JP2016042226A - Traffic light detection device and traffic light detection method - Google Patents

Abstract

PROBLEM TO BE SOLVED: To accurately detect a traffic light ahead of a vehicle which the vehicle should follow.SOLUTION: A traffic light detection device is configured to perform the steps of: acquiring an image by imaging surroundings of a vehicle by using a vehicle-mounted imaging part (11) and detecting a vehicle's own position; estimating a relative position of traffic lights relative to the vehicle on the basis of map information containing position information on the traffic lights around the vehicle and the vehicle's own position; configuring a detection area for the position of the traffic lights on the image predicted based on the relative position of the traffic lights (S01); and detecting the traffic lights from the detection area configured on the image and determining the traffic light which the vehicle should follow (S09). In a case where the traffic lights predicted to be imaged on the detection area are two or more, if all the two or more traffic lights can be detected from the detection area (YES at S07), the traffic light which the vehicle should follow is determined among all the two or more traffic lights (S09).SELECTED DRAWING: Figure 4

Description

  The present invention relates to a traffic signal detection apparatus and a traffic signal detection method.

  2. Description of the Related Art Conventionally, an outside-vehicle recognition device that recognizes a traffic light or a stop sign from an image obtained by capturing the traveling direction of a vehicle is known in order to enable appropriate traveling control on the vehicle (Patent Document 1). In Patent Document 1, when a plurality of traffic lights or stop signs appear in an image, the traffic light or stop sign that is most visible in the image is determined as a traffic light or a stop sign that is closest to the vehicle. .

JP 2007-257299 A

  However, the traffic light may be blocked by the presence of a preceding vehicle or a roadside tree and may not appear in the image. When the near-side traffic signal that is present at the closest position to the vehicle is blocked, in Patent Document 1, another traffic signal that appears in the image is erroneously recognized as the traffic signal that is present at the closest position to the vehicle.

  The present invention has been made in view of the above problems, and an object thereof is to provide a traffic signal detection device and a traffic signal detection method capable of accurately detecting a traffic signal on the near side that a vehicle should follow.

  A traffic light detection apparatus according to an aspect of the present invention captures an image of the surroundings of a vehicle using an imaging unit mounted on the vehicle, and detects the self-position of the vehicle. Estimate the relative position of the traffic signal to the vehicle from the map information including the traffic signal position information around the vehicle and the self-position, set the detection area at the position of the traffic signal on the image predicted from the relative position of the traffic light, A traffic light is detected from the detection area set on the image, and the traffic light to be followed by the vehicle is specified. If the number of traffic signals predicted to appear in the detection area is two or more and all of the two or more traffic lights can be detected from the detection area, the traffic signal to be followed by the vehicle from all of the two or more traffic lights is determined. Identify.

  According to the traffic signal detection apparatus and the traffic signal detection method, when only some traffic signals are reflected in one detection region among two or more traffic signals predicted to be captured in one detection region, By detecting the traffic signal from the vehicle, erroneous recognition of the traffic signal on the near side is suppressed, and the traffic signal on the near side that the vehicle should follow can be detected with high accuracy.

FIG. 1 is a block diagram illustrating information input to and output from the traffic signal detection apparatus 100 according to the embodiment. FIG. 2 is a block diagram illustrating the configuration and data flow of the traffic light detection apparatus 100 according to the embodiment. FIG. 3 is a block diagram showing a detailed configuration of the detection target specifying unit 20 of FIG. FIG. 4 is a flowchart illustrating an example of detailed processing procedures of the detection target specifying unit 20 and the traffic light detection unit 28. FIG. 5 is a table for explaining an example of determination in step S07 of FIG. FIG. 6 is a table for explaining an example of determination in step S11 of FIG. FIG. 7 is a table for explaining an example of determination in step S13 of FIG.

  Hereinafter, embodiments of the present invention will be described with reference to the drawings. The same members are denoted by the same reference numerals and the description thereof is omitted.

  With reference to FIG. 1, the information input / output to / from the traffic signal detection apparatus 100 according to the embodiment will be described. The traffic signal detection apparatus 100 detects a traffic signal installed around a road on which the vehicle travels from an image captured by an imaging unit (camera) mounted on the vehicle.

