JP7621801B2 - Traffic light recognition method and traffic light recognition device - Google Patents

Description

The present invention relates to a traffic light recognition method and a traffic light recognition device.

Patent Document 1 discloses a method for recognizing three-light traffic lights that have green, yellow, and red lights. This method extracts red, blue, and yellow parts from a color image input from a camera that captures a scene including a traffic light, and derives the circularity for each color part to extract traffic light candidates from the image.

JP 2005-301518 A

According to the method disclosed in Patent Document 1, the same judgment conditions are used to extract traffic light signals whether the vehicle is located outside the intersection or inside the intersection. Regardless of the relationship between the vehicle and the intersection, the judgment accuracy for judging traffic light signals is the same, which can make it difficult to make smooth traffic judgments.

The present invention was made in consideration of these problems, and its purpose is to provide a traffic light recognition method and a traffic light recognition device that enable smooth traffic judgment.

In a traffic light recognition method according to one aspect of the present invention, a controller estimates a traffic light area in an image where a target traffic light is located, and performs a light color determination to determine the color of the target traffic light based on a chromaticity determination range prepared for each light color of the target traffic light and the chromaticity of each pixel in the traffic light area. The controller determines whether or not a vehicle is located in the intersection, and if it is determined that a vehicle is located in the intersection, it performs a process of switching the chromaticity determination range to a narrower range than the chromaticity determination range that is set when it is determined that a vehicle is not located in the intersection.

The present invention enables smooth traffic decisions.

FIG. 1 is a block diagram showing the configuration of a traffic light recognition device according to this embodiment. FIG. 2 is a diagram illustrating a hue determination range for determining the color corresponding to a traffic light. FIG. 3 is an explanatory diagram showing the relationship between the position of a vehicle relative to an intersection and the lights of traffic lights. FIG. 4 is a flowchart showing the process flow of the traffic light recognition method according to the first embodiment. FIG. 5 is a flowchart showing the flow of a light color determination process according to the first embodiment. FIG. 6 is an explanatory diagram showing the concept of light color determination according to the second embodiment. FIG. 7 is a flowchart showing the process flow of the traffic light recognition method according to the second embodiment. FIG. 8 is a flowchart showing the flow of a light color determination process according to the second embodiment. FIG. 9 is a flowchart showing the process flow of the traffic light recognition method according to the third embodiment. FIG. 10 is a flowchart showing the flow of a light color determination process according to the third embodiment.

Embodiments of the present invention will be described below with reference to the drawings. In the description of the drawings, identical parts are given the same reference numerals and the description will be omitted.

The configuration of the traffic light recognition device according to this embodiment will be described with reference to FIG. 1. The traffic light recognition device includes an imaging unit 10 and a controller 30. The controller 30 is connected to the imaging unit 10 via wired or wireless communication. The controller 30 is also connected to the driving assistance device 50 via wired or wireless communication.

The imaging unit 10, the controller 30, and the driving assistance device 50 are mounted on a vehicle (not shown). However, the controller 30 may be installed outside the vehicle.

The imaging unit 10 captures images in the direction of travel of the vehicle, for example, in front of the vehicle. The imaging unit 10 is a digital camera equipped with a solid-state imaging element such as a CCD or CMOS, and captures images of the surroundings of the vehicle to obtain image data of the surrounding area. The imaging unit 10 captures an image of a specified range in front of the vehicle by setting the focal length, the lens angle of view, the vertical and horizontal angles of the camera, etc.

The images captured by the imaging unit 10 are input to the controller 30 and stored in a storage unit (not shown). The imaging unit 10 captures images at a predetermined cycle, and the images are input to the controller 30 sequentially.

The controller 30 processes the images input from the imaging unit 10. The controller 30 is a general-purpose microcomputer equipped with a CPU (Central Processing Unit), memory, and input/output units, and is an example of a control device or processing device. A computer program for causing the controller 30 to function as part of the traffic light recognition device is installed in the controller 30. By executing the computer program, the controller 30 functions as multiple information processing circuits equipped in the traffic light recognition device.

