KR101681938B1 - Apparatus for preventing drowsy driving - Google Patents

Abstract

The present invention relates to a drowsiness driving preventing device capable of accurately detecting the eyes of a driver without increasing the irradiating light or increasing the sensitivity of the camera unit and greatly reducing the amount of calculation for eye detection.
An apparatus for preventing drowsiness driving according to the present invention includes: an illumination unit that emits light in an infrared wavelength band; A prism lens for dispersing incident infrared rays in a first direction and a second direction; An infrared pass filter provided on one surface of the prism lens for blocking visible light and passing infrared light; A neutral density filter provided on the first direction optical path of the prism lens; An antireflective (AR) layer disposed on the second directional optical path of the prism lens; A first image pickup element for detecting light incident upon the neutral density filter and generating a first image; A second image pickup element for generating a second image by detecting light incident upon the non-reflective layer; And extracting the area including the eyes as a region of interest after recognizing the driver's face region from the first image generated by the first image pickup element, extracting from the region corresponding to the region of interest the eyes of the driver And a central processing unit for detecting the central processing unit.

Description

{APPARATUS FOR PREVENTING DROWING DRIVING}

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a drowsiness driving prevention system, and more particularly, to a drowsiness driving prevention device and an optical system for drowsiness driving prevention device capable of discriminating whether or not drowsiness is possible by using a near infrared ray camera and a near infrared ray camera.

Generally, since a plurality of vehicles run at the same time in each lane when the vehicle is running on the road, the driver must pay attention to the brake operation and the steering wheel operation so that the driver can run safely without colliding with the vehicle in the next lane.

However, when the driver does not know that the vehicle ahead is stopped due to drowsy driving and can not properly operate the brakes, a collision accident may be caused. In the case of steering the steering wheel to the right and left unintentionally during drowsiness operation, The vehicle enters the side lane and collides with another vehicle in the lane of the lane, failing to guarantee safe driving.

In order to prevent such drowsy driving, various technologies have been proposed. In general, the drowsiness driving prevention device detects an eye image from a driver image input from a camera and determines whether the drowsiness operation is performed using the detected eye image.

In addition, when it is determined that the driver is sleepy driving, a predetermined event such as a warning sound is generated to alert the driver, thereby preventing traffic accidents caused by sleepy driving.

For example, in the prior art 1, switching means for outputting an on / off signal, a camera for detecting a facial image of a driver, and a camera are analyzed to determine a drowsy operation according to the open / close rate of eyes and a control signal for alarm is output Which is capable of outputting an alarm signal during a drowsy operation through a configuration of a video signal operating unit and a control unit.

However, such a conventional drowsiness driving prevention apparatus or method has a problem in that when the general CCD or CMOS camera is used, the driving environment is dark, such as at night, or an image sufficient to discriminate the driver's eye condition when the driver wears sunglasses There is a problem that it is difficult to do.

In order to solve such a problem, there is a method of acquiring an image using an infrared camera (see prior patents 2 and 3). However, in this case, there is a limit in obtaining satisfactory images for various illuminance environments.

In order to solve the limitations of the prior patents 2 and 3, particularly in the case of sunglasses, the amount of light in the eye region transmitted to the infrared camera is very low, ranging from 2.25 to 6.25% In order to obtain the amount of light, the illumination should be 10 to 30 times brighter or the sensitivity of the camera lens and sensor (hereinafter, 'camera unit') should be increased by 10 to 30 times. When the influence of external light sources such as strong backlight, You have to attenuate the light and lower the light and sensitivity.

However, if the irradiated light is increased or the sensitivity of the camera unit is increased in order to acquire the image of the eye region through the sunglasses, another problem occurs as follows.

First, when the irradiation light of the illumination is increased, the size of the illumination, the current consumption, the heat generation, the price, and the like increase, and when the sensitivity of the camera unit is increased, the size and the price of the lens and the image sensor increase.

Next, when the irradiation light increase or the camera unit sensitivity increasing method is applied individually or simultaneously, it is possible to secure enough brightness to transmit the sunglasses to check the opening state of the eye, but image saturation occurs in the area other than the sunglasses generated at this time There is a problem.

In addition, when the sunlight is transmitted through the sunglasses to increase the irradiation light to detect the eye area, or when a problem occurs in the face recognition operation due to strong external light such as backlighting or spotlighting, the operation itself becomes impossible after a malfunction, It is not possible to determine the area of interest, which increases the amount of computation.

