KR20170058512A - Apparatus for traffic signs recognition and method thereof - Google Patents

Abstract

The present invention relates to a traffic sign recognition technology. A traffic sign recognition device according to the present invention comprises: an opposing vehicle detecting unit for capturing an image using a camera which detects the front of a user vehicle, and detecting an opposing vehicle on the acquired image; a driving lane determining unit for setting a central reference which is the vertical direction of the user vehicle, and determining the driving lane of the user vehicle based on the relationship between the central reference and the opposing vehicle; a region of interest (ROI) setting unit for setting three ROI having a certain range with respect to the image, and setting an analysis sequence of the ROI based on the driving lane; and a traffic sign recognition unit for recognizing a traffic sign in three ROI according to the analysis sequence, and ending a traffic sign formula if the traffic sign is recognized in one or more ROI among the three ROI according to the sequence. As described above, according to the present invention, it is possible to use less data and a processor having lower performance by recognizing the traffic sign only in a portion where the traffic sign exists in an image taken by a camera.

Description

[0001] APPARATUS FOR TRAFFIC SIGNS RECOGNITION AND METHOD THEREOF [0002]

The present invention relates to traffic sign recognition technology.

The vehicle is equipped with a camera for monitoring the front of the vehicle. The camera can recognize an object including a vehicle, a traffic sign, and the like existing in the image by performing an image processing technique on the image captured by the camera.

However, in order to recognize an object existing in an image by performing an image processing technique or the like, a lot of data is required and a high-performance processor is required.

Such a problem can be solved by recognizing only the necessary part without recognizing the premise part of the image.

In detail, when the traffic sign is recognized, it is possible to use less data by recognizing the traffic sign by applying the image processing technique or the like only to the portion where the traffic sign is present in the image taken by the camera, .

However, at present, there is a problem that a device which recognizes a traffic sign only for a part where a traffic sign is present is set in the image taken by the camera, is not developed.

SUMMARY OF THE INVENTION In view of the foregoing, an object of the present invention is to provide a technique for setting a portion where a traffic sign is present in an image photographed by a camera and recognizing a traffic sign only for the relevant portion.

According to an aspect of the present invention, there is provided an image processing apparatus for acquiring an image using a camera for sensing the front of a vehicle, an opposite vehicle detecting unit for detecting an opposite vehicle in the captured image, A driving lane determining unit for determining a driving lane of the subject vehicle based on the relationship between the center reference and the opposing vehicle, and three regions of interest (ROI) having a certain range with respect to the image, If the traffic sign is recognized in one or more areas of interest among the three areas of interest according to the analysis sequence, the traffic sign is recognized in three areas of interest according to the analysis area, And a traffic sign recognition unit for terminating the recognition of the sign.

According to another aspect of the present invention, there is provided a driving method of a vehicle, comprising the steps of: acquiring an image using a camera that senses the front of the vehicle; detecting an opposing vehicle from the acquired image to determine a driving lane of the vehicle; Setting a region of interest (ROI) having a certain range and setting an analysis order of the ROI on the basis of the driving lane, and identifying the traffic sign in the ROI according to the analysis sequence, And a traffic sign recognition step of terminating the traffic sign recognition if a traffic sign is recognized in at least one region of interest among the three regions of interest according to the analysis order.

As described above, according to the present invention, it is possible to use less data by recognizing a traffic sign only for a portion where a traffic sign exists in an image taken by a camera, and to use a lower performance processor .

1 is a view illustrating a traffic sign recognition apparatus according to an embodiment of the present invention.
FIG. 2 is a diagram illustrating an operation of the anti-vehicle detection unit according to an embodiment of the present invention. Referring to FIG.
3 is a diagram illustrating an example of an operation of the driving lane determination unit according to an embodiment of the present invention.
4 is a view illustrating another example of the operation of the driving lane determining unit according to an embodiment of the present invention.
FIG. 5 is a diagram illustrating an example of an operation of the ROI setting unit according to an embodiment of the present invention. Referring to FIG.
6 is a view illustrating another example of operations of the ROI recognition unit and the ROI recognition unit according to an embodiment of the present invention.
7 is a flowchart illustrating a method of recognizing a traffic sign according to an exemplary embodiment of the present invention.