This controller 30 has multiple information processing circuits, including a position acquisition unit 31, a traffic light area setting unit 32, an intersection determination unit 33, a determination condition switching unit 34, a light color determination unit 35, and an output unit 36.

The position acquisition unit 31 acquires the position of the vehicle in real space. The traffic light area setting unit 32 converts the position of the vehicle in real space to the position of the vehicle on the map based on a map that is stored in advance and includes the position information of the intersection (stop line) and the traffic light. The traffic light area setting unit 32 sets a traffic light area on the image based on the position of the vehicle and the position of the target traffic light on the map. The traffic light area is an area on the image where the target traffic light, that is, the traffic light installed at the intersection located immediately adjacent to the vehicle in the traveling direction of the vehicle, exists. Since the installation range of the traffic light in the height direction is regulated by law, the area where the target traffic light exists on the image can be set based on the vehicle attitude, the imaging direction of the imaging unit 10, the position of the vehicle on the map, and the position of the target traffic light. In other words, the traffic light area can be set if the imaging direction of the imaging unit 10 taking into account the vehicle attitude and the relative position of the traffic light with respect to the vehicle are known. Note that the traffic light area setting unit 32 is not necessarily limited to setting traffic light areas based on a map as described above, and can also set traffic light areas by searching for the external shape of traffic lights in all areas of the image using pattern matching, for example.

The intersection determination unit 33 determines whether or not the vehicle is located in an intersection where a target traffic light is installed, based on the position of the vehicle on the map. The determination condition switching unit 34 switches the determination conditions used by the light color determination unit 35, which in this embodiment is the hue determination range (chromaticity determination range), depending on the determination result of the intersection determination unit 33.

The light color determination unit 35 performs light color determination to determine the color of the light of the target traffic light (the color of the light that is turned on). When each pixel of the image is expressed by an RGB value, the light color determination unit 35 converts the RGB value for each pixel from RGB space to HSV space to obtain the hue. As shown in FIG. 2, the chromaticity of each pixel is represented by a hue ranging from, for example, 0 degrees (red) through 180 degrees (blue) to 360 degrees (red).

The light color determination unit 35 holds a range (180-TH to 180+TH) that takes into account a predetermined range setting value TH on both the plus and minus sides, centered on the hue (180 degrees) that represents the light of the blue light device 111, as a determination range for a hue that corresponds to blue. Similarly, the light color determination unit 35 holds a range (60-TH to 60+TH) that takes into account a predetermined range setting value TH on both the plus and minus sides, centered on the hue (60 degrees) that represents the light of the yellow light device 112, as a determination range for a hue that corresponds to yellow. The light color determination unit 35 also holds a range (180-TH to 0+TH) that takes into account a predetermined range setting value TH on both the plus and minus sides, centered on the hue (0 degrees (360 degrees)) that represents the light of the red light device 113, as a determination range for a hue that corresponds to red.

The light color determination unit 35 determines the color of the light of the target traffic light based on the hue determination range prepared for each color of the light of the target traffic light and the hue of each pixel within the traffic light area.

The output unit 36 outputs the determination result of the light color determination unit 35. The output unit 36 outputs the determination result to, for example, the driving assistance device 50.

The driving assistance device 50 performs driving assistance to assist the driving of a vehicle (not shown) based on the traffic light recognition results obtained by the controller 30. The driving assistance may be, for example, driving the vehicle by automatic driving along a predetermined driving route, or determining whether to proceed or stop, or it may be assisting the driving operation of the vehicle occupant. The driving assistance may also be controlling a human interface (display device and speaker) to notify the vehicle occupant of the traffic light recognition results.

The driving assistance device 50 does not need to be configured as hardware independent of the controller 30. The functions executed by the driving assistance device 50 may be realized as one of a plurality of information processing circuits included in the controller 30.

Referring to Figure 3, the relationship between the position of the vehicle relative to the intersection and the traffic judgment regarding the light of the target traffic light will be explained. In Figure 3, the vehicle is indicated by the symbol A, and the target traffic light is indicated by the symbol 100.