  Map information D02, landmark information D01, and camera information D03 are input to the traffic signal detection apparatus 100. The map information D02 includes traffic signal position information associated in advance between the real environment and the map. The landmark information D01 is used to calculate the vehicle's own position in the real environment. The landmark includes a feature (ground landmark) provided on the ground and a GPS satellite that transmits a GPS signal that can be received by the vehicle. In the embodiment, a ground landmark will be described as an example. The landmark information D01 includes, for example, ground landmark position information. The camera information D03 is used to extract a video around the vehicle (for example, forward) from the imaging unit. The traffic signal detection apparatus 100 outputs a traffic signal detection result as traffic signal information D04 based on the information D01 to D03.

  With reference to FIG. 2, the structure and data flow of the traffic light detection apparatus 100 according to the embodiment will be described. The traffic signal detection apparatus 100 includes an imaging unit 11, a self-position detection unit 12, a traffic signal position estimation unit 13, a detection target specifying unit 20, and a traffic signal detection unit 28.

  The imaging unit 11 is mounted on a vehicle and captures an image of the surroundings of the vehicle. The imaging unit 11 is a camera including a solid-state imaging device, for example, a CCD and a CMOS, and acquires an image that can be processed. The imaging unit 11 sets the angle of view of the lens and the vertical and horizontal angles of the camera based on the camera information D03, and outputs the acquired image as image data D08.

  The self position detector 12 detects the self position of the vehicle based on the landmark information D01. The landmark information D01 is, for example, information on the relative position of the ground landmark (store, famous place, sightseeing spot) with respect to the vehicle detected by a vehicle-mounted camera or a sensing means such as a laser radar. In the map information D02, position information of ground landmarks is registered in advance. The self-position of the vehicle can be detected by comparing the landmark information D01 with the information on the relative position of the ground landmark. Here, the “position” includes coordinates and orientation. Specifically, the position of the ground landmark includes the coordinates and orientation of the ground landmark, and the position of the vehicle includes the coordinates and orientation of the vehicle. The self-position detector 12 outputs the coordinates (x, y, z) in the reference coordinate system and the posture (pitch, yaw, roll) that is the rotation direction of each coordinate axis as the self-position information D05.

  For example, the self-position detecting unit 12 detects the initial position of the vehicle using the landmark information D01. The initial position is a position of the vehicle that can be directly obtained from the landmark information D01, that is, a coordinate and a posture. Then, the vehicle's own position is calculated by cumulatively adding the movement amount of the vehicle to the initial position. The self-position detection unit 12 can estimate the amount of movement of the vehicle per unit time, that is, the amount of change in coordinates and posture, using odometry, a radar device, a gyro sensor, a yaw rate sensor, and a rudder angle sensor.

  The traffic signal position estimation unit 13 estimates the relative position of the traffic signal with respect to the vehicle from the map information D02 and the self-position information D05. In the map information D02, traffic signal position information (coordinate information) is registered in advance. The relative coordinates of the traffic signal with respect to the vehicle can be obtained from the coordinates of the traffic signal and the coordinates and posture of the vehicle. The traffic signal position estimation unit 13 outputs the estimated relative coordinates of the traffic signal as relative position information D06. The traffic signal whose relative position is estimated by the traffic signal position estimation unit 13 is a traffic signal that presents a signal to the vehicle, in other words, a traffic signal that the vehicle should follow. For example, the traffic signal position estimating unit 13 extracts the intersection where the traffic signal is installed that is the nearest intersection where the vehicle will enter from now on as the “nearest intersection”. At least one traffic signal that is present at the “closest intersection” and that the vehicle of the self-location information should follow is identified as the “closest traffic signal”. Then, the relative position of the “nearest signal” with respect to the vehicle is output as relative position information D06.

  The detection target specifying unit 20 specifies, as a detection target, a traffic light on the near side that should be followed by the vehicle among the traffic lights predicted to appear in the image. Specifically, the detection target specifying unit 20 estimates the position of the traffic light on the image from the relative position of the traffic light with respect to the vehicle. And the detection area | region where it is estimated that a traffic light is reflected in the image is set. The detection area specified by the detection target specifying unit 20 is output as detection target information D09. The detailed configurations of the detection target specifying unit 20 and the traffic light detection unit 28 will be described later with reference to FIG. Detailed processing flows of the detection target specifying unit 20 and the traffic light detection unit 28 will be described later with reference to FIG.

