the search area, and
An eye detecting step for performing detection, wherein the nostril detecting step determines the captured face image in a predetermined manner.
Detects nostrils by binarizing according to the threshold of
And the eyes are searched based on the position of the nostrils.
If the eye could not be detected in the eye detection step,
In this case, the threshold value for the binarization process is changed,
A method for detecting eyes in a face image, which detects the nostrils again .