The target traffic light 100 installed at an intersection is, for example, a three-light traffic light, and is a device that lights up to represent specific color signals, for example, green, yellow, and red. The target traffic light 100 is equipped with a blue light 111, a yellow light 112, and a red light 113. In the three-light traffic light main body 110, the three color lights 111, 112, and 113 are lined up horizontally. The blue light 111 is located at the left end, the yellow light 112 is located in the center, and the red light 113 is located at the right end.

Blue light unit 111 is a light unit that has a light source housed inside and provides blue lighting. The light shape provided by blue light unit 111 is round. The light provided by blue light unit 111 indicates a green traffic light. Like blue light unit 111, yellow light unit 112 and red light unit 113 are lights that have a light source housed inside and provide yellow and red lighting. The light shape provided by yellow light unit 112 and red light unit 113 is round. The light provided by yellow light unit 112 and red light unit 113 indicates a yellow traffic light and a red traffic light.

The meaning of the lights on a three-light traffic signal is determined by traffic laws that vehicles must follow. For example, a green light means "you may proceed," and a red light means "stop at the stop position." A yellow light means "stop at the stop position, unless you are too close to the stop position to stop safely."

Here, "green" light refers to the three-color light of the target traffic light 100, excluding yellow and red. In other words, the "green" light does not have to be blue in actual chromaticity, but may be a chromaticity equivalent to green. Furthermore, the target traffic light 100 is not limited to a horizontal traffic light, but may be a vertical traffic light with three colored lights 111, 112, 113 lined up vertically. Furthermore, the light shape may be rectangular.

If the distance D from vehicle A to the intersection, e.g., the stop line, is greater than zero, i.e., vehicle A is not located in the intersection, vehicle A or the driver makes the following judgment regarding the light of the target traffic light 100. That is, if the target traffic light 100 has a green light, vehicle A judges that it is OK to proceed into the intersection. On the other hand, if the target traffic light 100 has a yellow or red light, vehicle A judges that it will stop at the stop line without proceeding into the intersection.

If vehicle A is not located within the intersection, vehicle A will refrain from entering the intersection later. Therefore, it is preferable to be able to determine early on the color of the light indicating a stop at the target traffic light 100 so that vehicle A can enter the intersection only when the light is green. In this embodiment, red is exemplified as the color of the light indicating a stop at the target traffic light 100, but yellow may also be used.

On the other hand, if the distance D from vehicle A to the stop line of the intersection is zero or less, i.e., if vehicle A is located inside the intersection, vehicle A or the driver makes the following judgment regarding the light of the target traffic light 100. That is, if the target traffic light 100 is lit yellow or red, vehicle A judges that it is necessary to quickly exit the intersection.

When vehicle A is located in an intersection, it is in a situation where vehicle A is about to exit the intersection. In this case, it is preferable to exit the intersection quickly when it is guaranteed that an oncoming vehicle will not enter the intersection. Therefore, it is preferable to be able to accurately determine whether the target traffic light 100 is yellow or red without erroneously determining that the light is yellow or red. In this embodiment, red is exemplified as the color of the light indicating a stop at the target traffic light 100, but yellow may also be included.

In this way, the desired accuracy of determining the color of the lights differs depending on whether vehicle A is located inside the intersection.

The following describes a series of processes executed by the traffic light recognition device, i.e., the flow of the traffic light recognition method, with reference to FIG. 3. The flowchart shown in FIG. 3 is called up in response to the timing at which an image is captured by the imaging unit 10, and is executed by the controller 30.

First, the traffic light area setting unit 32 acquires the image input from the imaging unit 10 (step S10).

