KR101748646B1 - Apparatus And Method Detecting Side Vehicle Using Camera - Google Patents

Abstract

The present invention relates to an apparatus and a method for detecting a side vehicle using a camera, and a side vehicle detecting apparatus according to the present invention detects a specific ROI from a side object and determines whether the detected specific ROI is a wheel common to all the vehicles And recognizes an object on a side surface. When the side object is detected as a vehicle, the distance to the detected vehicle is detected. When the distance approaches the distance, a warning is given to the driver or a steering wheel is forcibly operated to prevent an accident. According to the present invention, it is possible to confirm whether or not a side object is present, and if it is confirmed that an object exists on the side surface of the vehicle, it can be confirmed whether or not an existing object is a vehicle. In addition, there is an advantage in that cost can be reduced by using a low-cost camera instead of expensive equipment such as a sensor.

Description

TECHNICAL FIELD [0001] The present invention relates to a side vehicle detecting apparatus and method using a camera,

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an apparatus and a method for detecting a side vehicle, and more particularly, to an apparatus and a method for detecting and responding to a side vehicle using a camera.

The conventional side vehicle detection technology can be largely divided into a method using a camera and a method using a sensor.

The camera uses a ROI (Region of Interest) having a certain size and position on the left and right sides. The motion vector is obtained in the corresponding ROI, and the object is recognized through the opposite motion of the background and the object. When the vehicle is in motion, the motion vector of the background appears from the front to the rear, but when the object is on the side, it does not occur or appears from the rear to the front.

The sensor uses a laser, infrared or ultrasound sensor, and measures the time it takes for the laser or light from the sensor to come back after it touches to determine the presence of an object in the area.

However, in the camera-based method, a presence or absence of a side object is determined through a motion vector of a specific region in an image input through a camera. Therefore, if the object exists on the side, the motion vector of the object and the background are the same Therefore, even if an object exists on the side surface, there is a problem that the object can not be detected because the motion vector does not occur.

In addition, since the method of using the sensor differs from the conventional road environment in that the accuracy of object detection is low because there are many similar radio waves other than infrared or ultrasonic waves outputted from the sensor, and when the vehicle is traveling at high speed, Calibration is required. In the case of a laser sensor, the accuracy is reliable, but it is expensive and difficult to popularize.

Finally, a common problem of the prior art is that it is possible to determine whether an object exists or not, but it is impossible to determine whether the object is a vehicle.

SUMMARY OF THE INVENTION The present invention has been made in view of the above problems, and it is an object of the present invention to provide a vehicle side vehicle detection system and a vehicle side vehicle detection method, which detects a specific ROI from a side object, detects whether the detected specific ROI is common to all vehicles, And an object of the present invention is to provide an apparatus and a method.

It is another object of the present invention to provide a side vehicle detecting apparatus using a camera for detecting a distance to a detected vehicle when a side object is detected as a vehicle and for warning a driver when the distance approaches, And a method.

According to an aspect of the present invention, there is provided a side vehicle detection apparatus using a camera, including: a camera that captures and transmits an image; A side vehicle detector for extracting a region of interest (ROI) region for side vehicle detection from the image transmitted from the camera to determine whether a side vehicle exists; An ECU that controls the steering angle or alerts the driver to avoid collision with the side vehicle if the side vehicle detection unit determines that the side vehicle exists; And a warning unit informing the driver of the possibility of collision with the side vehicle or warning of the possibility of collision with the side vehicle.

According to another aspect of the present invention, there is provided a lateral vehicle detection method using a camera, comprising: extracting a region of interest (ROI) region for detecting a lateral vehicle in an image photographed by a camera, Controlling the steering angle to avoid collision with the side vehicle if it is determined that the side vehicle exists; And warning the driver of the possibility of collision with the side vehicle or warning of the possibility of collision with the side vehicle.

According to the present invention, it is possible to confirm whether or not a side object is present, and if it is confirmed that an object exists on the side surface of the vehicle, it is possible to confirm whether or not an existing object is a vehicle.

In addition, there is an advantage in that cost can be reduced by using a low-cost camera instead of expensive equipment such as a sensor.

1 is a block diagram for explaining a side vehicle detection apparatus using a camera according to an embodiment of the present invention;
2 is an exemplary view for explaining a position of a side vehicle;
3 is an exemplary view for explaining a side vehicle in an actual image.
4 is a flowchart for explaining the operation of the side vehicle detection unit.
5 is an exemplary view for explaining a position of an ROI in an image input from a camera;
6 is an exemplary view for explaining a position of a lower end of a wheel;
FIG. 7 is an exemplary view for explaining the degree of visibility of a wheel according to an angle in an image. FIG.
8 is an exemplary view for explaining an angle between a camera and a wheel;
Fig. 9 is an exemplary diagram for explaining how to divide the wheel ROI into quarters. Fig.