Previous Patent 1. Korean Patent No. 0295849 (May 30, 2001) Prior Patent 2. Korean Patent Publication No. 2006-0066807 (Jun. 19, 2006) Prior Patent 3. Korean Patent Publication No. 1997-0036013 (July 22, 1997)

It is an object of the present invention to provide a drowsiness driving prevention device capable of accurately discriminating whether or not drowsy driving is performed even when a driver's sunglasses is worn as well as various illuminance environments.

In particular, it is a principal object of the present invention to provide a drowsiness driving prevention device capable of detecting a driver's eye without increasing the irradiated light or increasing the sensitivity of the camera unit, and capable of reducing a calculation amount for eye detection.

According to an aspect of the present invention, there is provided an apparatus for preventing drowsiness in an eye, comprising: an illumination unit that emits light of an infrared wavelength band toward a face of a driver; A prism lens for dispersing the infrared rays incident on the driver after projecting in a first direction and a second direction; An infrared ray pass filter provided on a surface of the prism lens on which the infrared rays are incident, the infrared rays passing through the prism lens while blocking visible rays and passing infrared rays; A neutral density filter provided on the first direction optical path of the prism lens; An antireflective (AR) layer disposed on the second directional optical path of the prism lens; A first image pickup element for detecting light incident upon the neutral density filter and generating a first image; A second image pickup element for generating a second image by detecting light incident upon the non-reflective layer; And extracting the area including the eyes as a region of interest after recognizing the driver's face region from the first image generated by the first image pickup element, extracting from the region corresponding to the region of interest the eyes of the driver And a central processing unit for detecting the central processing unit.

According to the drowsiness prevention device of the present invention, it is possible to acquire an image suitable for distinguishing drowsiness not only in various illuminance environments but also when wearing sunglasses of a driver, and in particular, It is possible to accurately detect the driver's eye without increasing the sensitivity and further improve the processing speed for eye detection by greatly reducing the image analysis workload or calculation amount for eye detection.

1 is a block diagram of a drowsiness driving prevention device according to the present invention;
2 is a view showing a detailed configuration of an optical system according to the present invention;
3 is a view showing a wavelength band of light emitted by the illumination unit according to the present invention.
FIG. 4 is a perspective view illustrating an example in which the drowsiness driving prevention device according to the present invention is installed inside a vehicle.
5 (a) and 5 (b) illustrate a first image, a region of interest, and a second image obtained by the first and second image pickup devices of the present invention.

The terminology used herein is for the purpose of describing particular embodiments only and is not intended to be limiting of the invention. The singular expressions include plural expressions unless the context clearly dictates otherwise. In this specification, the terms "comprises" or "having" and the like refer to the presence of stated features, integers, steps, operations, elements, components, or combinations thereof, But do not preclude the presence or addition of one or more other features, integers, steps, operations, elements, components, or combinations thereof.

Also, in the present specification, the term " above or above "means to be located above or below the object portion, and does not necessarily mean that the object is located on the upper side with respect to the gravitational direction. It will also be understood that when a section of an area, plate, or the like is referred to as being "above or above another section ", this applies not only to the case where the other section is " And the like.

Also, in this specification, when an element is referred to as being "connected" or "connected" with another element, the element may be directly connected or directly connected to the other element, It should be understood that, unless an opposite description is present, it may be connected or connected via another element in the middle.

Also, in this specification, the terms first, second, etc. may be used to describe various components, but the components should not be limited by the terms. The terms are used only for the purpose of distinguishing one component from another.

In the following, preferred embodiments, advantages and features of the present invention will be described in detail with reference to the accompanying drawings.

FIG. 1 is a block diagram of an apparatus for preventing drowsiness according to the present invention, FIG. 2 is a view showing a detailed configuration of an optical system according to the present invention, FIG. 3 is a graph showing the wavelength band of light emitted by the illumination unit according to the present invention FIG. 4 is an example of an installation example in which the drowsiness driving prevention device according to the present invention is installed inside a vehicle, and FIG. 5 is a view showing an example in which the drowsiness prevention device according to the present invention is acquired by the first and second image pickup devices 31 and 32 An area of interest, and a second image.

1 to 5, the drowsiness prevention device according to the present invention includes an illumination unit 10, an optical system 20, an infrared camera 30, and a central processing unit 40, The optical system 20 includes a prism lens 21, an infrared ray passing filter 22, a neutral density filter 23, an anti-reflection layer 24, , And an objective lens (25).

The illumination unit 10 of the present invention includes a light emitting element that emits infrared rays, and the light emitting element is preferably configured to emit light in a near infrared ray wavelength range.

