KR101241225B1 - Image recognition system and method for car - Google Patents

Abstract

An image recognition system mounted on a vehicle to capture an image for recognizing road information may include a camera for photographing a road on which the vehicle travels at one or more predefined locations of the vehicle, and an image acquisition for obtaining one or more road images through the camera. The exposure control unit may recognize the road information through the road image, and the vehicle control unit for controlling the operation of the vehicle according to the road information and the image acquisition unit periodically adjust the exposure value of the camera so as to acquire a plurality of road images having different brightness. Include.

Description

Vehicle image recognition system and method thereof {IMAGE RECOGNITION SYSTEM AND METHOD FOR CAR}

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an image recognition system and a method thereof, and more particularly, to an image recognition system and method for photographing an image mounted on a vehicle for recognizing road information.

In the case of the camera technology installed on the vehicle, the image acquired through the camera suddenly becomes too bright or too dark in a bad condition where the ambient illumination is suddenly changed, causing frequent recognition of the lane and other information. This is because the camera's exposure (image brightness) adjustment does not immediately respond to the sudden change in ambient light.

Here, examples of a sudden change in ambient illuminance may include various conditions such as when entering / exiting a tunnel, driving under an overpass, direct sunlight, or direct headlights.

This problem leads to a decrease in overall camera recognition performance, and at the same time, when applied to safety systems such as collision avoidance and lane recognition, it results in lower reliability and increased risk.

Accordingly, an object of the present invention is to provide an image recognition system and method for a vehicle that alternately acquire images having different exposure values or image brightness of a camera.

According to one aspect of the invention there is provided an image recognition system. The system includes an image recognition system mounted on a vehicle for capturing an image for recognizing road information, the system comprising: a camera for photographing a road on which the vehicle travels at one or more predefined locations of the vehicle; An image acquisition unit obtaining one or more road images through the camera; A vehicle control unit recognizing the road information through the road image and controlling the operation of the vehicle according to the road information; And an exposure controller configured to periodically adjust an exposure value of the camera to acquire a plurality of road images having different brightnesses.

According to another feature of the invention there is provided an image recognition method. The method includes a method of recognizing an image of a video recognition system mounted on a vehicle to capture an image for recognizing road information, the method comprising: periodically adjusting an exposure value of a camera; And acquiring a road image according to the adjusted exposure value of the camera.

According to an exemplary embodiment of the present invention, by acquiring images having different exposures (or image brightnesses) at the time of acquiring a camera image, the difference of information including the lanes in a section (or moment) in which the peripheral brightness (illuminance) changes abruptly. Recognition problems can be solved to improve the overall camera information recognition rate.

In addition, in the situation where the brightness of the light source should be distinguished, it may be used to classify the light source brightness by acquiring images of different illuminances rather than uniform exposure values. In other words, when the surrounding illumination suddenly changes due to surrounding facilities (tunnels, under an overpass, surrounding skyscrapers, etc.), it is possible to solve a problem of unrecognized road information (lanes, signs, etc.) that occurred when the illumination suddenly changed. This can improve overall camera performance and recognition rate.

In addition, when the amount of light directly incident on the camera increases (eg, headlamps or surrounding lighting systems of opposite cars), cross-acquisition of dark images can prevent unrecognition and misunderstanding and improve the recognition rate, resulting in high beam assist (HBA) performance. Can improve.

Also, if it is difficult to detect without brightness adjustment regardless of ambient lighting (when distinguishing headlights, streetlights, and sign reflectors), cameras can be recognized by detecting the light source and the intensity of light by cross-acquiring a darker image at night. It can improve performance.

In addition, it is possible to cope by adjusting the optimum exposure after a certain number of frames or a time lapse in preparation for the case where the sudden change in brightness is maintained while continuously acquiring images having different exposure and brightness.

As such, according to embodiments of the present invention, camera performance is ultimately improved, which leads to an improvement in system reliability, convenience, and safety.