BRIEF DESCRIPTION OF THE DRAWINGS The advantages and features of the present invention and the manner of achieving them will become apparent with reference to the embodiments described in detail below with reference to the accompanying drawings. The present invention may, however, be embodied in many different forms and should not be construed as being limited to the embodiments set forth herein. Rather, these embodiments are provided so that this disclosure will be thorough and complete, and will fully convey the scope of the invention to those skilled in the art. And is intended to enable a person skilled in the art to readily understand the scope of the invention, and the invention is defined by the claims. It is to be understood that the terminology used herein is for the purpose of describing particular embodiments only and is not intended to be limiting of the invention. In the present specification, the singular form includes plural forms unless otherwise specified in the specification. It is noted that " comprises, " or "comprising," as used herein, means the presence or absence of one or more other components, steps, operations, and / Do not exclude the addition.

The present invention is a device for recognizing a side vehicle through an image through a camera in a running vehicle, and preventing an accident through a collision warning or a forced vehicle operation according to the distance.

Hereinafter, a side vehicle detection apparatus using a camera according to an embodiment of the present invention will be described with reference to FIGS. FIG. 1 is a block diagram for explaining a side vehicle detection apparatus using a camera according to an embodiment of the present invention, FIG. 2 is an exemplary view for explaining a position of a side vehicle, and FIG. Fig.

1, a side vehicle detection apparatus 100 using a camera according to the present invention includes a camera 110, a side vehicle detection unit 120, an ECU 130, a steering angle control unit 140, a brake control unit 150, And a warning unit 160. [

The camera 110 captures an image in real time and transmits it to the side vehicle detection unit 120. Here, the camera 110 is a calibrated camera. The reason why the camera 110 needs to be calibrated is to calculate the distance to the recognized vehicle and the size of the wheel according to the distance from the image.

The side vehicle detection unit 120 determines whether a side vehicle exists in the image transmitted from the camera 110. [ 2, the vehicle located in area 2 is referred to as a side vehicle, and the vehicle located in area 1 is referred to as a forward vehicle. If this is applied to the actual image, as shown in FIG. 3, the vehicles No. 5 and No. 6 are referred to as side vehicles, and the vehicles No. 1 to No. 4 are referred to as the forward vehicle.

The side vehicle detection unit 120 transmits a danger signal to the ECU 130 when the vehicle is detected on the side surface and the distance to the detected vehicle is within a predetermined danger distance.

When the ECU 130 receives the danger signal from the side vehicle detection unit 120, the ECU 130 controls the steering angle control unit 140 to adjust the steering angle in a direction opposite to that in which the side vehicle exists or to cause the brake control unit 150 to operate the brake It is possible to avoid a collision with the side vehicle in an appropriate manner according to the driving situation of the vehicle. At the same time, the ECU 130 sends a signal to the warning unit 160 to cause the warning unit 160 to warn the driver of the possibility of collision with the side vehicle.

As described above, according to the present invention, it is possible to confirm whether or not a side object is present, and when it is confirmed that an object exists on the side of the vehicle, it is possible to check whether an existing object is a vehicle. You can take steps to avoid and warn the driver of the possibility of a crash. In addition, there is an advantage in that cost can be reduced by using a low-cost camera instead of expensive equipment such as a sensor.

1 to 3, a side vehicle detection apparatus using a camera according to an embodiment of the present invention has been described. Hereinafter, referring to FIGS. 4 to 9, a camera according to an embodiment of the present invention The side vehicle detection method will be described. FIG. 4 is a flow chart for explaining the operation of the side vehicle detection unit, FIG. 5 is an exemplary view for explaining the position of the ROI in the image input from the camera, and FIG. 6 is an exemplary view for explaining the position of the bottom of the wheel. FIG. 8 is a view for explaining the angle between the camera and the wheel, and FIG. 9 is a view for explaining how the wheel ROI is divided into quadrants in the wheel ROI. Fig.

As shown in FIG. 4, when an image is input from the camera 110 (S400), an ROI (Region Of Interest) region is extracted as shown in FIG. 5 (S410). For example, when an image is transferred from the camera 110, an ROI area of size M x N is stored in R (x, y).

The extracted ROI region is preprocessed (S420). For example, in the ROI R (x, y) extracted, the image is flattened and reduced to half the image size as shown in Equation (1). That is, Equation (1) simultaneously performs planarization and reduction of an image. This is to remove the noise of the image and speed up the calculation.

In Equation 1, x = 0, 2, 4, ..., M-1, and y = 0, 2, 4, ..., N-1.

The vertical edge map VE (x, y) is obtained by using only the vertical filter among the sobel operators in the obtained P (x, y), as shown in Equation 2 and Equation (3).

In Equation (4), max represents the largest brightness value among 8-neighborhood pixels at pixel (x, y), and min represents the lowest brightness value. From VE (x, y) thus obtained, profile is obtained as shown in Equation (5) in the vertical direction.

In Equation (5), i = 0, 1, 2, ..., (M / 2) -1.

