JP2008191785A - Eye closure detection apparatus, doze detection apparatus, eye closure detection method, and program for eye closure detection - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide an eye closure detection apparatus, a doze detection apparatus, an eye closure detection method, and a program for eye closure detection that can detect eye positions even if nostrils cannot be detected in an eye detection apparatus for estimating eye positions from nostril positions. <P>SOLUTION: The eye closure detection apparatus 10, which has a nostril position determination means 12b for determining nostril positions from feature points extracted from a face image and an opening/closing detection means 12d for setting eyeball tracking areas based on the nostril positions and searching for eyes in the eyeball tracking areas, is provided with a mouth position determination means 12c for determining a mouth position if nostrils cannot be detected. The opening/closing detection means 12d sets eyeball tracking areas based on the mouth position and searches for eyes in the eyeball tracking areas. <P>COPYRIGHT: (C)2008,JPO&INPIT

Description

  The present invention relates to an eye-closed detection device, a doze detection device, a closed-eye detection method, and a closed-eye detection program for detecting opening and closing of an eye from a face image. The present invention relates to a closed eye detection device, a drowsiness detection device, a closed eye detection method, and a closed eye detection program.

A dozing detection technology has been proposed that, when a driver's face is photographed by processing the driver's face while driving and detecting the driver's dozing, alerting the driver to return to the awake state (For example, refer to Patent Document 1). In Patent Literature 1, a nostril is detected as a feature of the driver's face, the eye position is estimated from the detected nostril, and eye opening / closing is detected in the eyeball tracking region set there. For example, when the eye-closing time continues for a predetermined time or more, it is determined that the driver's arousal level is low, and the driver is alerted with an alarm sound or the like.
Japanese Patent Laid-Open No. 10-86696

  In this way, the nostril is rarely covered with glasses, hair, etc., and it is relatively easy to detect because it is separated from the eyebrows, and since it does not open and close unlike the mouth, the eye tracking region is based on the nostril position. It can be said to be effective.

  However, when the driver is facing sideways, even when facing the front, if the face is exposed to direct sunlight from the side and the nostrils are shaded, the nostrils cannot be detected due to individual differences in the shape of the nose, etc. There is.

  In view of the above problems, the present invention provides a closed eye detection device that estimates the position of an eye from a nostril position, an closed eye detection device capable of detecting an eye position even when a nostril cannot be detected, a doze detection device, a closed eye detection method, and closed eye detection. It aims at providing the program for.

  In order to solve the above-described problems, the present invention searches for an eye in the eyeball tracking area by setting a nostril position determining means for determining a nostril position from the feature points extracted from the face image, and an eyeball tracking area based on the nostril position. And an open / close detection means that, when no nostril is detected, has a mouth position determination means for determining a mouth position, and the open / close detection means sets an eyeball tracking region based on the mouth position, and The eye of the tracking area is searched.

  According to the present invention, it is possible to detect an eye more smoothly when detection of a nostril is difficult by detecting a mouth that is detected more easily than an eye but has a lower detection accuracy than a nostril.

  The present invention also provides a nostril position determining means for determining a nostril position from a feature point extracted from a face image, an open / close detection means for setting an eyeball tracking area based on the nostril position and searching for an eye in the eyeball tracking area. When the nostril is not detected in the closed eye detection device, the mouth position determination means for determining the mouth position, the nostril position determination means determines the nostril position based on the mouth position, the open / close detection means An eyeball tracking area is set based on a nostril position determined based on the mouth position, and an eye in the eyeball tracking area is searched.

  According to the present invention, since the nose is easier to determine the center line of the face than the mouth, the detection accuracy can be improved by first detecting the nostrils from the mouth rather than detecting the eyes directly from the mouth There is.

  Further, in one aspect of the present invention, when the above-mentioned eye-closing detection device, eye-closing time measuring means for measuring a continuous eye-closing time based on the eye opening detected by the opening / closing detection means, and when the eye-closing time is a predetermined time or more And warning means for warning the driver.