The position acquisition unit 31 acquires the position of the vehicle in real space (S11). For example, the position acquisition unit 31 acquires the position of the vehicle using a position estimation technique based on information input from a GPS (Global Positioning System) receiver (not shown). The position acquisition unit 31 may also acquire the position of the vehicle using the positional relationship between the object around the vehicle and the vehicle based on information about the object around the vehicle input from an object sensor such as a laser radar (not shown). In addition, the position acquisition unit 31 may acquire the position of the vehicle using a position estimation technique such as odometry based on information input from a sensor that acquires the behavior of the vehicle, such as an inertial navigation system, a sensor provided on the brake pedal or accelerator pedal, a wheel speed sensor, or a yaw rate sensor.

The traffic light area setting unit 32 sets a traffic light area on the image based on the position of the vehicle (step S12). The position in the image of a traffic light that is within the imaging range of the imaging unit 10 can be estimated based on the imaging direction of the imaging unit 10, the position and attitude of the vehicle at the time of imaging, and the position of the target traffic light relative to the vehicle. The determination unit 150 sets, for example, a part of the captured image that includes the estimated position of the target traffic light in the image, as the traffic light area.

Based on the vehicle's position, the intersection determination unit 33 calculates the distance D from the vehicle to the stop line of the intersection where the target traffic light is located in the vehicle's traveling direction (step S13). The intersection determination unit 33 can identify the position of the stop line by referring to the distance to the target traffic light or stop line obtained from an object sensor such as a laser radar, or by referring to the vehicle's position and map information around the vehicle.

The intersection determination unit 33 determines whether the distance D is zero or less (step S14). If the distance D is greater than zero (NO in step S14), the intersection determination unit 33 determines that the vehicle is located before the intersection. On the other hand, if the distance D is zero or less (YES in step S14), the intersection determination unit 33 determines that the vehicle is located inside the intersection.

Here, the intersection determination unit 33 determines that the vehicle is located before the intersection when the distance D is greater than zero, and determines that the vehicle is located inside the intersection when the distance D is equal to or less than zero, but this is not limited to the above example. For example, the vehicle speed of the vehicle may be detected, the stopping distance of the vehicle (the distance required for the vehicle to stop) may be calculated based on the vehicle speed, and the vehicle may be determined to be before the intersection when the distance D is greater than the stopping distance, and the vehicle may be determined to be inside the intersection when the distance D is equal to or less than the stopping distance. In other words, if the distance D is equal to or less than the stopping distance, it may be determined that the vehicle will definitely enter the intersection after that, and if the distance D is equal to or less than the stopping distance, it may be determined that the vehicle is located inside the intersection. In this way, before entering the intersection, it may be determined whether the vehicle is located inside the intersection based on the future stopping position of the vehicle. In this embodiment, as described above, it is determined whether the vehicle is located inside the intersection based on whether the distance D is equal to or less than zero.

If the distance D is greater than zero (NO in step S14), the judgment condition switching unit 34 sets the range setting value TH, which specifies the judgment range of the hue used by the light color judgment unit 35, to a first value (step S15). On the other hand, if the distance D is equal to or less than zero (YES in step S14), the judgment condition switching unit 34 sets the range setting value TH to a second value smaller than the first value (step S16). In other words, when it is determined that a vehicle is located in the intersection, the judgment condition switching unit 34 performs a process of switching the judgment range of the hue used to judge the light color to a narrower range than the judgment range of the hue set when it is determined that a vehicle is not located in the intersection.

The light color determination unit 35 performs a light color determination to determine the color of the light of the target traffic light (step S17). The light color determination process in step S17 will be described with reference to FIG. 5.

The light color determination unit 35 converts the RGB values for each pixel from RGB space to HSV space to obtain the hue of each pixel in the traffic light area (step S100). The light color determination unit 35 then obtains the hue of the traffic light area by statistically processing the hue of each pixel in the traffic light area. For example, the light color determination unit 35 determines the median hue or the most frequently occurring hue of the hues present in the traffic light area as the hue of the traffic light area.

If the hue of the traffic light area falls within the judgment range of the hue corresponding to red, the light color judgment unit 35 judges the hue of the traffic light area to be red (YES in step S101). Then, the light color judgment unit 35 judges the light of the target traffic light to be red (step S102).