  When all of the traffic signals predicted to appear in the detection region set by the detection target specifying unit 20 can be detected from the detection region, the traffic signal detection unit 28 has a traffic signal shielded by the presence of a preceding vehicle or a roadside tree. It can be judged not to. Therefore, the traffic signal detection unit 28 specifies the near side traffic signal to be followed by the vehicle from the detection area.

  On the other hand, when a part of the traffic signal predicted to appear in the detection area cannot be detected from the detection area, it can be determined that there is a traffic signal that is blocked by the presence of the preceding vehicle or the roadside tree. Therefore, the detection target specifying unit 20 does not specify the near side traffic signal to be followed by the vehicle from the detection area.

  The traffic light detection unit 28 detects the presentation color of the traffic light that the specified vehicle should follow from the detection area including the detection target. The detection result is output as traffic signal information D04.

  In addition, the self-position detection part 12, the traffic signal position estimation part 13, the detection target specific | specification part 20, and the traffic signal detection part 28 are realizable using the microcontroller provided with CPU, memory, and an input-output part, for example. A plurality of information processing units (12, 13, 20, 28) included in the traffic light detection device 100 are configured by the CPU executing a computer program installed in advance in the microcontroller.

  With reference to FIG. 3, the detailed structure of the detection target specific | specification part 20 and the traffic signal detection part 28 of FIG. 2 is demonstrated. The detection target specifying unit 20 includes a detection area setting unit 21, a priority order setting unit 22, and a traffic light number calculation unit 23.

  The detection area setting unit 21 sets, in the image, a detection area in which the traffic light is predicted to be captured from the relative position of the traffic light. Since the imaging unit 11 is fixed to the vehicle, the position on the image where the traffic light is expected to be captured can be specified in the image if the angle of view and direction taken by the imaging unit 11 are determined. . Based on the position on the image, the detection area setting unit 21 sets a detection area in which the traffic light is predicted to be captured in the image. The detection area setting unit 21 can set a plurality of detection areas in accordance with the number of traffic signals expected to appear in the image and the distribution of traffic signals. For example, when a traffic signal located on the own lane side where the vehicle travels and a traffic signal located on the opposite lane side are predicted as the “closest traffic signal”, an area on the own lane side of the image, and an area on the opposite lane side A detection area may be set for each of the above. The detection area setting unit 21 outputs a detection area where the traffic signal is predicted to be captured as “detection area information”.

  The priority order setting unit 22 sets a priority order between two or more detection areas set by the detection area setting unit 21 based on the possibility that the traffic signal is blocked. Depending on the position of the traffic signal with respect to the road and the position of the traffic signal with respect to the vehicle, there is a difference in the likelihood that the traffic signal will be reflected in the image, that is, the traffic signal may be blocked by obstacles around the vehicle. For example, due to the presence of a preceding vehicle, a traffic signal located on the own lane side where the vehicle travels is easily shielded, but a traffic signal located on the opposite lane side is difficult to shield. The priority of the detection area including the traffic signal located on the opposite lane side is set higher than the detection area including the traffic signal located on the own lane side. As described above, when two or more detection areas are set in the image, the priority order setting unit 22 sets the priority order between the two or more detection areas so that the lower the possibility that the traffic signal is blocked, the higher the priority order is. Set the priority.

  The traffic light number calculation unit 23 calculates the number of traffic signals predicted to appear in the detection area from the relative position of the traffic light. For example, when a vehicle is traveling on a straight road, a plurality of traffic lights appear close to each other on an image obtained by imaging the front of the road. For this reason, a plurality of traffic lights may be included in one detection area. By using the “detection area information” and the relative position information D06, it is possible to estimate the number of traffic signals expected to be included in the detection area. The calculation result is output as “estimated signal number information”.

  The traffic light detection unit 28 includes a traffic light number detection unit 24, a comparison unit 25, a distance determination unit 26, and a size estimation unit 27.