  According to the present invention, the driver is warned when the eye-closing time continues for a predetermined time or more, so that the driver can be awakened when he falls asleep.

  To provide a closed eye detection device, a doze detection device, a closed eye detection method, and a closed eye detection program capable of detecting the position of an eye even when the nostril cannot be detected in the closed eye detection device that estimates the eye position from the nostril position. it can.

  Hereinafter, the best mode for carrying out the present invention will be described with reference to the accompanying drawings. The closed eye detection device 10 of the present embodiment determines the nostril position from the face image and detects the opening / closing of the eye in the eye tracking region determined based on the nostril position, but can determine the nostril position based on the driver's face orientation and the like. If not, the mouth position is detected and the eye tracking region is determined based on the mouth position, or the nostril is detected again based on the mouth position, and the opening / closing of the eye is detected based on the nostril position. Therefore, even when the nostril position is not detected, the opening / closing of the eyes can be detected by detecting the mouth position.

  FIG. 1 shows a system configuration diagram of a dozing detection device 1 including the closed eye detection device 10 of the present embodiment. The closed eye detection device 10 includes a face camera 11 and a face image processing ECU (Electronic Control Unit) 12. The closed eye detection device 10 further includes a collision determination ECU 13, a combination meter 14, and the like via a vehicle LAN such as a CAN (Controller Area Network). The doze detecting device 1 is configured by being connected to the brake ECU 15 and the millimeter wave radar device 16.

  The face camera 11 is disposed at a position facing the driver's face from the front or slightly below, for example, in the combination panel or on the steering column. The face camera 11 has, for example, a CMOS (Complementary Metal Oxide Semiconductor) or a CCD (Charge Coupled Device) photoelectric conversion element, and photoelectrically converts incident light according to its intensity to obtain a predetermined luminance gradation (for example, 256-tone) digital image (face image) is output.

  The face image processing ECU 12 performs image processing on the face image taken by the face camera 11 to indicate how much the driver's face is directed to the left and right with respect to the front direction, and how much the eyes are open. The eye opening (the distance between the upper eyelid and the lower eyelid) is sent to the collision determination ECU 13.

  The collision determination ECU 13 detects a side-viewing operation from the degree of face orientation and its duration, detects a drowsy operation from the eye opening and its duration, sounds an alarm sound from the combination meter 14, and lights a warning lamp To do. The collision determination ECU 13 determines the possibility of an obstacle collision based on the distance to the obstacle and the relative speed detected by the millimeter wave radar device 16, and emits an alarm sound when the possibility of the collision is high. When a collision is unavoidable, the brake ECU 15 is requested to forcibly brake the vehicle. Further, when the driver's face orientation degree is equal to or greater than a predetermined value or when closed eyes for a predetermined time or longer are continuously detected, the collision determination ECU 13 performs warning sound and vehicle braking early.

  The face image processing ECU 12 and the collision determination ECU 13 are a CPU that executes a program, a RAM that is a work area for executing the program and that temporarily stores data, and an EEPROM (Electronically Erasable and Programmable Read Only Memory) that retains data even when the ignition is turned off. ), An input / output interface serving as a data interface, a communication controller that communicates with other ECUs, a ROM that stores programs, and the like are configured by a microcomputer connected by a bus.

  When the CPU executes a program (a closed eye detection program in claims) 17, a face direction detection unit 12 a that detects the degree of face orientation based on the front direction, and a nostril position determination unit that determines the nostril position 12b, mouth position determining means 12c for determining the mouth position, and opening / closing detecting means 12d for searching for an eye in the eyeball tracking region and detecting opening / closing. Further, the CPU of the collision determination ECU 13 executes the program, thereby realizing the eye-closing time measuring unit 13a that measures the continuous eye-closing time.

[Detection of eye opening and closing]
First, detection of opening and closing of the eye when a nostril is detected will be described. Face orientation and eye opening / closing can be detected by the following well-known methods.