If the hue of the traffic light area is not red (NO in step S101), the light color determination unit 35 determines whether the hue of the traffic light area is within the judgment range for a hue corresponding to yellow (step S103). If the hue of the traffic light area falls within the judgment range for a hue corresponding to yellow, the light color determination unit 35 determines the hue of the traffic light area to be yellow (YES in step S103). Then, the light color determination unit 35 determines the light of the target traffic light to be yellow (step S104).

If it is determined that the hue of the traffic light area is not yellow (NO in step S103), the light color determination unit 35 determines whether the hue of the traffic light area is within the determination range for a hue corresponding to blue (step S105). If the hue of the traffic light area is within the determination range for a hue corresponding to blue, the light color determination unit 35 determines that the hue of the traffic light area is blue (YES in step S105). Then, the light color determination unit 35 determines that the light of the target traffic light is blue (step S106).

If it is determined that the hue of the traffic light area is not blue (NO in step S105), the light color determination unit 35 determines that the color of the traffic light is unknown (step S107).

In this manner, in the traffic light recognition method of this embodiment, the controller 30 estimates the traffic light area in which the target traffic light is located from within the image, and performs a light color determination to determine the color of the target traffic light's light based on the chromaticity determination ranges prepared for each light color of the target traffic light and the chromaticity of each pixel within the traffic light area. The controller 30 determines whether or not a vehicle is located within the intersection, and if it is determined that a vehicle is located within the intersection, it performs processing to switch the chromaticity determination range to a narrower range than the chromaticity determination range that is set when it is determined that no vehicle is located within the intersection.

When a vehicle is outside the intersection, it is preferable to be able to determine the red light early, since it is a situation in which the vehicle is refraining from entering the intersection. On the other hand, when a vehicle is inside the intersection and the light state of the target traffic light changes to red, the vehicle needs to quickly exit the intersection. Therefore, when a vehicle is inside the intersection, it is preferable not to erroneously determine the red light. In this way, the preferable determination accuracy for determining the color of the light differs depending on whether or not the vehicle is inside the intersection. In this respect, according to the present embodiment, when it is determined that the vehicle is inside the intersection, the chromaticity determination range is switched to a narrower range than when it is determined that the vehicle is not inside the intersection. This makes it possible to suppress erroneous determination of the red light when the vehicle is inside the intersection. Therefore, the determination accuracy of the light color can be switched depending on the positional relationship between the host vehicle and the intersection, making it possible to make smooth traffic decisions.

In the above embodiment, the judgment condition switching unit 34 switches the chromaticity judgment range for all colors between inside and outside the intersection. However, the judgment condition switching unit 34 may switch only the chromaticity judgment range for red.

Second Embodiment
A traffic light recognition device and a traffic light recognition method according to a second embodiment will be described. Hereinafter, the description of the second embodiment will be omitted, and the differences from the first embodiment will be mainly described.

As shown in FIG. 6, in the traffic light recognition device of the second embodiment, the light color determination unit 35 stores the light color determination result for each frame of the image captured by the imaging unit 10. In the example shown in FIG. 6, the light color determination unit 35 stores the past eight consecutive determination results. Then, the light color determination unit 35 determines, as the light color determination result, the light color that has been determined a predetermined number of times, for example, five times or more, among the past eight determination results.

In the example shown in FIG. 6, the target traffic light has switched from a green light to a yellow light in the image of frame tc. The light color determination unit 35 determines that the light of the target traffic light is yellow based on the image of frame tc, but the light of the target traffic light has been blue five or more times out of the past eight determination results. Therefore, the light color determination unit 35 determines that the light color determination result is blue.

On the other hand, as the periodic capture of images progresses and the number of times the target traffic light is determined to be yellow based on the images of each frame increases, at the time the image of frame tj is acquired, the target traffic light will be yellow in five or more of the past eight determination results. As a result, the light color determination unit 35 determines yellow as the light color determination result.