Hereinafter, some embodiments of the present invention will be described in detail with reference to exemplary drawings. It should be noted that, in adding reference numerals to the constituent elements of the drawings, the same constituent elements are denoted by the same reference numerals even though they are shown in different drawings. In the following description of the present invention, a detailed description of known functions and configurations incorporated herein will be omitted when it may make the subject matter of the present invention rather unclear.

In describing the components of the present invention, terms such as first, second, A, B, (a), and (b) may be used. These terms are intended to distinguish the constituent elements from other constituent elements, and the terms do not limit the nature, order or order of the constituent elements. When a component is described as being "connected", "coupled", or "connected" to another component, the component may be directly connected to or connected to the other component, It should be understood that an element may be "connected," "coupled," or "connected."

1 is a view illustrating a traffic sign recognition apparatus according to an embodiment of the present invention.

A traffic sign recognition apparatus according to an embodiment of the present invention includes an opposite vehicle detection unit that acquires an image using a camera that detects the front of the vehicle and detects an opposite vehicle from the acquired image, A driving lane determining unit for determining a driving lane of the vehicle based on the relationship between the center reference and the counter vehicle, and three regions of interest (ROI) having a certain range for the image, And a traffic sign is recognized in one or more areas of interest among the three areas of interest according to the order of traffic, And a traffic sign recognition unit for terminating the identification of the sign.

Referring to FIG. 1, a traffic sign recognition apparatus 100 according to an embodiment of the present invention includes an image capturing unit 120 for capturing an image using a camera for sensing the front of the vehicle, (110).

Generally, a vehicle is equipped with a front camera for storing images at the time of a traffic accident. The images stored by these forward cameras play an important role in judging whether or not the accident is responsible.

As described above, the front camera can be used for a traffic accident, but the present invention is not limited to this, but it is a tendency to use the camera for other purposes.

The counter vehicle detection unit 110 can detect the other vehicle using the image processing technique programmed in the forward image of the subject vehicle obtained through the forward camera.

A brief description of image processing techniques is a technique for more accurately detecting features including shapes or colors, and can perform one or more of the following steps.

The image processing technique may first perform an image acquisition step. This simply means obtaining a digital image, and includes a preprocessing such as scaling.

Once image acquisition is complete, an image enhancement step may be performed. The image improvement step is a process of manipulating the image more suitably for a specific application, and can be improved by applying various methods based on human subjective preference.

When the image enhancement is completed, an image restoration step can be executed. The image reconstruction step also improves the image, but unlike the image enhancement step, the image can be improved based on mathematical or probabilistic models.

When the image restoration is completed, a color image processing step can be executed. It applies color models and basic color processing in the digital domain.

When the color image processing is completed, a wavelet step can be executed. This is for displaying images at various resolutions.

Once the wavelet is complete, a compression step can be performed. By executing the compression step, it is possible to reduce the bandwidth required to transmit the storage device or image necessary for storing the image.

Once the compression step is complete, a morphological processing step may be performed. This may be to extract image components that are useful for shape representation and description.

Once the ecological processing step is complete, a segmentation step may be performed that divides the image into its constituent parts or constituent objects.

Once the segmentation step is complete, a Representation & description step may be performed that transforms the incoming data into a computer process suitable for processing.

Once the presentation and description steps are complete, you can perform an object recognition step that labels the object based on the descriptors.

By performing the image processing technique including at least one of the above-described steps, the counter vehicle detecting unit 110 can accurately detect the other vehicle in the image acquired from the camera.

In addition, the counter vehicle detecting unit 110 can calculate the speed of the subject vehicle from the forward image and the relative speed of the detected other vehicles.

For example, if a forward camera operating at 30 [frame / s] is installed in the vehicle, and the distance change value of the fixed object measured between the frames by the forward camera is 3 [m / frame] The unit 110 can calculate the speed of the subject vehicle at 90 [m / s].

The counter vehicle detecting unit 110 can calculate the relative speed of the other vehicle with respect to the subject vehicle using the distance change value of the other vehicle instead of the fixed object.

For example, when the distance change value of another vehicle measured between frames by the front camera is 0.6 [m / frame], the counter vehicle detecting unit 110 sets the relative speed of the other vehicle with respect to the subject vehicle at 18 [m / s ] Can be calculated.