  The traffic signal number detection unit 24 detects the number of traffic signals actually reflected in the detection area using the image data D08 of the detection area. As the detection area image data D08, the detection area data cut out from the entire image data D08 can be used. Alternatively, the detection area image data D08 may be enlarged image data obtained by enlarging the detection area. Thereby, since the resolution of a detection area increases, the recognition accuracy of a traffic light improves. In this case, what is necessary is just to control the attitude | position with respect to the vehicle of the imaging part 11, and the angle of view of the imaging part 11 so that an enlarged image can be acquired. In other words, the enlarged image may be captured by narrowing down the entire image space to the detection area using the imaging unit 11 having the pan / tilt / zoom mechanism. Similarly, the traffic signal detection unit 28 uses detection area data cut out from the entire image data D08 or enlarged image data obtained by enlarging the detection area when detecting the presentation color of the traffic signal to be detected. be able to.

  The comparison unit 25 compares the number of traffic lights calculated by the traffic signal number calculation unit 23 with the number of traffic lights detected by the traffic signal number detection unit 24, and determines whether or not they match. In other words, the comparison unit 25 is configured such that only some of the two or more traffic signals predicted to appear in the detection region are actually reflected in the detection region, or all of the two or more traffic signals are detected in the detection region. It is judged whether it is actually reflected in.

  The distance determination unit 26 estimates a distance between two or more traffic signals predicted to appear in the detection area, and determines whether the estimated distance is longer than a predetermined threshold value. The distance determination unit 26 can estimate the distance between two or more traffic lights, specifically the distance in the traveling direction of the vehicle, by using the relative position information D06 and the map information D02.

  If the distance between two or more traffic lights is short, the difference in the size of the traffic lights appearing on the image becomes small. For this reason, it becomes difficult to specify the traffic signal on the near side that the vehicle should follow on the basis of the size of the traffic signal in the image. Specifically, it becomes difficult for the size estimation unit 27 to determine whether or not the size of the traffic light shown on the image matches the size of the traffic light on the near side estimated by the size estimation unit 27 described later. There is a risk of misrecognizing the traffic light on the near side that the vehicle should follow. On the other hand, if the distance between two or more traffic lights is sufficiently long, the difference in the size of the traffic lights appearing on the image will increase, and the traffic lights on the near side that the vehicle should follow should be specified based on the size of the traffic light appearing on the images. It becomes easy to do.

  Therefore, the distance determination unit 26 sets a threshold value in advance in order to determine whether or not it is possible to specify the traffic signal on the near side that the vehicle should follow on the basis of the size of the traffic signal that appears in the image. To do.

  The size estimation unit 27 estimates the size of the traffic signal on the image predicted to appear in the detection area from the relative position of the traffic light. Specifically, the size estimation unit 27 estimates the size of the traffic light on the image from the relative position of the traffic light on the near side that the vehicle should follow. When the distance determination unit 26 determines that the distance between two or more traffic signals is longer than the threshold value, it is possible to specify the traffic signal on the near side that the vehicle should follow based on the size of the traffic signal reflected in the image. . Therefore, in this case, the traffic signal detection unit 28 in FIG. 2 selects a traffic signal closest to the size of the traffic signal on the image estimated by the size estimation unit 27 from two or more traffic signals, and selects the presentation color of the traffic signal. To detect.

  With reference to FIG. 4, an example of a detailed processing procedure of the detection target specifying unit 20 and the traffic light detection unit 28 will be described. The example shown in FIG. 4 shows an example in which the detection area setting unit 21 sets two or more detection areas in the image.

  First, in step S01, the detection area setting unit 21 sets, in the image, two or more detection areas in which a traffic light is predicted to be captured from the relative position of the traffic light. Proceeding to step S03, the priority order setting unit 22 sets the priority order between two or more detection areas set by the detection area setting unit 21 based on the possibility that the traffic light is blocked.

  In step S05, the detection target specifying unit 20 selects one detection area from two or more detection areas. For example, the detection area with the highest priority is selected. Proceeding to step S07, the traffic light number calculation unit 23 calculates the number of traffic signals predicted to appear in the detection area from the relative position of the traffic light. The traffic signal number detection unit 24 detects the number of traffic signals actually reflected in the detection area using the image data D08 of the detection area. Then, the comparison unit 25 compares the number of traffic lights calculated by the traffic signal number calculation unit 23 with the number of traffic lights detected by the traffic signal number detection unit 24, and determines whether or not they match.