  FIG. 2A shows an example of a face image. The closed eye detection device 10 sequentially identifies the face outline, center line, nostril position, and eye position from the sequentially input face images, and sets a predetermined eyeball tracking region including the eye position. Once the eyeball tracking area is set, the detection of the opening and closing of the eyes in the eyeball tracking region is repeated, and when the state where the opening and closing of the eyes is not detected continues, the detection is performed again from the face outline.

  First, the face direction detection means 12a detects the face outline as the approximate position of the face from the face image. For example, the face orientation detection unit 12a stores a difference between pixel values of pixels having different pixel values (luminance) more than a predetermined value between the sequentially input face images together with the position information, and the predetermined number of face images. The difference between the pixel values at the same position is counted, a histogram in the vertical direction of the counted values is created, and the peak of the integrated value is set as the facial contour position in the horizontal direction. Although the background and the like are shown in the face image, since the background is stationary, it is possible to detect the outline of the face in the horizontal direction from the left and right ends of the region where the pixel value changes.

  The contour in the vertical direction of the face is detected from the edge information. Based on the edge information, pixels having a large luminance change compared to the skin, such as facial parts, eyebrows, eyelids, nostrils, mouth corners, and upper and lower lip borders, are detected. For example, Sobel's edge detection algorithm is used to extract edge information, thereby obtaining the contour of each part of the face surrounded by two types of edge information of low brightness to high brightness and high brightness to low brightness.

  In addition, since the parts such as the eyes and nose of the human face are arranged symmetrically, the face direction detection means 12a detects the center line of the face so that the number of left and right edge information is substantially equal. Since the relative position of the center line and the contour in the horizontal direction changes when the driver turns to the horizontal direction, the face direction can be tracked by monitoring the center line.

  Then, a black pixel continuous region (region surrounded by edge information), for example, on the left side from the center line of the obtained face is scanned from the upper side (eyebrow side) as scanning lines a to c. Detect eyebrows. Since the eyebrows may be considered to be bilaterally symmetric, if there is a continuous region of similar black pixels on the opposite side of the center line, the right and left eyebrow positions are determined as the upper contour of the face. Further, edge information that is lower than the eyebrow position and crosses the center line of the face and continues in the horizontal direction more than a predetermined amount is detected, and the edge information is detected based on the width, strength, etc. of the edge information. It is considered as a border and determined as the lower contour of the face. In this way, the face contour position is obtained.

  Next, the nostril position determining means 12b detects the nostril position. The nostril position is detected in order to set the eye tracking area, but the eye tracking area can be set with reference to the mouth position. However, as will be described later, when detecting the position of the eyelid, a change in the position of the upper eyelid due to blinking is used, but the mouth is also opened and closed by conversation, breathing, etc., so it is difficult to distinguish from the eyelid. Further, the mouth position is an edge in the horizontal direction, but the face image has many portions having edge information continuous in the horizontal direction, such as eyebrows, eyes, and lower nose.

  On the other hand, since there are two nostrils symmetrically with respect to the center line, it is considered that the accuracy when detected is higher than the mouth position. Moreover, the mouth position is determined where for a long edge laterally or transversely of the center becomes difficult, nostril position so good a midpoint of two nostrils as the central lateral, right eye tracking region It is easy to set in a convenient place.

  The nostril position determining means 12b sets a slightly longer nostril detection region that passes through the center line of the face above the lips. The reason for making the nostril detection region vertically long is to easily eliminate the influence of moles and light, and to reduce the burden of image processing when detecting nostrils by pattern matching. The nostril position determining means 12b detects a nostril by extracting feature points having nostril characteristics such as shape, size, and luminance. For example, the brightness of the nostril detection area is clarified by binarization processing, etc., whether pattern matching with the standard pattern of either the right or left nostrils, or whether a block of black pixels with a predetermined area or more is detected, etc. Detect either the right or left nostril. When one nostril is detected, a nostril that is substantially line-symmetric with respect to the center line is detected. When two nostrils are detected horizontally, the nostril position is determined.