The counter vehicle detecting unit 110 may calculate the absolute speed of the other vehicle by adding the speed of the subject vehicle to the relative speed of the other vehicle with respect to the subject vehicle, and the counter vehicle detecting unit 110 may calculate the absolute speed of the other vehicle It is possible to detect an opposing vehicle by comparing it with a preset speed threshold value.

For example, if the speed threshold value is set to 0 [m / s], the counter vehicle detecting section 110 detects another vehicle having an absolute velocity smaller than 0 [m / s] can do.

The traffic sign recognition apparatus 100 according to an embodiment of the present invention divides the image into center reference in the longitudinal direction of the vehicle and determines the driving lane of the vehicle based on the relationship between the center reference and the counter vehicle (120). ≪ / RTI >

The driving lane decision unit 120 can set the center reference in the vertical direction in the image based on the installed position of the camera.

The center reference can be set as a line as a reference for separating the opposite road from the running road of the subject vehicle. However, the center reference set as the line may have an error in separating the opposite road from the running road of the vehicle according to the traveling road of the vehicle. To solve this error, the center reference may be set to a certain area.

However, in the following description, the case where the center reference is set as a line will be mainly described.

The driving lane determining unit 120 determines that the number of opposing vehicles detected on the right side of the center reference is greater than the number of opposing vehicles detected on the left side of the center reference and the number of opposing vehicles detected on the right side of the center reference is greater than the number threshold The number of opposing vehicles detected on the left side of the center reference is greater than the number of opposing vehicles detected on the right side of the center reference and the number of opposing vehicles detected on the left side of the center reference is greater than the number threshold It is possible to judge the driving lane to the right.

The number threshold value is a value for blocking the driving lane judging error of the driving lane judging unit 120 and can be set through experimental data.

The traffic sign recognition apparatus 100 according to an embodiment of the present invention sets three ROIs having a certain range with respect to an image and sets an analysis order of ROIs based on the driving lane And an interest area setting unit 130. [

The interest area setting unit 130 sets the upper-side interest area, the right-side interest area, and the left-side interest area that do not overlap with each other, and when the driving lane is determined as the right side, When the driving lane is determined as the left side, the analysis order can be set to the upper interest region, the left interest region, and the right interest region.

The upper area of interest, the right area of interest and the area of interest of the left may be areas of interest for detecting traffic signs located on the upper middle of the road, traffic signs located on the right side of the road, and traffic signs located on the left side of the road, The right interest area and the left interest area may be set so as not to overlap the recognition time, and further, the frames of the upper interest area, the right interest area, and the left interest area may be set to not match each other.

In addition, the ROI setting unit 130 may further include a function of resetting the predetermined range if the traffic sign recognition unit 140 does not recognize the traffic sign.

However, since the traffic sign recognition unit 140 may not recognize the absence of the traffic sign, the region-of-interest setting unit 130 may reset a certain range by comparing it with the threshold value for the non-recognition.

The threshold value for the non-recognition can be calculated through the experimental data, and can be the value of the number of times of non-recognition.

The traffic sign recognition apparatus 100 according to an exemplary embodiment of the present invention identifies traffic signs in three areas of interest according to a set analysis procedure and identifies traffic signs in one or more areas of interest And a traffic sign recognizing unit 140 for terminating the traffic sign recognition if it is recognized.

The traffic sign recognizing unit 140 may identify the traffic sign using an image processing technique similar to the countermeasure detecting unit 110. [ However, the traffic sign recognizing unit 140 recognizes the traffic sign based on the characteristics including the shape and color of the traffic sign, So that traffic signs can be detected.

FIG. 2 is a diagram illustrating an operation of the anti-vehicle detection unit according to an embodiment of the present invention. Referring to FIG.

The opposing vehicle detecting unit according to an embodiment of the present invention calculates the absolute speed of other vehicles based on the speed of the subject vehicle and the relative speed of other vehicles acquired through the image, To detect an opposing vehicle. According to the setting, the absolute speed of the opponent vehicle may be smaller than the speed threshold value.

Referring to FIG. 2, the counter vehicle detecting unit may obtain the speed of the subject vehicle and the relative speed of other vehicles in the image obtained from the camera sensing the front of the subject vehicle (S200).

For example, if a forward camera operating at 30 [frame / s] is installed in the vehicle, and the distance change value of the fixed object measured between the frames by the forward camera is 3 [m / frame] The velocity of the vehicle can be calculated as 90 [m / s].