  If the two match, that is, if all of the two or more traffic lights can actually be detected in the detection area (YES in S07), the traffic signal shielded by the presence of the preceding vehicle or the roadside tree is in the detection area. It can be determined that it does not exist. Then, it progresses to step S09 and the traffic signal detection part 28 specifies the near side traffic signal which a vehicle should follow from the said detection area | region. Specifically, among the two or more traffic signals reflected in the detection area, the traffic signal having the maximum size on the image is specified as the traffic signal on the near side that the vehicle should follow. Thereafter, the process proceeds to step S19.

  On the other hand, if the two do not match, that is, if only some of the traffic lights predicted to appear in the detection area are actually detected in the detection area (NO in S07), the preceding vehicle or street It can be determined that a traffic signal shielded by the presence of a tree is present in the detection area. Then, it progresses to step S11 and the distance determination part 26 estimates the distance between two or more traffic lights estimated to appear in a detection area by using relative position information D06 and map information D02. Then, it is determined whether or not the estimated distance is longer than a predetermined threshold value.

  When the distance is equal to or smaller than a predetermined threshold value (NO in S11), the difference in the size of the traffic signal appearing on the image becomes small. It can be determined that it cannot be specified based on the standard. Note that the same determination can be made when the distance between two or more traffic lights cannot be estimated in step S11. Then, it progresses to step S17 and the traffic signal detection part 28 selects another detection area from two or more detection areas. For example, the detection area with the next highest priority is selected. Thereafter, the process returns to step S07.

  On the other hand, if the distance is longer than a predetermined threshold value (YES in S11), it can be determined that the near-side traffic signal to be followed by the vehicle can be specified based on the size of the traffic signal reflected on the image. Then, it progresses to step S13 and the size estimation part 27 estimates the magnitude | size on the image of the said traffic signal from the relative position of the traffic signal of the near side which a vehicle should follow. Then, the size estimation unit 27 determines whether there is a traffic signal having a size equal to the estimated size among the traffic signals reflected in the detection area.

  When a traffic light having a size equal to the estimated size is reflected (YES in S13), it can be determined that the front traffic light to be followed by the vehicle is not shielded. Then, it progresses to step S15 and the detection target specific | specification part 20 specifies the said signal apparatus as a detection target. Thereafter, the process proceeds to step S19.

  On the other hand, when a traffic light having a size equal to the estimated size is not shown (NO in S13), it can be determined that the traffic light on the near side that the vehicle should follow is shielded. Therefore, it progresses to step S17 and the detection area | region with the next highest priority is selected. Thereafter, the process returns to step S07.

  In step S19, the presenting color of the traffic light identified in step S09 or step S15 is detected.

  Here, the processing procedure shown in FIG. 4 is classified as the first to fourth embodiments as follows.

  In the first embodiment, all of the two or more traffic signals are actually detected in the detection region (YES in S07), and the traffic signal having the largest size on the image among the two or more traffic signals reflected in the detection region is selected. It is specified as a traffic light to be followed by the vehicle (S09).

  In the second embodiment, only some of the two or more traffic signals predicted to appear in the detection area are actually detected in the detection area (NO in S07) and predicted to appear in the detection area. The distance between the two or more traffic signals is equal to or less than a predetermined threshold value (NO in S11), and another detection area is selected (S17).

  In the third embodiment, only some of the two or more traffic signals predicted to appear in the detection area are actually detected in the detection area (NO in S07) and predicted to appear in the detection area. The distance between the two or more traffic signals is longer than a predetermined threshold (YES in S11), and no traffic signal having a size equal to the estimated size is captured (NO in S13). Is selected (S17).

  In the fourth embodiment, only some of the two or more traffic signals predicted to appear in the detection area are actually detected in the detection area (NO in S07) and predicted to appear in the detection area. Since the distance between the two or more traffic signals is longer than a predetermined threshold (YES in S11) and a traffic signal having a size equal to the estimated size is reflected (YES in S13), the traffic signal is detected. The target is specified (S15).