  When the nostril position can be determined, the open / close detection means 12d uses the statistical data on the relationship between the nostril and the eye position to set a predetermined area for the nostril position as the eye tracking area. FIG. 2C is a diagram for explaining the eyeball tracking area. If the face is in a horizontal state, two nostrils are arranged horizontally, and the position of the eyeball with respect to the midpoint is designated by the distance r and the direction θ. Since the range of the distance r and the range of the direction θ where the eyeball is present at a predetermined ratio or more are clear from the statistical data, the open / close detection unit 12d sets a rectangular area covering this as the eyeball tracking area.

  The opening / closing detection means 12d detects the opening / closing of the eye by monitoring the upper eyelid and lower eyelid of the eyeball tracking area. The open / close detection means 12d acquires edge information from the eyeball tracking area, searches the pixel array at the upper left vertex of the eyeball tracking area for an edge from luminance large to small (skin color to wrinkle) from top to bottom, and the edge Can be searched, the same search is performed for the right pixel column. Therefore, if this search is completed to the right end, the upper pixel position is obtained. Similarly, the open / close detection means 12d acquires the lower eyelid pixel position.

  Further, the open / close detection means 12d determines whether or not the upper eyelid is detected depending on whether or not the position of the upper eyelid thus detected changes greatly in the face images within a predetermined number. This is because the driver blinks within a predetermined time, so it is known that the position of the upper eyelid moves more greatly than other eye parts (for example, lower eyelids, eyebrows and eyeglass frames). This is because it is considered that the position of the upper eyelid is accurately detected when the position of the eyelid moves greatly.

  Next, the open / close detection means 12d calculates the difference (up and down distance) between the upper and lower eyelid pixel positions in order from the left, and sets the maximum number of pixels as the eye opening. Even when the driver closes his eyes, the eye opening is detected from the upper and lower eyelid positions through the same process.

  The open / close detection means 12d calculates the difference (up and down distance) between the upper and lower eyelid pixel positions in order from the left, and sets the maximum number of pixels as the eye opening. Even when the driver closes his eyes, the eye opening is detected from the upper and lower eyelid positions through the same process.

  The opening / closing detection means 12d detects the opening degree of the eye from the face image taken every cycle time, and sequentially sends it to the collision determination ECU 13. Then, the eye-closing time measuring means 13a of the collision determination ECU 13 compares the threshold for determining the opening and closing of the eye with the opening of the eye, and determines that the eye is open when it is greater than or equal to the threshold and closed when it is less than the threshold. When the closed eyes continue, the closed eye time is continuously measured.

  It is preferable to set the eye tracking area from the position of the nostril, but if the driver turns sideways more than the specified or if the nostrils are shaded by the influence of light even if they are facing the front, an individual with a nose shape A nostril may not be detected due to a difference or the like. Therefore, in the present embodiment, a description will be given of the closed eye detection device 10 that detects the mouth position and sets the eye tracking region using statistical data based on the mouth position when the nostril position cannot be determined.

  FIG. 3 shows an example of a face image after image processing in which a center line or the like is detected. The face image in FIG. 3 is obtained by extracting edge information after binarization and detecting the outline of the face, and the area other than the outline (hatched area) is masked with black pixels because it is unnecessary for detection of closed eyes.

  In FIG. 3, the contours of the facial parts of the eyebrows 31, the eyelids 32, and the mouth 33 are detected, but the contours of the nostrils 34 (shown by dotted lines) that should be detected are not detected. As described above, when the nostril position determining unit 12b cannot determine the nostril position, the mouth position determining unit 12c determines the position of the mouth 33.

  FIG. 4A shows a diagram for explaining the determination of the mouth position. Since the contour and the center line are detected by the face direction detection unit 12a, the mouth position determination unit 12c sets a horizontally long mouth position detection region near the lower contour. Since the mouth position can be detected from the brightness change occurring in the upper lip, the lower lip, and the boundary thereof, the mouth position determining means 12c obtains edge information by differentiating the brightness value of the mouth position detection area in the vertical direction. Then, an edge is searched from the pixel row at the upper left vertex of the mouth position detection area from the top to the bottom, and the detected edge information and its intensity are stored for each pixel. When this is performed up to the rightmost pixel row, the continuous edge in the horizontal direction, the width, and the intensity generated by the luminance change such as the upper lip, the lower lip, and the boundary thereof are acquired.