According to a preferred embodiment, the lighting unit 10 is installed in a vehicle interior so as to project infrared light to at least the face portion including the pupil of the driver. For example, the lighting unit 10 may be mounted on the front end region of the roof panel 3, May be mounted near the top region of the windshield (1), near the driver's sun visor (5), or near the room mirror.

According to a preferred embodiment, the illumination unit 10 is composed of a lamp capable of emitting near-infrared rays in a wavelength range of 760 to 1550 nm, and the lamp may be composed of a plurality of IR LED chips.

The illuminating unit 10 may be provided as an independent object and may be mounted on the front area of the driver or may be disposed integrally with the infrared camera 30 and disposed in the front area of the driver.

Meanwhile, the illumination unit 10 may further include an illumination lens, a light diffusion plate, a light diffusion film, a reflector, and the like, for the purpose of light uniformity of the projection illumination.

The illuminating unit 10 can be controlled on / off and output by the central processing unit 40 for extracting a region of interest and eye detection, which will be described later. A detailed description thereof will be given later.

The illumination unit 10 may include a plurality of infrared LED chips mounted on the periphery of the infrared camera 30. Preferably, a plurality of infrared LED chips may be arranged at equal intervals on the outer circumference of the infrared camera 30 so that uniform and sufficient illumination may be projected onto the driver's face.

The optical system 20 of the present invention has a structure that functions to acquire the first image P1 and the second image P2 which will be described later by dividing the incident light I1 into two lights, ), A prism lens 21, an infrared ray pass filter 22, a neutral density filter 23, and an anti-reflection layer 24

The objective lens 25 of the optical system 20 condenses the infrared light reflected after projecting onto the driver's face and provides it to the prism lens 21.

The objective lens 25 is disposed on the path of the light I1 incident on the prism lens 21 after being projected on the driver's face and is preferably disposed on the prism lens 21 ). ≪ / RTI >

The prism lens 21 of the optical system 20 transmits the infrared light I1 passing through the objective lens 25 in the first direction K1 and in the second direction K2 different from the first direction K1 Is a dispersion refracting configuration.

The infrared pass filter 22 is disposed on one surface (hereinafter referred to as a first surface) of the prism lens 21 through which the infrared light I1 reflected after being projected on the driver's face by the illumination unit 10 is incident It is deployed to block visible light and pass infrared rays.

According to a preferred embodiment, the infrared pass filter 22 is an optical pass band filter (OBPF) or optically passable filter (OBPF) that blocks visible light and allows near infrared rays to pass only near infrared rays of a specific band among the light I1 incident on the prism lens 21, And a low-pass filter (OLPF).

Here, the 'near infrared ray of a specific band' passed by the optical band pass filter (OBPF) may be any one of a plurality of bands classified in the near-infrared wavelength band. For example, the first band B1, The second band B2, or the third band B3.

The illumination unit 10 includes a first band B1, a second band B2, and a third band B3. The first band B1, the second band B2, and the third band B3 are separated by a narrow-band short wavelength of the near- And is selectively configured to emit light.

Here, the bandwidth of the first band B1 can be configured to be at most 760 to 880 nm in the range of 840 to 880 nm at the minimum, and the bandwidth of the second band B2 can be set at 760 to 1550 nm at the minimum in the range of 840 to 1020 nm And the bandwidth of the third band B3 can be configured to a maximum of 880 to 1550 nm in a range of a minimum of 920 to 1020 nm.

Therefore, when the infrared pass filter 22 is formed of an optical bandpass filter, the infrared pass filter 22 is a narrow band pass filter for passing only the light of the first band B1, A wide band pass filter for passing only the light of the second band B2 or a narrow band pass filter for passing only the light of the third band B3.

The light I1 incident on the first surface side of the prism lens 21 passes through the infrared ray passing filter 22 and is filtered with a near infrared ray component of a specific band and is then filtered by the prism lens 21 in the first direction K1 ) And the second direction (K2).

The neutral density filter (NDF) 24 of the optical system 20 functions to evenly reduce the amount of light for each wavelength of (near) infrared rays traveling in the first direction K1. That is, the neutral density filter (NDF) 24 is a structure for more accurately recognizing the facial area from the dynamic driver image. Specifically, the depth of field is shallow, the background outside the driver is distorted and the driver is emphasized. So that it can be extracted more accurately.

The neutral density filter (NDF) 24 is provided on the optical path which is incident on the prism lens 21 and proceeds in the first direction K1. Preferably, the neutral density filter 24 Plane. ≪ / RTI >

(Near) infrared rays that are incident on the first surface side of the prism lens 21 and are refracted in the second surface side (i.e., in the first direction K1) pass through the neutral density filter 23, And is then incident on the first image pickup device 31 to generate the first image P1. Then, the region of interest D1 to be described later can be extracted from the first image P1 thus generated.