1 is a block diagram showing the configuration of an image recognition system according to an embodiment of the present invention.
2 illustrates an image frame according to an embodiment of the present invention.
3 illustrates the types of image frames according to an embodiment of the present invention.
4 is a flowchart illustrating an image recognition method according to an embodiment of the present invention.

DETAILED DESCRIPTION Hereinafter, exemplary embodiments of the present invention will be described in detail with reference to the accompanying drawings so that those skilled in the art may easily implement the present invention. The present invention may, however, be embodied in many different forms and should not be construed as limited to the embodiments set forth herein. In the drawings, parts irrelevant to the description are omitted in order to clearly describe the present invention, and like reference numerals designate like parts throughout the specification.

Throughout the specification, when an element is referred to as "comprising ", it means that it can include other elements as well, without excluding other elements unless specifically stated otherwise.

Also, the term "part" in the description means a unit for processing at least one function or operation, which may be implemented by hardware, software, or a combination of hardware and software.

Hereinafter, an image recognition system and a method thereof for a vehicle according to an exemplary embodiment of the present invention will be described in detail with reference to the accompanying drawings.

1 is a block diagram showing a configuration of an image recognition system according to an embodiment of the present invention, Figure 2 shows an image frame according to an embodiment of the present invention, Figure 3 is a type of image frame according to an embodiment of the present invention Indicates.

Referring to FIG. 1, the image recognition system 100 includes a camera 101, an image acquisition unit 103, an image processing unit 105, a vehicle control unit 107, and an exposure control unit 109.

At least one camera 101 is mounted at a specific position of the vehicle to photograph a road on which the vehicle travels.

The image acquisition unit 103 acquires one or more road images through the camera 101.

The image processing unit 105 recognizes one or more road images acquired by the image acquisition unit 103 through lanes, preceding cars, opposing cars, road information, and the like through predefined image processing techniques.

The vehicle controller 107 controls the vehicle based on the information recognized by the image processor 105. For example, lane recognition using a camera, lane departure warning, lane maintenance assistance system, road information (sign) recognition and display, preceding vehicle and opposing vehicle recognition, collision avoidance, high beam control, and the like can be performed.

The exposure controller 109 is connected to the camera 101 and the image acquisition unit 103 and periodically adjusts the exposure value so that the image acquisition unit 103 of the camera 101 acquires a plurality of road images having different brightness.

At this time, the exposure control unit 109 adjusts the exposure value of the camera 101 so that the optimum frame, the bright frame, the optimum frame, and the dark frame can be obtained by crossing over time. That is, the reference exposure value for acquiring the optimal road image, the first exposure value for acquiring the road image brighter than the optimal road image, and the second exposure value for acquiring the road image darker than the optimal road image are periodically provided to the camera 101. do.

In this case, alternately provided order, as in Figure 2, the reference exposure value-> first exposure value-> reference exposure value-> second exposure value-> reference exposure value-> first exposure value... The order of may be set to repeat repeatedly.

Here, although the first exposure value is disposed before the second exposure value, the second exposure value may be disposed before the first exposure value.

In addition, the order may be variously arranged, but is not limited to the above.

In addition, the exposure controller 109 may provide the reference exposure value, the first exposure value, and the second exposure value to the camera 101 for a predetermined time to acquire a plurality of image frames according to each exposure value. That is, the optimal exposure image, the bright exposure image, and the dark exposure image may be acquired in units of 1 second each.

In this case, the first exposure value may be set to one exposure value.

Alternatively, the first exposure value may be configured of a plurality of first exposure values in which the exposure value is gradually increased so that the brightness becomes brighter in stages. In this case, a plurality of bright image frames, which are gradually brighter than the optimal image frame, are generated.

In addition, the second exposure value may also be set to one exposure value.

Alternatively, the second exposure value may also be configured as a plurality of second exposure values in which the exposure value is gradually decreased so that the brightness becomes dark step by step. In this case, a plurality of dark image frames are generated, the brightness of which becomes darker than the optimal image frame.