If the distance change values of the first vehicle and the second vehicle measured between the frames by the front camera are 0.6 [m / frame] and -3.1 [m / frame], respectively, The relative speeds of the first and second vehicles can be calculated as 18 [m / s] and -93 [m / s], respectively.

If the speed of the subject vehicle and the relative speed of the other vehicles are obtained through step S200, the counter vehicle detecting unit may calculate the absolute speed of the other vehicles by adding the speed of the subject vehicle to the relative speed of the other vehicle with respect to the subject vehicle S210).

For example, in step S200, the vehicle speed, the absolute speed of the first vehicle, and the absolute speed of the second vehicle are 90 [m / s], 18 [m / s] , The counter vehicle detecting unit can calculate the absolute speed of the first vehicle and the absolute speed of the second vehicle at 108 [m / s] and -3 [m / s].

When the absolute speed of the other vehicles is calculated by performing step S210, the opposite vehicle detection unit may detect the opposite vehicle by comparing the absolute speed of the calculated other vehicles with a preset speed threshold (S220).

For example, if the absolute velocity of the first vehicle and the absolute velocity of the second vehicle are calculated as 108 [m / s] and -3 [m / s] in step S210, respectively, Is compared with a preset speed threshold value of 0 [m / s], and a second vehicle having a value smaller than a speed threshold value of 0 [m / s] can be detected as an opposing vehicle.

In the above description, the case where the speed threshold value is set to 0 [m / s] has been described. However, the present invention is not limited to this and can be set to other values based on data through experiments.

3 is a diagram illustrating an example of an operation of the driving lane determination unit according to an embodiment of the present invention.

The driving lane determining unit according to an embodiment of the present invention divides the image obtained through the camera into center reference in the vertical direction of the vehicle and detects the number of opposing vehicles detected on the right side of the center reference as the counterpart detected on the left side of the center reference If the number of opposing vehicles detected on the right side of the center reference is greater than the number threshold and the number of opposing vehicles detected on the left side of the center reference is greater than the number of opposing vehicles detected on the right side of the center reference If the number of opposing vehicles detected on the left side of the center reference is greater than the number of vehicles, the driving lane can be judged as the right side if it is larger than the number threshold.

Referring to FIG. 3, an opposing vehicle 330 can be detected by an opposing vehicle detecting unit described using FIG. 2 in an image including other vehicles 310, 320, and 330 obtained through a camera.

In the above-described situation, the driving lane determining unit divides the image by the center reference 340, which is the longitudinal direction of the vehicle, and determines that the number 1 of the opposing vehicles 330 detected on the left side of the center reference 340 is the number threshold value 0 It is possible to judge the driving lane of the host vehicle to the right side.

Although the number threshold has been described as being set to zero in the above description, it may be set to an integer greater than 0 in order to increase the accuracy of the driving lane determination unit.

For example, if the number threshold is set to 0, the vehicle is running on the right road, and the other vehicle is running backward on the right outer lane, the driving lane judge may erroneously judge the lane to the left. The determination as such an erroneous driving lane can be resolved by setting the number threshold to an integer greater than zero.

Hereinafter, the operation of the driving lane determining unit described with reference to Fig. 3 will be described in detail with reference to Fig.

4 is a view illustrating another example of the operation of the driving lane determining unit according to an embodiment of the present invention.

Referring to FIG. 4, the driving lane determining unit may set a center reference in the vertical direction of the vehicle (S400).

The center criterion can be set based on the position and monitoring angle of the installed camera.

For example, in an image photographed from a camera installed at the right side, the center reference may be set to the left side of the center reference in the image photographed from the camera installed at the left side position.

In addition, the center reference in the image taken from the camera installed with the viewing angle directed to the right side can be set to the left side of the center reference in the image taken from the camera installed with the monitoring angle directed to the left side.

Furthermore, in order not to be affected by the driving lane of the subject vehicle, the center reference may be set to an area other than a line.

More specifically, if the center reference is set as a line based on the case where the vehicle travels in the first lane on the right road, if the same vehicle runs in a three-lane line, the center reference is between the two lanes and the third lane There may be a problem in judging the driving lane.

The above-mentioned problem can be solved by setting the center reference to the area of the primary and secondary lanes rather than the line as the center reference.