With this configuration, although there is a certain delay before the color of the light of the target traffic light is determined as the result of the light color determination, the light color can be determined with high accuracy. In addition, if the number of times determination value used by the light color determination unit 35 is switched to a larger value, it becomes difficult to quickly determine the change in the light color, but erroneous determination of the light color can be suppressed.

The following describes a series of processes executed by the traffic light recognition device, i.e., the flow of the traffic light recognition method, with reference to FIG. 7. The flowchart shown in FIG. 7 is called up in response to the timing at which an image is captured by the imaging unit 10, and is executed by the controller 30.

In the flowchart shown in FIG. 7, the processes from step S10 to step S16 are the same as those shown in the first embodiment. In the traffic light recognition method according to this embodiment, the processes of step S20 and step S21 are added.

If the distance D is greater than zero (NO in step S14), the judgment condition switching unit 34 sets the range setting value TH (step S15) and sets the number judgment value THA used by the light color judgment unit 35 to the first number (step S20). The first number value is, for example, 5 times.

On the other hand, if the distance D is equal to or less than zero (YES in step S14), the judgment condition switching unit 34 sets the range setting value TH (step S16) and then sets the number of times judgment value THA to a second number of times greater than the first number of times (step S20). The second number of times is, for example, seven times. In other words, when it is determined that a vehicle is located within the intersection, the judgment condition switching unit 34 performs a process of switching the number of times judgment value THA for judging the light color to a value greater than the number of times judgment value THA that is set when it is determined that a vehicle is not located within the intersection.

The light color determination process will be described with reference to FIG. 8. In the flowchart shown in FIG. 8, the processes from step S100 to step S107 are the same as those shown in the first embodiment. In the traffic light recognition method according to this embodiment, the processes from step S110 to step S116 are added.

In each process from step S110 to step S113, the light color determination unit 35 stores the light color determination result. The light color determination unit 35 has a memory unit that can store, for example, eight determination results, and stores the past eight determination results including the most recent determination result.

The light color determination unit 35 determines whether there is a light color that is equal to or greater than the number of times determination value THA (step S114). If there is a light color that is equal to or greater than the number of times determination value THA (YES in step S114), the light color determination unit 35 determines the light color that is equal to or greater than the number of times determination value THA as the determination result of the light color determination (step S115). On the other hand, if there is no light color that is equal to or greater than the number of times determination value THA (NO in step S114), the light color determination unit 35 determines the determination result of the light color determination as unknown (step S116).

In this manner, in the traffic light recognition method of this embodiment, the controller 30 stores the light color judgment results for each image frame, and determines the light color judgment result for the light color that is equal to or exceeds a predetermined number of judgment results among the judgment results for a predetermined number of times. When it is determined that a vehicle is located within the intersection, the controller 30 further performs a process of switching the number of judgment value to a value greater than the number of judgment value that is set when it is determined that a vehicle is not located within the intersection.

According to this configuration, when it is determined that the vehicle is located at the intersection, the number judgment value is switched to a larger value than when it is determined that the vehicle is not located at the intersection. This makes it possible to prevent erroneous judgment of red lights when the vehicle is located at the intersection. Therefore, the judgment accuracy of the light color judgment can be switched depending on the positional relationship between the vehicle and the intersection, making it possible to make smooth traffic judgments.

Third Embodiment
A traffic light recognition device and a traffic light recognition method according to a third embodiment will be described. Hereinafter, the description of the second embodiment will be omitted, and the differences from the third embodiment will be mainly described.

In the traffic light recognition device of the third embodiment, the light color determination unit 35 calculates the circularity of a light determined to be red, yellow, or blue. Specifically, the light color determination unit 35 extracts a plurality of pixels corresponding to the light, and identifies the shape of a pixel set that is a set of the extracted pixels. The shape is identified by, for example, calculating the circularity, as shown in the following formula 1.
(Formula 1)
Circularity = 4π × (area of pixel group) ÷ (perimeter of pixel group) ²

The circularity shown in Equation 1 approaches 1 the closer the shape of the pixel group is to a circle. In other words, this circularity indicates the degree of similarity to the light shape of the target traffic light. If the circularity is equal to or greater than the shape judgment value (an example of a similarity judgment value), the light color judgment unit 35 determines the color of the traffic light based on the image as the judgment result of the light color judgment.