  In FIG. 4A, an upper lip edge 35, a lip border edge 36, a lower lip edge 37, and a shadow edge 38 caused by a shadow under the lips are detected. The mouth position determining means 12c detects the edge (lip border edge 36) that is preferably the longest and the largest in width and strength from a plurality of edges in the mouth position detection region, and if there is no such edge, a predetermined priority (for example, The lip border edge 36 is detected at the longest edge). Note that the lip is an area characterized by a vermilion color with respect to the surrounding skin color, so the mouth position may be specified using color information.

  The mouth position determining means 12c determines, for example, the intersection of the center line and the lip border edge 36 as the mouth position (the center of the mouth in the lateral direction). In addition, since the edge part of the edge in the horizontal direction may be unclear, the mouth position may be determined from the average of the intersections of a plurality of edges having a predetermined length or more and the center line.

  When the mouth position is determined, the closed eye detection unit 12d sets an eyeball tracking region using statistical data on the relationship between the mouth position and the eyeball position. FIG. 4B is a diagram illustrating an example of the relationship between the mouth position and the eyeball position. The position of the eyeball with respect to the mouth position is specified by the distance D / direction a, and the range of the distance D / direction a where the eyeball is present at a predetermined ratio or more is clear from the statistical data. A rectangular area that covers the eyeball tracking area is set.

  Thereafter, the opening / closing detection means 12d can detect the opening / closing of the eye from the eyeball tracking area, as in the case where the nostril position is determined.

  FIG. 5 shows an example of a flowchart in which the closed eye detection device 10 detects opening / closing of the eye using the mouth position when the nostril position cannot be determined.

  The face orientation detection means 12a detects the contour and the center line from the face image taken by the face camera 11 (S1).

  Next, the nostril position determining means 12b sets a vertically elongated nostril detection region, and determines whether or not the nostril can be detected using the characteristics of the nostril (S2). In this determination, it may be determined that the nostril cannot be detected when two or one of the two nostrils cannot be detected, and it may be difficult to detect the nostrils when the face orientation is equal to or greater than a predetermined value. Since it is known, it may be determined that the nostril cannot be detected without detecting one or two nostrils when the face orientation is greater than or equal to a predetermined value. Thereby, the burden of image processing is reduced.

  When the nostril is detected (Yes in S2), the open / close detection means 12d sets an eyeball tracking region based on the nostril position (S3), detects eyelids from the eyeball tracking region, and detects the eye from the distance between the upper and lower eyelids. The opening degree is detected (S4).

  When the nostril is not detected (No in S2), the mouth position determination unit 12c sets the mouth position detection region using the face outline and the center line (S5). The mouth position determination unit 12c determines the mouth position using a horizontally long edge such as the lip border edge 36, and the closed eye detection unit 12d sets the eye tracking region using statistical data on the relationship between the mouth position and the eyeball position. (S6). Similarly to the case where the nostril is detected, the open / close detection means 12d detects eyelids from the eyeball tracking region, and detects the eye opening from the distance between the upper and lower eyelids (S4).

  In the closed eye detection device 10 of the present embodiment, the detection accuracy is lower than that of the nostril, but when the nostril is difficult to detect directly, the opening of the eye is detected after detecting the mouth position easier than detecting the eyeball. Thus, the opening degree of the eyes can be detected more smoothly.

  In the first embodiment, the eye tracking region is set by using statistical data based on the mouth position. However, in this embodiment, after determining the mouth position, the nostril detection region is set based on the mouth position and the nostril is detected therefrom. In addition, the closed eye detection device 10 that detects the opening degree of the eye from the determined nostril position will be described.