  With reference to FIG. 5, the determination example of step S07 of FIG. 4 will be described. In all the embodiments, the traffic light number calculation unit 23 calculates the number of traffic signals predicted to appear in the detection region RG as 2. In the first embodiment, the traffic signal number detection unit 24 detects the number of traffic signals (TS1, TS2) actually reflected in the detection region RG as 2. On the other hand, in the second to fourth embodiments, the rear traffic light (TS2) is shielded by the shield SH (for example, a preceding vehicle or a roadside tree). For this reason, the traffic signal number detection unit 24 detects the number of traffic signals TS1 actually reflected in the detection region RG as 2.

  Therefore, in the first embodiment, the comparison unit 25 determines that the calculated number of traffic lights matches the number of detected traffic lights, and in step S07, all of the two or more traffic lights are actually detected in the detection region. Judge that On the other hand, in the second to fourth embodiments, the comparison unit 25 determines that the calculated number of traffic signals and the number of detected traffic signals do not match, and in step S07, a part of two or more traffic signals is detected. It is determined that it is not actually detected.

  With reference to FIG. 6, the example of determination of step S11 of FIG. 4 is demonstrated. The determination in step S11 is performed in the second to fourth examples. In the third and fourth embodiments, the distance determination unit 26 has a distance between two or more traffic signals (TS1, TS2) predicted to appear in the detection region RG longer than a predetermined threshold (S11). YES). On the other hand, in the second embodiment, the distance determination unit 26 has a distance between two or more traffic signals (TS1, TS2) predicted to appear in the detection region RG is equal to or less than a predetermined threshold (S11). NO).

  With reference to FIG. 7, the determination example of step S13 of FIG. 4 will be described. The determination in step S13 is made when the distance between two or more traffic signals (TS1, TS2) predicted to appear in the detection region RG is longer than a predetermined threshold in the third and fourth embodiments. (YES in S11) is performed. If the distance between the traffic lights (TS1, TS2) is long, the traffic lights (TS1, TS2) can be identified based on the size of the traffic lights. Therefore, the size estimation unit 27 estimates the size of the traffic light (EST) on the image from the relative position of the traffic light (EST) on the near side that the vehicle should follow.

  In the fourth embodiment, among the two or more traffic signals (TS1, TS2), the rear traffic signal TS2 is shielded by the shield SH and is not reflected in the image data D08. It is reflected. Therefore, the size estimation unit 27 compares the size of the estimated traffic light (EST) with the traffic light (TS1) actually reflected. Then, it is determined that the traffic signal (TS1) having the same size as the estimated traffic signal (EST) is reflected in the detection region RG (YES in step S13).

  On the other hand, in the third embodiment, among the two or more traffic signals (TS1, TS2), the traffic signal TS1 on the near side that the vehicle should follow is shielded by the shielding object SH and is not reflected in the image data D08. TS2 is shown in the image data D08. Therefore, the size estimation unit 27 compares the size of the estimated traffic light (EST) with the actual traffic light (TS2). Then, it is determined that a traffic light having a size equal to the estimated traffic light (EST) is not captured in the detection region RG (NO in step S13).

  As described above, according to the embodiment, the following operational effects can be obtained.

  The vehicle needs to travel according to the traffic light on the near side that is present at the closest position on the travel path. However, the traffic lights may be hidden by factors (shields) that cannot be specified from the map information, such as preceding vehicles, roadside trees, and construction, and may not be visible from the vehicle. In this case, the blocked traffic signal does not appear in the image. As shown in FIG. 7, when the front traffic light (TS1) is shielded and the back traffic light (TS2) ahead of the front traffic light (TS1) appears in the image, the back traffic light (TS2) ) May be misrecognized as a traffic light on the near side. When the display colors of the traffic light on the near side and the traffic light on the far side are different, it is difficult to realize appropriate traveling control for the vehicle.

  The image acquired by the imaging unit 11 shows not only the traffic light on the near side but also the traffic light on the far side ahead. However, as shown in FIG. 5, it is difficult to clearly identify the front traffic light from the position and size of the traffic light on the image.