 Meanwhile, according to FIG. 3, the optimum frame is an exposure compensation frame suitable for ambient illumination, and is used for recognizing road information in a general situation. Therefore, the reference exposure value is set so that the camera 101 can obtain an optimal image frame.

The bright frame is a brighter frame than the optimal frame, and is used for recognizing road information in a section in which peripheral illuminance is sudden. Therefore, the first exposure value is set to a value larger than the reference exposure value.

The dark frame is a darker frame than the optimum frame, and is used for recognizing road information in a situation in which ambient illumination is rapidly increasing and brightness needs to be distinguished. Therefore, the second exposure value is set to a value smaller than the reference exposure value.

In addition, the exposure controller 109 may analyze the image frame acquired by the image acquirer 103 to determine whether the threshold condition is met, and change the reference exposure value when the threshold condition is met. Here, the threshold condition may be set to the number of images so that the reference exposure value may be automatically changed when the total number of images exceeds a predetermined number or more. In addition, the threshold condition may be set to a shooting time of the camera so that the reference exposure value may be automatically changed after a certain time.

The image recognition process of the image recognition system 100 described above is sequentially illustrated in FIG. 4. At this time, it will be described in connection with the configuration of Figure 1, and the same reference numerals as in FIG.

4 is a flowchart illustrating an image recognition method according to an embodiment of the present invention.

Referring to FIG. 4, the camera 101 is turned on (S101) to set a predetermined optimal exposure value (reference exposure value) as a camera exposure value (S103) to photograph a road. Then, the image acquisition unit 103 obtains the optimal exposure image (S105).

At this time, the exposure control unit 109 determines whether the exposure value adjustment cycle has arrived (S107). That is, it may be determined whether a specific time has passed or whether the number of optimal exposure image frames exceeds a specific number. However, the exposure value adjustment cycle is not limited to the above, and various methods and conditions can be set.

If the exposure value adjustment cycle has not arrived at step S107, the exposure controller 109 performs step S105 again to capture an image with a reference exposure value.

However, the exposure controller 109 sets the exposure value of the camera 101 as the first exposure value if the exposure value adjustment cycle has arrived in step S107. That is, the brightness is brighter by increasing the exposure value than the reference exposure value (S109). Then, the image acquisition unit 103 obtains a bright image compared to the optimal exposure image (S111).

The exposure controller 109 determines whether the exposure value adjustment cycle has arrived in the same manner as in step S107 (S113).

At this time, if the exposure value adjustment cycle has not arrived, step S111 is performed again. On the other hand, when the exposure value adjustment cycle has arrived, the exposure controller 109 resets the exposure value of the camera 101 to a reference exposure value, that is, the optimum exposure value (S115). Then, the image acquisition unit 103 obtains the optimal exposure image (S117).

The exposure controller 109 determines whether the exposure value adjustment cycle has arrived in the same manner as in step S107 (S119).

At this time, if the exposure value adjustment cycle has not arrived, step S117 is performed again. On the other hand, if the exposure value adjustment cycle has arrived, the exposure control unit 109 sets the exposure value of the camera 101 as the second exposure value. That is, the brightness is darkened by reducing the exposure value than the reference exposure value (S121). Then, the image acquisition unit 103 obtains a dark image compared to the optimal exposure image (S123).

On the other hand, the exposure control unit 109 determines whether or not the predetermined threshold condition is met (S125), if the threshold condition is met, after changing the optimal exposure value, that is, the reference exposure value (S127), and starts again from step S105.

Here, the threshold condition may be the number of images acquired by the image acquisition unit 103 and the total photographing time of the camera. For example, when the exposure control unit 109 counts 30 images output from the image acquisition unit 103, the exposure control unit 109 may change the reference exposure value (S127). Alternatively, the exposure controller 109 may change the reference exposure value in units of 1 sec (S127).

In this case, step S125 is described as being performed after acquiring a dark image, but may be performed at each step after step S105.