With this configuration, by evaluating the shape in addition to the color, the light color of the target traffic light can be determined with high accuracy. In addition, by switching the shape determination value used by the light color determination unit 35 to a larger value, erroneous determination of the light color can be suppressed.

The following describes a series of processes executed by the traffic light recognition device, i.e., the flow of the traffic light recognition method, with reference to FIG. 9. The flowchart shown in FIG. 9 is called up in response to the timing at which an image is captured by the imaging unit 10, and is executed by the controller 30.

In the flowchart shown in FIG. 9, the processes from step S10 to step S16 are the same as those shown in the first embodiment. In the traffic light recognition method according to this embodiment, the processes of step S22 and step S23 are added.

If the distance D is greater than zero (NO in step S14), the judgment condition switching unit 34 continues to set the range setting value TH (step S15), and sets the shape judgment value THB used by the light color judgment unit 35 to the first shape value (step S22). The first shape value is a value equal to or less than 1.

On the other hand, if the distance D is zero or less (YES in step S14), the judgment condition switching unit 34 sets the range setting value TH (step S16) and then sets the shape judgment value THB to the second shape value (step S23). The second shape value is a value equal to or less than 1 and is greater than the first shape value. In other words, when it is determined that a vehicle is located in the intersection, the judgment condition switching unit 34 performs a process of switching the shape judgment value THB, which determines the light color, to a value greater than the shape judgment value THB that is set when it is determined that a vehicle is not located in the intersection.

The light color determination process will be described with reference to FIG. 10. In the flowchart shown in FIG. 8, the processes from step S100 to step S107 are the same as those shown in the first embodiment. In the traffic light recognition method according to this embodiment, the processes from step S120 to step S123 are added.

In step S120, the light color determination unit 35 calculates the circularity of the light determined to be red, yellow, or blue (step S20). The circularity is calculated, for example, as shown in Equation 1.

The light color determination unit 35 determines whether the calculated circularity is equal to or greater than the shape determination value THB (step S121). If the circularity is equal to or greater than the shape determination value THB (YES in step S121), the light color determination unit 35 determines the color of the light whose circularity was calculated as the determination result of the light color determination (step S122). On the other hand, if the circularity is not equal to or greater than the shape determination value THB (NO in step S121), the light color determination unit 35 determines the determination result of the light color determination as unknown.

In this manner, in the traffic light recognition method of this embodiment, the controller 30 confirms the color of the light determined in the light color determination when the similarity between the shape of a set of pixels in the traffic light area, which is a set of pixels having chromaticity corresponding to the traffic light, and the shape of the traffic light is equal to or greater than a predetermined similarity determination value. When it is determined that a vehicle is located within the intersection, the controller 30 further performs a process of switching the similarity determination value to a value greater than the previous similarity determination value that is set when it is determined that a vehicle is not located within the intersection.

According to this configuration, when it is determined that the vehicle is located at the intersection, the judgment value is switched to a larger value than when it is determined that the vehicle is not located at the intersection. This makes it possible to prevent erroneous judgment of red lights when the vehicle is located at the intersection. Therefore, the judgment accuracy of the light color judgment can be switched depending on the positional relationship between the vehicle and the intersection, making it possible to make smooth traffic judgments.

In each of the above-described embodiments, the judgment conditions (hue judgment range, number judgment value THA, shape judgment value THB) for judging the traffic light color are switched after the vehicle enters the intersection. However, if the color of the target traffic light is judged to be red even once after switching the judgment conditions, the judgment condition switching unit 34 may restore the judgment conditions to the conditions before switching. In this way, once the color of the light indicating a stop is judged, the judgment conditions are restored, so that the color of the light can be stably judged thereafter.

The traffic light recognition method and traffic light recognition device have been described above using the first, second, or third embodiment. The traffic light recognition device according to each embodiment executes the traffic light recognition method described in the respective embodiment, and achieves the same actions and effects as the traffic light recognition method.