The wheel ROI is extracted (S430). For example, it is assumed that the position with the highest frequency in the profile obtained as described above is Xp and the position in the x-direction of the center of the wheel. Use the value of the profile (Xp) to find the y position of the lower part of the wheel. For example, the positions of Xp and Yp to be obtained from the ROI are as shown in FIG. To obtain Yp, the value of Xp position in VE (x, y) is stored in YE (i) as shown in Equation (6), and YE '(i) is obtained through KDE as shown in Equation (7).

In equations (6) and (7), i = 0, 1, 2, ... , (N / 2) -1.

When the obtained YE '(i) satisfies the condition of Equation (8), the value of i is used as Yp.

In Equation (8), i = (N / 2) -1, (N / 2) -2, ... , 1.

The size of the wheel is calculated through coordinates (Xp, Yp) and camera calibration information obtained through the above-described process. The following assumptions are made to calculate the size of the wheel.

As shown in FIG. 7, when the value of? Is 0, the shape of the wheel coming into the image is a circle, and when the value of? Is 90 degrees, the wheel is not seen at all. For this reason, it is assumed that the diameter of the wheel is 0 when the angle of the wheel is 0 degrees, and the diameter of the wheel is changed to 90 degrees according to the cosine function.

The above-described assumption is as shown in Fig. The angle of the wheel with respect to the coordinates (RXp, RYp) in the real world of the coordinates (Xp, Yp) of the camera 110 and the image is obtained as shown in Equation (9).

Therefore, the left and right coordinates in the real world of the wheel region are (RXp - Rx × COS (θ / 2), RYp) and (RXp + Rx × COS (RXp, RYp - Ry) in the real world. The wheel ROI can be extracted by converting the coordinates of the real world into image coordinates according to the calibration information.

The feature is detected (S440). For example, the feature is detected in the extracted wheel ROI and the feature used is Histogram Of Gradient (HOG). In order to extract the HOG, as shown in FIG. 9, the wheel ROI is divided into quadrants, and 18 dimensions of HOG are extracted from the respective regions, and then connected in order to extract a total of 72 dimensions.

The extracted features are determined (460) by the widely known classifier 450 such as Adaboost, SVM or Neural Network.

As a result of the determination, if there is a wheel, it is determined that the side vehicle is located in the corresponding ROI, and the distance between the driver vehicle and the side vehicle is determined by camera calibration information (470).

 It is determined whether the distance between the driver's vehicle and the side vehicle is within a predetermined danger distance (S480).

As a result of the determination, if it is within the danger distance, an alarm signal is transmitted (S490).

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 embodiments, but, on the contrary, is intended to cover various modifications and equivalent arrangements included within the spirit and scope of the appended claims. Therefore, the scope of the present invention should not be limited by the illustrated embodiments, but should be determined by the scope of the appended claims and equivalents thereof.

110: camera 120: side vehicle detection unit
130: ECU 140: steering angle control section
150: Brake control unit 160: Warning unit

Claims (6)

A camera for capturing and delivering images;
A side vehicle detector for extracting a region of interest (ROI) region for side vehicle detection from the image transmitted from the camera to determine whether a side vehicle exists;
An ECU for controlling the steering angle or warning the driver to avoid collision with the side vehicle if the side vehicle detection unit determines that the side vehicle exists; And
And a warning unit informing the driver of the possibility of collision with the side vehicle or warning of the possibility of collision with the side vehicle,
The side vehicle detection unit extracts an ROI (Region Of Interest) from the delivered image, extracts the x-axis and y-axis coordinates of the wheel by reducing the extracted ROI to a specific size, Determining whether the wheel exists or not, determining that the side vehicle exists if the wheel exists and the distance from the wheel is within a predetermined danger distance
Side vehicle detecting apparatus using a camera. delete The method according to claim 1,
The ECU may adjust the steering angle in the opposite direction in which the side vehicle exists or operate the brake to avoid collision with the side vehicle
Side vehicle detecting apparatus using a camera. Extracting a region of interest (ROI) region for side vehicle detection from the image photographed by the camera, and determining whether a side vehicle exists;
Controlling the steering angle to avoid collision with the side vehicle if it is determined that the side vehicle exists; And
Notifying a driver of a collision with a side vehicle or warning of a possibility of collision with the side vehicle,
Wherein the determining step comprises:
Extracting a region of interest (ROI) from the delivered image and extracting the x-axis and y-axis coordinates of the wheel by reducing the extracted ROI to a specific size;
Determining the presence or absence of a wheel by calculating the size of the wheel based on the extracted coordinates; And
And judging that the side vehicle exists if the wheel exists and the distance to the existing wheel is within a predetermined danger distance
A side vehicle detection method using a camera. delete 5. The method of claim 4,
And performing at least one of adjusting the steering angle in the opposite direction in which the side vehicle exists, or operating the brake to prevent traveling in the direction in which the side vehicle exists
A side vehicle detection method using a camera.

Images