If the center reference is set by performing step S400, the driving lane determining unit determines whether the number of opposing vehicles detected on the right side of the center reference is greater than the number of opposing vehicles detected on the left side of the center reference (S410).

If it is determined in step S410 that the number of opposing vehicles detected on the right side of the center reference is larger than the number of opposing vehicles detected on the left side of the center reference (YES), the driving lane determining unit determines that the number of opposing vehicles detected on the right side of the center reference Is greater than a predetermined number threshold (S420).

In step S420, the number threshold may be set to an integer greater than 0 through experiment data in order to prevent a judgment error of the driving lane determining unit in a situation where a backing vehicle is detected as an opposing vehicle.

If it is determined in step S420 that the number of detected opposing vehicles detected on the right side of the center reference is larger than the number threshold value (YES), the driving lane determining unit may determine the driving lane of the vehicle as the left side (S430).

If it is determined in step S410 that the number of opposing vehicles detected on the right side of the center reference is not larger than the number of opposing vehicles detected on the left side of the center reference (NO), the driving lane determining unit determines that the counter vehicle Is greater than a predetermined number threshold (S440).

Similar to step S420, in step S440, in order to prevent a judgment error of the driving lane determining unit in a situation where a reversing vehicle is detected as an opposing vehicle, the coefficient threshold may be the same as the coefficient threshold used in step S420.

If it is determined in step S440 that the number of opposing vehicles detected on the left side of the center reference is larger than the number threshold value (YES), the driving lane determining unit may determine the driving lane of the vehicle as the right side (S450).

FIG. 5 is a diagram illustrating an example of an operation of the ROI setting unit according to an embodiment of the present invention. Referring to FIG.

The ROI setting unit according to an embodiment of the present invention sets three ROIs, an upper ROI, a right ROI, and a ROI, which have a certain range with respect to an image acquired by the camera, If the driving lane is judged to be on the right side, the analysis order is set as the upper interest area, the right interest area, and the left interest area. If the driving lane is judged as the left side, the analysis order is the upper interest area, the left interest area, Can be set.

5, the ROI setting unit may set an ROI 510, a ROI 520, and a ROI 530, which are three ROIs having a certain range .

The upper area of interest 510 can be set based on the position of the traffic sign installed in the center upper side irrespective of the right driving lane and the left driving lane.

The right interest area 520 can be set based on the position of the traffic sign 540 installed on the right road side in the right driving lane.

The left interest area 530 may be set based on the position of the traffic sign installed on the left side of the left driving lane.

However, the upper interest area 510, the right interest area 520, and the left interest area 530 are set so as not to overlap with each other in order to reduce the time of traffic sign recognition, and further, the respective areas can be set not to touch each other.

5, it can be determined that the driving lane of the subject vehicle is on the right side by the operation of the opposing vehicle detecting unit and the driving lane determining unit, and the region of interest setting unit 510 sets the upper region of interest 510, the right region of interest 520, The analysis order for recognizing traffic signs for the ROI 530 can be set as the ROI 510, the ROI 520, and the ROI 530.

The attention area setting unit may further include a function of resetting a certain range. If the traffic sign recognition unit does not recognize the traffic sign in all of the upper interest area 510, the right interest area 520 and the left interest area 530 The upper region 510, the right region 520 and the left region 530, respectively.

According to the present invention, as the certain range is narrowly set, the time for recognizing a traffic sign can be reduced, but the traffic sign may not be recognized because it is set too narrow. This problem can be solved by resetting a certain range when the traffic sign is not recognized in both the set upper interest area 510, the right interest area 520 and the left interest area 530. [

6, the operation performed by the RO area setting unit described with reference to FIG. 5 and the operation of the traffic sign recognizing unit will be described.

6 is a view illustrating another example of operations of the ROI recognition unit and the ROI recognition unit according to an embodiment of the present invention.

Referring to FIG. 6, the interest area setting unit determines whether the driving lane is determined to be right in the driving lane determining unit (S600).

If it is determined in step S600 that the driving lane has been determined to the right (YES), the ROI setting unit sets the ROI in the order of the upper ROI, the right ROI, and the left ROI, An analysis procedure is set (S610).

If the analysis order is set in step S610, the traffic sign recognition unit confirms the traffic sign in the upper interest area according to the set analysis order (S620)

Briefly, a traffic sign recognition unit can recognize a traffic sign based on the shape or color of the traffic sign on the acquired image.