In the first, second or third embodiment described above, the traffic light recognition device includes a driving assistance device 50. However, the traffic light recognition device may be configured with only the imaging unit 10 and the controller 30.

In the first and second embodiments described above, the light shape of each color light that constitutes the main body of the three-light traffic light is circular. However, the light shape of each color light may be polygonal, such as rectangular.

In the first and second embodiments, an example is shown in which the multiple information processing circuits provided in the controller 30 are realized by software, but it is of course possible to configure the information processing circuits by preparing dedicated hardware for realizing the functions of each information processing circuit. Also, the multiple information processing circuits may be configured by individual hardware.

As described above, an embodiment of the present invention has been described, but the descriptions and drawings that form part of this disclosure should not be understood as limiting this invention. Various alternative embodiments, examples, and operating techniques will become apparent to those skilled in the art from this disclosure.

REFERENCE SIGNS LIST 10 Imaging unit 30 Controller 31 Position acquisition unit 32 Traffic light area setting unit 33 Intersection determination unit 34 Determination condition switching unit 35 Light color determination unit 36 Output unit 50 Driving support device

Claims (5)

An image of an area in front of the vehicle is captured using an imaging unit mounted on the vehicle;
A controller that processes the image input from the imaging unit,
A traffic light area including a target traffic light provided at an intersection located immediately adjacent to the vehicle in the traveling direction of the vehicle is estimated from the image;
performing a light color determination for determining the color of the light of the target traffic light based on a chromaticity determination range prepared for each color of the light of the target traffic light and the chromaticity of each pixel within the traffic light area;
determining whether the vehicle is located within the intersection;
A traffic light recognition method which, when it is determined that the vehicle is located within the intersection, switches the chromaticity judgment range for judging the light of a color indicating a stop at the target traffic light to a narrower range than the chromaticity judgment range which is set when it is determined that the vehicle is not located within the intersection. The controller:
storing a result of the light color determination for each frame of the image captured by the imaging unit;
A light color that is equal to or exceeds a predetermined number of judgment results among the judgment results of the predetermined number of times is determined as a judgment result of the light color judgment,
The traffic light recognition method according to claim 1, further comprising the step of: when it is determined that the vehicle is located within the intersection, switching the number judgment value to a value greater than the number judgment value that is set when it is determined that the vehicle is not located within the intersection. The controller:
In the traffic light area, when a similarity between a shape of a set of pixels, which is a set of pixels having a chromaticity corresponding to the traffic light of the traffic light, and a shape of the traffic light of the traffic light is equal to or greater than a predetermined similarity determination value, the color of the traffic light determined in the light color determination is confirmed;
3. The traffic light recognition method according to claim 1, further comprising the step of: when it is determined that the vehicle is located within the intersection, switching the similarity judgment value to a value greater than the previous similarity judgment value that is set when it is determined that the vehicle is not located within the intersection. The controller:
4. A traffic light recognition method as claimed in claim 1, further comprising: returning the judgment conditions to the conditions before the switching if the color of the light of the target traffic light is judged to be a color indicating a stop even once after the judgment conditions for judging the color of the light of the traffic light are switched. An imaging unit mounted on a vehicle to capture an image of an area in front of the vehicle;
a controller for processing the image input from the imaging unit;
The controller:
A traffic light area including a target traffic light provided at an intersection located immediately adjacent to the vehicle in the traveling direction of the vehicle is estimated from the image;
performing a light color determination for determining the color of the light of the target traffic light based on a chromaticity determination range prepared for each color of the light of the traffic light and the chromaticity of each pixel within the traffic light area;
determining whether the vehicle is located within the intersection;
When it is determined that the vehicle is located within the intersection, the traffic light recognition device performs a process of switching the chromaticity judgment range for judging the light color indicating a stop at the traffic light to a narrower range than the chromaticity judgment range that is set when it is determined that the vehicle is not located within the intersection.

Images