The anti-reflection layer 24 of the optical system 20 is provided on the optical path that proceeds in the second direction K2. Preferably, the anti-reflection layer 24 of the optical system 20 is a third surface different from the above- (Near) infrared ray traveling in the second direction K2 may be formed on the surface through which the prism lens 21 is transmitted and externally emitted.

For reference, the second image P2, which will be described later, is an image used to substantially detect an eye at a portion D2 corresponding to the region of interest D1 acquired in the first image P1, P2 are generated from (near) infrared rays traveling in the second direction K2. Therefore, the noise of the second image P2 is blocked as much as possible, and a clearer image is required than the first image P1.

Specifically, infrared rays (near infrared rays) which are diffracted in the third surface side (i.e., the second direction K2) after being incident on the first surface side of the prism lens 21 are incident on the second image pickup element 32, Reflection layer 24 is formed so that the external light is reflected by the prism lens 21 when the (near) infrared ray in the second direction K2 is incident on the second image pickup element 32, Thereby preventing the noise from being introduced during the generation of the second image P2.

According to one embodiment, the non-reflective layer 24 is formed by coating a third layer of the prism lens 21 on the third surface of the prism lens 21 to absorb external incident light, And the like.

The infrared camera (30) of the present invention includes an image sensor capable of detecting an infrared ray projected on a face of a driver and generating an image. If the illuminating unit 10 is configured to emit light in the near infrared ray wavelength range, the infrared ray camera 30 is composed of a near infrared ray camera. Preferably, the near-infrared ray camera 30 may be a near-infrared ray CCD or a CMOS camera.

The infrared camera 30 is mounted in the driver's front area so as to capture at least the face area of the driver. The image signal obtained by the infrared camera 30 is input to the central processing unit 40. The central processing unit 40 receives the image signal and analyzes the image to determine whether the driver is drowsy.

In particular, the infrared camera 30 of the present invention is characterized by being divided into a first imaging device 31 and a second imaging device 32. [

Specifically, the first image pickup element 31 detects the light that is incident on the prism lens 21 and the neutral density filter 23 after passing sequentially through the first direction K1, that is, the light traveling in the first direction K1, P1).

The first image P1 obtained by the first image pickup device 31 is used for recognizing the driver's facial region and setting the region of interest D1.

The second image pickup device 32 detects the light that has passed through the prism lens 21 and the non-reflective layer 24 in the second direction K2 and then enters the second image K2 to generate the second image P2 do.

The second image P2 obtained by the second image pickup element 32 is used to substantially detect the driver's eye based on the region of interest D1 set by the first image P1.

The process of detecting the eyes by analyzing the first image P1 and the second image P2 obtained by the first image pickup device 31 and the second image pickup device 32 as described above is performed by a central processing unit (40).

Hereinafter, the central processing unit 40 determines whether or not the driver is drowsy based on the processing method for detecting eyes based on the first image P1 and the second image P2, .

The central processing unit 40 of the present invention firstly recognizes the driver's facial region from the first image P1 generated by the first image pickup device 31 and then sets the region including the eye to the region of interest D1 .

Here, the 'area including the eye' refers to the entire face or part of the driver's face, including the eye area including the pupil and the iris.

When the setting of the region of interest D1 is completed, the central processing unit 40 then exposes the region D2 corresponding to the region of interest D1 extracted from the first image P1 of the second image P2 And the eyes of the driver included in the second image P2 are detected by performing image control by focusing.

In other words, the central processing unit 40 can accurately extract the area of interest D1 by first recognizing the driver's face area on the weakened first image P1, and after extracting the area of interest D1, Is configured to detect an eye in another clear image corresponding to the region D1 (i.e., a portion D2 corresponding to the region of interest D1 in the second image P2).

Accordingly, it is possible to improve the eye detection accuracy of the central processing unit 40. In particular, only the region D2 matching the region of interest D1 in the second image P2 needs to be analyzed, The amount of computation is greatly reduced and the processing speed for eye detection can be improved.

When the detection of the driver's eyes is completed by the above-described processing, the central processing unit 40 analyzes this to determine whether or not the driver is drowsy.

Specifically, the central processing unit 40 compares the detected eye image with the normal-state eye region image stored in the memory 60, or analyzes the eye opening / closing (for example, opening or aperture ratio) It can be configured to determine if you are drowsy.