Thereafter, the traffic sign recognition unit determines whether the traffic sign is recognized in step S620 (S625).

If it is determined in step S625 that the traffic sign is not recognized (NO), the traffic sign recognizing unit recognizes the traffic sign on the right interest area, which is the next analysis sequence of the upper interest area (S630).

This can identify the traffic sign similar to the step S620.

Thereafter, the traffic sign recognition unit determines whether the traffic sign is recognized in step S630 (S635).

If it is determined in step S635 that the traffic sign is not recognized (NO), the traffic sign recognition unit identifies the traffic sign in the left interest area, which is the next analysis sequence of the right interest area (S640).

Alternatively, if it is determined in step S600 that the driving lane is not on the right side (NO), it is determined that the ROI is a ROI (Top Of Interest Area), a ROI The analysis order is set in order (S650).

If the analysis sequence is set in step S650, the traffic sign recognition unit confirms the traffic sign in the upper interest area according to the set analysis order (S660).

Thereafter, the traffic sign recognition unit determines in step S660 whether the traffic sign is recognized (S665).

If it is determined in step S665 that the traffic sign is not recognized (NO), the traffic sign recognizing unit recognizes the traffic sign in the left interest area, which is the next analysis sequence of the upper interest area (S670).

Thereafter, the traffic sign recognition unit determines whether the traffic sign is recognized in step S670 (S675).

If it is determined in step S675 that the traffic sign is not recognized (NO), the traffic sign recognition unit identifies the traffic sign on the right interest area, which is the next analysis sequence of the left interest area (S680).

If it is determined at step S625, S635, S665, and S675 that the traffic sign has been recognized (YES), the traffic sign recognition unit recognizes the traffic sign of the area of interest in the next analysis procedure It can be terminated without being identified.

As described above, according to the traffic sign recognition apparatus of the present invention, a traffic sign can be recognized within a shorter time by recognizing a traffic sign only for a region of interest having a certain range rather than the entire image acquired by the camera.

Hereinafter, a traffic sign recognition method, which is an operation performed by the traffic sign recognition apparatus described with reference to FIGS. 1 to 6, will be briefly described.

7 is a flowchart illustrating a method of recognizing a traffic sign according to an exemplary embodiment of the present invention.

A traffic sign recognition method according to an embodiment of the present invention includes the steps of: acquiring an image using a camera that detects the front of a vehicle; detecting a vehicle in the acquired image to determine a driving lane of the vehicle; (ROI) having a certain range with respect to the image and setting an analysis sequence of the ROI on the basis of the driving lane, A traffic sign recognition step of recognizing the sign and ending the traffic sign recognition if the traffic sign is recognized in one or more areas of interest among the three areas of interest in the sequence.

Referring to FIG. 7, a traffic sign recognition method according to an embodiment of the present invention includes acquiring an image using a camera for sensing the front of the vehicle, and detecting an opposing vehicle from the acquired image (S700 ).

Generally, a vehicle is equipped with a front camera for storing images at the time of a traffic accident. The images stored by these forward cameras play an important role in judging whether or not the accident is responsible.

As described above, the front camera can be used for a traffic accident, but the present invention is not limited to this, but it is a tendency to use the camera for other purposes.

The counter vehicle detection step S700 may detect the other vehicle using the image processing technique programmed in the forward image of the subject vehicle obtained through the forward camera.

A brief description of image processing techniques is a technique for more accurately detecting features including shapes or colors, and can perform one or more of the following steps.

The image processing technique may first perform an image acquisition step. This simply means obtaining a digital image, and includes a preprocessing such as scaling.

Once image acquisition is complete, an image enhancement step may be performed. The image improvement step is a process of manipulating the image more suitably for a specific application, and can be improved by applying various methods based on human subjective preference.

When the image enhancement is completed, an image restoration step can be executed. The image reconstruction step also improves the image, but unlike the image enhancement step, the image can be improved based on mathematical or probabilistic models.

When the image restoration is completed, a color image processing step can be executed. It applies color models and basic color processing in the digital domain.

When the color image processing is completed, a wavelet step can be executed. This is for displaying images at various resolutions.

Once the wavelet is complete, a compression step can be performed. By executing the compression step, it is possible to reduce the bandwidth required to transmit the storage device or image necessary for storing the image.