  If the mouth position and the center line are known, the estimation accuracy of the nostril position is also improved. Therefore, the threshold value for detecting the feature point corresponding to the nostril can be lowered, and the nostril can be easily detected.

  FIG. 6 is a diagram illustrating an example of the relationship between the mouth position and the nostril position. Although it is the same in that the nostril detection region is set from the mouth position using the statistical data, the distance L1 between the mouth position and the nostril in the vertical direction is shorter than the position between the eye and the nostril, and the individual difference is also small. Therefore, the vertical width W1 of the nostril detection region can be reduced. Further, the lateral distance L2 between the mouth position and the nostril may be within a predetermined distance from the center line with little individual difference, but there may be an error in the position of the detected center line when the face orientation increases. Further, since the relative position of the center line and the left and right nostrils changes, the width W2 in the lateral direction of the nostril detection region is set to be approximately the same as W1. Therefore, after the mouth position is detected, the vertically elongated nostril detection region can be reduced.

  Then, after setting the nostril detection region by narrowing the nostril position in this way, the risk of false detection of the nostril due to the influence of moles and light is reduced, so the threshold for nostril detection can be lowered.

  Here, the threshold value is a threshold value for binarizing the nostril detection region, for example. In this case, it is easy to detect a feature point by a nostril that should be photographed with a small luminance value (dark) by lowering the threshold value. become. Also, in the pattern matching process, when the degree of coincidence when the pixel values match most within the nostril detection region is greater than or equal to a predetermined value, it is determined that there is a nostril there. Therefore, it becomes easy to detect a nostril that is not photographed in a normal form. In addition, when detecting a block of black pixels having a predetermined area or more, if a predetermined number of pixels (for example, N pixels) are processed, the black pixels are considered to be continuous even if they are not continuous. When lowering the pixel threshold, for example, if it is 2 × N pixels, it is considered that the black pixels are continuous even if the black pixels are not continuous. It is possible to easily detect a block of black pixels having a predetermined area or larger when the image is taken.

  Even when the threshold value is lowered, for example, when the original threshold value is 1.0, by setting a threshold value of about 0.7 as the lower limit, it is possible to prevent erroneous detection of moles and shadows.

  If the nostril is detected from the mouth position in this way, the midpoint of the left and right nostrils can be obtained, so the center line of the face can be determined more easily than the mouth position, and as a result, the eye tracking region can be set with high accuracy. It is also possible to detect the degree of opening with high accuracy.

  FIG. 7 shows an example of a flowchart for detecting the nostril using the mouth position when the closed eye detection device 10 cannot determine the nostril position and detecting the opening and closing of the eye based on the nostril position. In FIG. 7, the same steps as those in FIG.

  The face orientation detection means 12a detects the contour and the center line from the face image taken by the face camera 11 (S1).

  Next, the nostril position determining means 12b sets a vertically elongated nostril detection region, and determines whether or not the nostril can be detected using the characteristics of the nostril (S2). In this determination, it may be determined that the nostril cannot be detected when two or one of the two nostrils cannot be detected, and it may be difficult to detect the nostrils when the face orientation is equal to or greater than a predetermined value. Since it is known, it may be determined that the nostril cannot be detected without detecting one or two nostrils when the face orientation is greater than or equal to a predetermined value. Thereby, the burden of image processing is reduced.

  When the nostril is detected (Yes in S2), the open / close detection means 12d sets an eyeball tracking region based on the nostril position (S3), detects eyelids from the eyeball tracking region, and detects the eye from the distance between the upper and lower eyelids. The opening degree is detected (S4).

  When the nostril is not detected (No in S2), the mouth position determination unit 12c sets the mouth position detection region using the face outline and the center line (S5). The mouth position determination means 12c determines the mouth position using a horizontally long edge such as the lip border edge 36.

  The nostril position determining means 12b uses the statistical data of the relationship between the mouth position and the nostril position to set a nostril detection area with a narrowed area, and to detect the nostril by lowering the threshold for detecting the nostril. (S10).