While the present invention has been particularly shown and described with reference to exemplary embodiments thereof, it is to be understood that the invention is not limited to the disclosed exemplary embodiments, It belongs to the scope of right.

Claims (13)

In an image recognition system mounted on a vehicle to take an image for recognizing road information,
A camera photographing a road on which the vehicle travels at one or more predefined locations of the vehicle;
An image acquisition unit obtaining one or more road images through the camera;
A vehicle control unit recognizing the road information through the road image and controlling the operation of the vehicle according to the road information; And
And an exposure controller configured to periodically adjust an exposure value of the camera so that the image acquisition unit acquires a plurality of road images having different brightnesses.
The exposure control unit,
A reference exposure value for acquiring an optimal road image, a first exposure value for acquiring a road image brighter than the optimal road image, and a second exposure value for acquiring a road image darker than the optimal road image are preset and stored, and And a reference exposure value, the first exposure value, and the second exposure value are alternately provided to the camera. The method of claim 1,
The exposure control unit,
Determine whether a threshold condition is connected to the image acquisition unit, wherein the threshold condition includes the number of images acquired by the image acquisition unit and a photographing time of the camera; and when the threshold condition is satisfied, the reference exposure value is changed. Image recognition system. delete The method of claim 1,
The exposure control unit,
An image provided to the camera by periodically repeating the reference exposure value, the first exposure value, the reference exposure value, and the second exposure value or the reference exposure value, the second exposure value, the reference exposure value, and the first exposure value Recognition system. The method of claim 1,
The first exposure value is composed of a plurality of first exposure values in which the exposure value becomes larger step by step from the reference exposure value, and the second exposure value is composed of a plurality of second exposure values in which the exposure value is gradually decreased from the reference exposure value,
The exposure control unit,
And provide the camera with the plurality of first exposure values and the plurality of second exposure values, respectively. The method of claim 1,
The exposure control unit,
And providing the reference exposure value, the first exposure value, and the second exposure value to the camera for a predetermined time. delete An image recognition method of an image recognition system mounted on a vehicle and capturing an image for recognizing road information.
Adjusting the exposure value of the camera periodically; And
And acquiring a road image according to the adjusted exposure value of the camera.
The adjusting step,
A reference exposure value for acquiring an optimal road image, a first exposure value for acquiring a road image brighter than the optimal road image, and a second exposure value for acquiring a road image darker than the optimal road image are periodically alternated to an exposure value of the camera. Set it up,
The obtaining step,
And obtaining different road images according to the reference exposure value, the first exposure value, or the second exposure value. 9. The method of claim 8,
The obtaining step,
And acquiring a plurality of road images corresponding to the exposure value for each exposure value by continuing the reference exposure value, the first exposure value, or the second exposure value for a predetermined time. 9. The method of claim 8,
The first exposure value is composed of a plurality of first exposure values in which the exposure value becomes larger step by step from the reference exposure value, and the second exposure value is composed of a plurality of second exposure values in which the exposure value is gradually decreased from the reference exposure value,
The obtaining step,
And obtaining a respective bright road image according to the plurality of first exposure values and a respective dark road image according to the plurality of second exposure values. 9. The method of claim 8,
The obtaining step,
Obtaining an optimal road image according to the reference exposure value;
Acquiring the bright road image according to the first exposure value;
Obtaining the optimal road image; And
Acquiring the dark road image according to the second exposure value
/ RTI > 9. The method of claim 8,
The obtaining step,
Obtaining an optimal road image according to the reference exposure value;
Acquiring the bright road image according to the second exposure value;
Obtaining the optimal road image; And
Acquiring the dark road image according to the first exposure value
/ RTI > 9. The method of claim 8,
After the obtaining step,
Determining whether a predetermined threshold condition, wherein the threshold condition includes a number of acquired images and a photographing time of a camera;
Adjusting the reference exposure value if the threshold condition is met;
Obtaining an optimal road image by applying the adjusted reference exposure value; And
Acquiring each road image according to the first exposure value or the second exposure value
Image recognition method further comprising.

Images