Once the compression step is complete, a morphological processing step may be performed. This may be to extract image components that are useful for shape representation and description.

Once the ecological processing step is complete, a segmentation step may be performed that divides the image into its constituent parts or constituent objects.

Once the segmentation step is complete, a Representation & description step may be performed that transforms the incoming data into a computer process suitable for processing.

Once the presentation and description steps are complete, you can perform an object recognition step that labels the object based on the descriptors.

By performing the image processing technique including at least one of the steps described above, the counter vehicle detection step (S700) can accurately detect the other vehicle in the image acquired from the camera.

Also, the counter vehicle detection step S700 may calculate the speed of the subject vehicle from the forward image and the relative speed of the detected other vehicles.

For example, if a forward camera operating at 30 [frame / s] is installed in the vehicle, and the distance change value of the fixed object measured between the frames by the forward camera is 3 [m / frame] In step S700, the speed of the subject vehicle can be calculated as 90 [m / s].

The counter vehicle detection step S700 may calculate the relative speed of the other vehicle with respect to the subject vehicle using the distance change value of the other vehicle instead of the fixed object.

For example, if the distance change value of another vehicle measured between the frames by the front camera is 0.6 [m / frame], the counter vehicle detection step S700 sets the relative speed of the other vehicle with respect to the vehicle to 18 [m / s].

The counter vehicle detection step S700 may calculate the absolute speed of the other vehicle by adding the speed of the subject vehicle to the relative speed of the other vehicle with respect to the subject vehicle. In the counter vehicle detection step S700, It is possible to detect an opposing vehicle by comparing the speed with a preset speed threshold.

For example, if the speed threshold value is set to 0 [m / s], the counter vehicle detection step S700 may set the other vehicle having an absolute velocity smaller than 0 [m / s] Can be detected.

A traffic sign recognition method according to an embodiment of the present invention includes dividing an image into a center reference in the longitudinal direction of the vehicle and determining a driving lane of the vehicle based on the relationship between the center reference and the counter vehicle, ).

In the driving lane determination step S710, the center reference in the vertical direction can be set in the image based on the installed position of the camera.

The center reference can be set as a line as a reference for separating the opposite road from the running road of the subject vehicle. However, the center reference set as the line may have an error in separating the opposite road from the running road of the vehicle according to the traveling road of the vehicle. To solve this error, the center reference may be set to a certain area.

However, in the following description, the case where the center reference is set as a line will be mainly described.

In the step S710, the number of opposing vehicles detected on the right side of the center reference is greater than the number of opposing vehicles detected on the left side of the center reference, and the number of opposing vehicles detected on the right side of the center reference is greater than the number threshold The number of opposing vehicles detected on the left side of the center reference is greater than the number of opposing vehicles detected on the right side of the center reference and the number of opposing vehicles detected on the left side of the center reference is greater than the number threshold It is possible to judge the driving lane to the right.

The number threshold value is a value for blocking the driving lane decision error of the driving lane determining step S710, and may be set through the experimental data.

A traffic sign recognition method according to an exemplary embodiment of the present invention includes setting three regions of interest (ROI) having a certain range for an image and setting an ROI for a ROI And a setting step S720.

The interest area setting step S720 sets the upper interest area, the right interest area, and the left interest area that do not overlap with each other, and when the driving lane is determined to be right, the analysis order is the order of the upper interest area, When the driving lane is determined as the left side, the analysis order can be set to the upper interest region, the left interest region, and the right interest region.

The upper area of interest, the right area of interest and the area of interest of the left may be areas of interest for detecting traffic signs located on the upper middle of the road, traffic signs located on the right side of the road, and traffic signs located on the left side of the road, The right interest area and the left interest area may be set so as not to overlap the recognition time, and further, the frames of the upper interest area, the right interest area, and the left interest area may be set to not match each other.

In addition, the interest area setting step S720 may further include a function of resetting a predetermined range to be widened if the traffic sign is not recognized in the traffic sign recognition step S730.

However, since it may not be recognized in the traffic sign recognition step 140 due to absence of the traffic sign, the interest area setting step S720 may reset a certain range by comparing with the threshold value for the insensitivity.

The threshold value for the non-recognition can be calculated through the experimental data, and can be the value of the number of times of non-recognition.