  Next, returning to step S2, the nostril position determining means 12b determines whether or not the nostril can be detected (S2). It should be noted that the determination in step S2 provides an upper limit for the number of times, thereby preventing the nostril detection from being repeated indefinitely. Further, each time it is repeated, the nostril detection region and the threshold for nostril detection may be changed so that the nostril can be easily detected.

  When the nostril is detected in step S2 (Yes in S2), the opening / closing detection means 12d can detect the opening degree of the eye (S4).

  According to the present embodiment, when the nostril is not detected in the normal nostril detection region and the threshold value, the mouth position is determined and the nostril is detected again from the mouth position. Therefore, the eye tracking region is set from the midpoint of the left and right nostrils. The opening degree of the eye can be detected with high accuracy.

  As described above, the closed eye detection device 10 according to the present embodiment determines the mouth position even when the nostril cannot be detected, sets the eye tracking region from the mouth position, or tries to detect the nostril again from the mouth position and confirms the nostril. Since the eye tracking region is set from the position, it is possible to detect the closed eye even when the nostril is difficult to detect.

It is a system configuration figure of a dozing detection device containing a closed eye detection device. It is a figure for demonstrating the detection of opening and closing of eyes from a face image. It is a figure which shows an example of the face image after the image process which detected the centerline etc. FIG. It is a figure for demonstrating determination of a mouth position. It is an example of the flowchart figure in which a closed eye detection apparatus detects opening and closing of eyes using a mouth position when a nostril position cannot be decided. It is a figure which shows an example of the relationship between a mouth position and a nostril position. It is an example of the flowchart figure which detects a nostril using a mouth position when a closed eye detection apparatus cannot determine a nostril position, and detects opening and closing of an eye on the basis of a nostril position.

Explanation of symbols

DESCRIPTION OF SYMBOLS 1 Dozing detection apparatus 10 Eye-closed detection apparatus 11 Face camera 12 Face image processing ECU
12a Face orientation detection means 12b Nostril position determination means 12c Mouth position determination means 12d Open / close detection means 13 Collision determination ECU
13a Eye-closing time measuring means 14 Combination meter 15 Brake ECU
16 Millimeter wave radar device 17 Program

Claims (5)

Closed eye having nostril position determining means for determining a nostril position from feature points extracted from a face image, and open / close detection means for setting an eyeball tracking region based on the nostril position and searching for an eye in the eyeball tracking region In the detection device,
When nostrils are not detected, it has mouth position determination means for determining the mouth position,
The open / close detection means sets an eyeball tracking region based on the mouth position, and searches for an eye in the eyeball tracking region;
A closed eye detection device characterized by the above. Closed eye having nostril position determining means for determining a nostril position from feature points extracted from a face image, and open / close detection means for setting an eyeball tracking region based on the nostril position and searching for an eye in the eyeball tracking region In the detection device,
If no nostril is detected, has mouth position determining means for determining the mouth position;
The nostril position determining means determines the nostril position with reference to the mouth position determined by the mouth position determining means;
The open / close detection means sets an eyeball tracking region based on the nostril position determined based on the mouth position, and searches for an eye in the eyeball tracking region;
A closed eye detection device characterized by the above. The closed eye detection device according to claim 1 or 2,
Eye-closing time measuring means for measuring a continuous eye-closing time based on the opening degree of the eyes detected by the opening / closing detection means;
Warning means for warning the driver when the eye-closing time is a predetermined time or more;
A dozing detection device characterized by comprising: Nostril position determining means determines the nostril position based on the feature points extracted from the face image;
If the nostril position is not determined, the mouth position determining means determines the mouth position;
An open / close detection means sets an eyeball tracking region based on the mouth position, and searches for an eye in the eyeball tracking region; and
A closed-eye detection method comprising: Nostril position determining means determines the nostril position based on the feature points extracted from the face image;
If the nostril position is not determined, the mouth position determining means determines the mouth position;
An open / close detection means sets an eyeball tracking region based on the mouth position, and searches for an eye in the eyeball tracking region; and
A program for detecting closed eyes, which causes a computer to execute.

Images