If it is determined that the driver is drowsy as a result of the determination, the warning event generator 50 is controlled to generate a warning to the driver. Here, the warning event generating unit 50 may employ at least one of a means such as a speaker, a vibration means, a cold air generator, or the like that can call a user's attention.

On the other hand, the central processing unit 40 of the present invention can be configured to further perform the following processing.

The central processing unit 40 recognizes the face after acquiring the first image P1 by limiting the illumination output of the illumination unit 10. [ If facial recognition fails at this time, the illumination output is changed to acquire an image again to perform facial recognition.

If the face recognition is successful, the second image P2 is acquired after extracting the region of interest D1 from the first image P1 and releasing the illumination output restriction. Then, the driver's eye is detected in an area D2 corresponding to the area of interest D1 among the second images P2.

At this time, if the eye detection fails, the illumination output is changed to acquire the second image P2 again and the eye is detected in the region D2 corresponding to the region of interest D1.

When the eye detection is successfully completed, it is possible to determine whether or not the driver is drowsy by analyzing the opening and closing of the eye (for example, whether or not the eye is open or the aperture ratio).

Thereafter, when it is determined that the user is drowsy, a warning event is generated, and then the process of acquiring the second image P2 is repeated and the process of detecting the eyes of the driver is continuously performed.

If the repetitive eye detection processing operation successively fails, the first image P1 acquisition and the interest area D1 setting operation after the illumination output limitation, and the second image P2 after the subsequent illumination output restriction release, To perform the eye detection performing operation of the camera.

As described above, when the first and second image acquiring processes according to the limitation and release of the illumination output are further included, there is an advantage that the eye can be detected more accurately.

While the preferred embodiments of the present invention have been described and illustrated above using specific terms, such terms are used only for the purpose of clarifying the invention, and it is to be understood that the embodiment It will be obvious that various changes and modifications can be made without departing from the spirit and scope of the invention. Such modified embodiments should not be understood individually from the spirit and scope of the present invention, but should be regarded as being within the scope of the claims of the present invention.

10: illumination part 20: optical system
21: prism lens 22: infrared pass filter
23: neutral density filter 24: non-reflecting layer
25: Objective lens 30: Infrared camera
31: First image pickup element
32: second image pickup element 40: central processing unit
50: Warning event generating unit P1: First video
P2: second image D1: region of interest
D2: region of interest matching

Claims (8)

An illuminator for emitting light of an infrared wavelength band toward a face of a driver;
A prism lens for dispersing the infrared rays incident on the driver side after projecting in a first direction and a second direction;
An infrared ray pass filter provided on a surface of the prism lens on which the infrared ray is incident, for blocking visible rays and passing infrared rays;
A neutral density filter provided on the first direction optical path of the prism lens;
An antireflective (AR) layer disposed on the second directional optical path of the prism lens;
A first image pickup element for detecting light incident upon the neutral density filter and generating a first image;
A second image pickup element for generating a second image by detecting light incident upon the non-reflective layer; And
Extracting a region including an eye as a region of interest after recognizing the face region of the driver from a first image generated by the first image pickup element, and extracting, from an area corresponding to the region of interest, And a central processing unit for detecting the drowsiness of the user.
The method according to claim 1,
The central processing unit,
Wherein when the extraction of the ROI is completed, the eye of the driver included in the second image is detected by performing an image control that focuses on an area corresponding to the ROI of the second image Drowsiness driving prevention device.
The method according to claim 1,
The central processing unit,
Acquiring the first image by limiting an illumination output of the illumination unit, extracting the ROI from the first image, recognizing the ROI,
And acquires the second image after releasing the illumination output limitation of the illuminating unit when the facial area recognition is successful and detects the eyes of the driver from an area corresponding to the area of interest of the second image A drowsiness driving prevention device.
The method of claim 3,
The central processing unit,
Wherein when the face region recognition fails, the illumination output is changed to acquire the first image again, and then the face region recognition is performed.
The method according to claim 1,
Further comprising an objective lens provided on the infrared pass filter.
The method according to claim 1,
Wherein the infrared pass filter is a band-pass filter that passes only a near infrared ray of a specific band of light incident on the prism lens.
The method according to claim 1,
Wherein the infrared ray emitted from the illumination unit is light in a near-infrared wavelength range.
The method according to claim 1,
The central processing unit compares the detected eye image with a steady-state eye region image stored in the memory or analyzes the opening and closing of the eye to determine whether the driver is drowsy. If it is determined that the driver is drowsy And generates a predetermined warning event.

Images