A traffic sign recognition method according to an exemplary embodiment of the present invention recognizes traffic signs in three areas of interest according to a set analysis procedure, and when traffic signs are recognized in one or more areas of interest among three areas of interest, And a traffic sign recognition step S730 for terminating the identification of the sign.

The traffic sign recognition step S730 may identify the traffic sign using an image processing technique similar to the counter vehicle detection step S700. However, the counter-vehicle detection step S700 detects the other vehicle in comparison with the feature including the shape and color of the vehicle, while the traffic sign recognition step S730 is a feature including the shape and color of the traffic sign The traffic sign can be detected.

In addition, the traffic sign recognition method of the present invention can perform all the operations performed by the traffic sign recognition apparatus of the present invention described with reference to FIGS. 1 to 6.

While the present invention has been described in connection with what is presently considered to be the most practical and preferred embodiments, it is to be understood that the invention is not limited to the disclosed embodiments. That is, within the scope of the present invention, all of the components may be selectively coupled to one or more of them. The foregoing description is merely illustrative of the technical idea of the present invention, and various changes and modifications may be made by those skilled in the art without departing from the essential characteristics of the present invention. The scope of protection of the present invention should be construed according to the following claims, and all technical ideas within the scope of equivalents should be construed as falling within the scope of the present invention.

Claims (8)

An opposing vehicle detection unit that acquires an image using a camera that detects the front of the vehicle, and detects an opposing vehicle from the acquired image;
A driving lane determining unit for setting the center of gravity of the vehicle in the vertical direction and determining the driving lane of the vehicle based on the relationship between the center reference and the opposing vehicle;
A ROI setting unit that sets three ROIs having a certain range for the image and sets an analysis order of the ROI based on the ROI; And
A traffic sign recognition unit for recognizing the traffic signs in the three areas of interest according to the analysis procedure and terminating the traffic sign recognition if one or more traffic signs are recognized in one or more areas of interest among the three areas of interest according to the analysis sequence; Includes traffic sign recognition device. The method according to claim 1,
Wherein the ROI setting unit comprises:
And sets the analysis order to the upper interest area, the right interest area, and the left interest area in order, if the driving lane is judged to be right, sets the upper order interest area, the right interest area and the left interest area, And sets the analysis order in order of the upper interest region, the left interest region, and the right interest region if the driving lane is determined to be left. The method according to claim 1,
The opposing vehicle detecting unit includes:
Calculating an absolute speed of the other vehicles based on the speed of the vehicle and the relative speed of the other vehicles obtained through the image, comparing the absolute speed of the other vehicles with a preset speed threshold, Wherein the absolute speed of the opponent vehicle is smaller than the speed threshold value. The method according to claim 1,
The driving-
When the number of counter vehicles detected on the right side of the center reference is larger than the number of counter vehicles detected on the left side of the center reference and the number of opposing vehicles detected on the right side of the center reference is greater than the number threshold, The number of opposing vehicles detected on the left side of the center reference is greater than the number of opposing vehicles detected on the right side of the center reference and the number of opposing vehicles detected on the left side of the center reference Is greater than the number threshold value, determines the driving lane to the right. The method of claim 1,
Wherein the ROI setting unit comprises:
Further comprising a function of resetting the predetermined range, and when the traffic sign recognition unit does not recognize the traffic sign, extends the predetermined range. A driving lane determining step of acquiring an image using a camera that detects the front of the vehicle and detecting an opposite vehicle in the acquired image to determine a driving lane of the vehicle;
A ROI setting step of setting three ROIs having a certain range with respect to the image and setting an analysis order of the ROI based on the driving lane; And
Recognizing the traffic sign in the area of interest according to the analysis procedure and terminating the traffic sign recognition if the traffic sign is recognized in one or more areas of interest among the three areas of interest according to the analysis sequence How to recognize traffic signs. The method according to claim 6,
The method of claim 1,
Sets the analysis order in the order of the upper interest region, the right interest region, and the left interest region when the driving lane is determined to be right, And the analysis order is set in the order of the upper interest area, the left interest area, and the right interest area, if the driving lane is determined as the left side. The method according to claim 6,
The method of claim 1,
Further comprising the step of resetting the predetermined range, and if the traffic sign is not recognized in the traffic sign recognition step, the predetermined range is widened.

Images