  Therefore, when only some of the two or more traffic signals predicted to appear in one detection region are detected in one detection region (NO in S07), it is predicted to appear in one detection region. It is determined that there is a traffic signal that is blocked by the presence of a preceding vehicle or a roadside tree. If it is difficult to determine whether or not the shielded traffic signal is a traffic signal on the near side from the position and size of the traffic signal on the image, the vehicle displays the traffic signal reflected in one detection area. A signal is detected from other detection areas without specifying the signal to be followed (S17).

  Thereby, the misrecognition of the traffic signal on the near side is suppressed, and the traffic signal on the near side that the vehicle should follow can be detected with high accuracy.

  When all the predicted traffic signals are detected in one detection area (YES in S07), it can be determined that there is no traffic signal shielded by the presence of a preceding vehicle or a roadside tree. Therefore, the traffic signal on the near side that the vehicle should follow can be detected from one detection region with high accuracy.

  When the imaging unit 11 mounted on the vehicle acquires an image, there is a difference in the possibility of the traffic signal being blocked by an obstacle around the vehicle, depending on the position of the traffic signal with respect to the road, that is, how easily the traffic signal is reflected in the image. is there. Therefore, when two or more detection areas are set in the image, the priority order setting unit 22 sets a priority order between the two or more detection areas. Then, the traffic light detection unit 28 selects a detection area having the second highest priority after the one detection area as another detection area, and detects the traffic signal. Thereby, it is possible to accurately detect the traffic signal on the near side that the vehicle should follow. In addition, it is possible to detect a traffic light at an early stage and reduce the processing load.

  When the distance between two or more traffic signals is less than or equal to the threshold value (NO in S11), as shown in FIG. 6, due to the difference in the size of the traffic signals (TS1, TS2) on the image, It is difficult to determine whether the traffic light is on the near side. Therefore, the traffic light reflected in one detection area is not set as a detection target, and detection of the traffic light is attempted from other detection areas. Thereby, the misrecognition of the traffic signal on the near side is suppressed, and the traffic signal on the near side that the vehicle should follow can be detected with high accuracy.

  The detection area setting unit 21 controls the attitude of the imaging unit 11 with respect to the vehicle and the angle of view of the imaging unit 11 so that an enlarged image obtained by enlarging the detection area can be acquired. Thereby, since the resolution of a detection area increases compared with the case where one detection area is cut out from an image, the recognition accuracy of a traffic light improves.

(First modification)
In the first and second modifications, a modification according to the first embodiment will be described. In the first modified example, when all of the two or more traffic signals are actually detected in the detection region (YES in S07), it is determined whether or not the distance between the two or more traffic signals is longer than the threshold value. . When the distance between the two or more traffic lights is longer than the threshold value, the size estimation unit 27 estimates the size of the traffic light on the near side that the vehicle should follow among the two or more traffic lights, and the estimated traffic light The traffic light closest to the size on the image is identified as the traffic light to be followed by the vehicle. That is, if YES is determined in step S07 in FIG. 4, step S11 is performed instead of step S09. If YES is determined in step S11 and YES is determined in subsequent step S13, the process proceeds to step S15, and a detection target is selected.

  According to the first modification, when the distance between two or more traffic lights is longer than the threshold value (YES in S11), the front traffic lights to be followed by the vehicle are determined based on the difference in the size of the traffic lights on the image. It can be specified with high accuracy. Therefore, erroneous detection of the traffic signal on the near side can be suppressed.

(Second modification)
In the second modified example, when all of the two or more traffic signals are actually detected in the detection region (YES in S07), it is determined whether or not the distance between the two or more traffic signals is longer than the threshold value. . When the distance between the two or more traffic lights is equal to or less than the threshold value, the detection target specifying unit 20 selects another detection area from among the two or more detection areas, for example, the detection area having the next highest priority. Select. Thereafter, the process returns to step S07. The same applies when the distance between two or more traffic lights cannot be estimated. That is, if YES is determined in step S07 in FIG. 4, step S11 is performed instead of step S09. If NO is determined in step S11, the process proceeds to step S17, another detection area is selected, and the process returns to step S07.

  According to the second modification, even when all of two or more traffic lights are actually detected in the detection area, if the distance between the two or more traffic lights is equal to or smaller than a threshold value, It is difficult to specify the traffic light on the near side because of the size of the signal. Therefore, the traffic light reflected in one detection area is not specified as a traffic light to be followed by the vehicle, and the traffic light is detected from other detection areas. Thereby, the misrecognition of the traffic signal on the near side is suppressed, and the traffic signal on the near side that the vehicle should follow can be detected with high accuracy.

  Although the contents of the present invention have been described with reference to the embodiments, the present invention is not limited to these descriptions, and it is obvious to those skilled in the art that various modifications and improvements can be made.

DESCRIPTION OF SYMBOLS 11 Image pick-up part 12 Self-position detection part 13 Traffic signal position estimation part 15 Traffic signal detection part 16 False detection degree evaluation part 20 Detection target specific | specification part 21 Detection area setting part 22 Priority order setting part 23 Traffic number calculation part 27 Size estimation part 28 Traffic signal detection Part 100 Traffic light detection device D02 Map information RG detection area

Claims (7)

An imaging unit mounted on a vehicle and capturing an image of the surroundings of the vehicle;
A self-position detector for detecting the self-position of the vehicle;
A traffic signal position estimation unit that estimates a relative position of the traffic signal to the vehicle from map information including positional information of traffic signals around the vehicle and the self-position;
A detection area setting unit for setting a detection area at the position of the traffic signal on the image predicted from the relative position of the traffic signal;
A traffic light detector that detects the traffic light from a detection area set on the image and identifies a traffic light to be followed by the vehicle;
The traffic signal detection unit is a case where there are two or more traffic signals predicted to appear in the detection area, and when two or more traffic signals are detected from the detection area, the two or more traffic signals are detected. A traffic light detection device that identifies a traffic light to be followed by the vehicle from among all traffic lights. A size estimation unit that estimates the size of the traffic light on the image predicted to be reflected in the detection area from the relative position of the traffic light;
The traffic signal detection unit is a case where the number of the traffic signals predicted to be captured in one of the detection areas is two or more and a part of the two or more traffic signals cannot be detected, and the two or more traffic signals are detected. When the distance between the two traffic lights is longer than a threshold value, the traffic light to which the vehicle should follow the traffic light closest to the size of the traffic light on the image estimated by the size estimation unit among the two or more traffic lights. The traffic light detection device according to claim 1, characterized by:   In the traffic signal detection unit, when two or more detection areas are set, the number of the traffic signals predicted to be captured in one of the detection areas is two and more, and the two or more traffic lights 3. The traffic light detection apparatus according to claim 1, wherein when a part of the traffic signal cannot be detected, the traffic light is detected from another detection area, and the traffic light to be followed by the vehicle is specified.   In the traffic signal detection unit, when two or more detection areas are set, the number of the traffic signals predicted to be captured in one of the detection areas is two and more, and the two or more traffic lights When a part of the signals cannot be detected and the distance between the two or more traffic signals predicted to be captured in the detection region is equal to or less than a threshold value or cannot be estimated, the traffic signals can be detected from other detection regions. The traffic light detection device according to any one of claims 1 to 3, wherein A priority setting unit that sets the priority between the two or more detection areas so that the priority is higher as the possibility that the traffic light is shielded is lower;
The traffic signal detection unit selects, as the other detection area, a detection area having the second highest priority after a detection area in which a part of the two or more traffic signals cannot be detected. 5. The traffic light detection device according to any one of 4.   6. The detection area setting unit according to claim 1, wherein the detection area setting unit controls an attitude of the imaging unit with respect to the vehicle and an angle of view of the imaging unit so that an enlarged image obtained by enlarging the detection area can be acquired. The traffic signal detection apparatus as described in any one of Claims. Using an imaging unit mounted on the vehicle, image the surroundings of the vehicle to obtain an image,
Detecting the position of the vehicle,
From the map information including the position information of traffic lights around the vehicle and the self-position, the relative position of the traffic lights to the vehicle is estimated,
Set a detection area at the position of the traffic light on the image predicted from the relative position of the traffic light;
A traffic signal detection method for detecting the traffic signal from a detection area set on the image and identifying a traffic signal to be followed by the vehicle,
When there are two or more traffic lights predicted to appear in the detection area, and all of the two or more traffic lights can be detected from the detection area, from among all the two or more traffic lights A traffic light detection method characterized by identifying a traffic light to be followed by the vehicle.

Images