JP2007066003A - Stop-line detection device - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a stop-line detection device capable of accurately detecting a stop line by suppressing detection error of the stop line. <P>SOLUTION: In the stop-line detection device 1 for detecting the stop-line on a road, based on the image photographing the road, the stop-line detection device 1 detects a crosswalk of the road, and sets a stop-line detection area for detecting the stop-line so as not to include crosswalks, and detects the stop line in the stop-line detection area. Thus, erroneous detection of lines, forming crosswalks as the stop line, is prevented, and the detection accuracy of the stop line can be improved. <P>COPYRIGHT: (C)2007,JPO&INPIT

Description

  The present invention relates to a stop line detection device that detects a stop line drawn on a road.

2. Description of the Related Art Conventionally, as an apparatus for detecting a stop line drawn on a road, an apparatus for detecting a stop line based on an image obtained by photographing the front of the host vehicle is known, as described in JP-A-2002-190100. ing. This device sets a stop line detection area in front of a preceding vehicle traveling in front of the host vehicle, and detects a white line having a length of a predetermined ratio or more as a stop line in the stop line detection area.
JP 2002-190100 A

  However, in such a device, there is a possibility that a pedestrian crossing provided near the stop line may be erroneously detected as a stop line. That is, the white line which comprises a pedestrian crossing may be provided in parallel with a stop line. In this case, it is difficult to distinguish the white line and the stop line constituting the pedestrian crossing, and the white line constituting the pedestrian crossing may be erroneously detected as the stop line.

  Therefore, an object of the present invention is to provide a stop line detection device that can detect stop lines with high accuracy while suppressing erroneous detection of stop lines.

  That is, the stop line detection device according to the present invention is a stop line detection device that detects a stop line drawn on the road based on an image of the road, and a pedestrian crossing detection unit that detects a pedestrian crossing of the road. A detection area setting means for setting a stop line detection area where the stop line should be detected so as not to include the pedestrian crossing, and a stop line detection means for detecting a stop line in the stop line detection area Has been.

  According to the present invention, a stop line detection area is set so as not to include a pedestrian crossing, and a stop line is detected in the stop line detection area, thereby erroneously detecting a line constituting the pedestrian crossing as a stop line. Can be prevented, and the detection accuracy of the stop line can be improved.

  In the stop line detection device according to the present invention, it is preferable that the detection area setting means sets the stop line detection area on the near side of the pedestrian crossing. According to the present invention, by setting the stop line detection area on the near side of the pedestrian crossing and limiting the detection area, it is possible to improve the detection accuracy and reduce the load of the detection process.

  In the stop line detection device according to the present invention, it is preferable that the stop line detection means detects the stop line based on continuity of horizontal edges and the thickness thereof. According to the present invention, by detecting the stop line based on the continuity of the horizontal edge and its thickness, it is possible to reduce erroneous recognition of the stop line and appropriately detect the stop line.

  Furthermore, in the stop line detection apparatus according to the present invention, it is preferable that the pedestrian crossing detection means detects the pedestrian crossing based on a frequency histogram at a vertical position of a vertical edge. According to the present invention, it is possible to accurately and easily detect a pedestrian crossing that includes many vertical edges.

  According to the present invention, it is possible to prevent erroneous detection of a pedestrian crossing as a stop line, and to detect a stop line with high accuracy.

  Hereinafter, embodiments of the present invention will be described in detail with reference to the accompanying drawings. In the description of the drawings, the same elements are denoted by the same reference numerals, and redundant description is omitted.

  FIG. 1 is a schematic configuration diagram of a stop line detection apparatus according to an embodiment of the present invention.

  As shown in FIG. 1, the stop line detection device 1 according to the present embodiment is a device that detects a stop line drawn on a road based on an image obtained by photographing the road, and is installed in a vehicle, for example. The stop line is a line provided to indicate the stop position of the vehicle. This stop line detection apparatus 1 is applied to a driving support system, and functions as a stop line detection unit of the driving support system. The driving support system is a system that supports driving of the vehicle, for example, detects a stop line of a road, calculates a distance from the vehicle to the stop line, and appropriately notifies the driver so that the vehicle stops at the stop line. Notification processing and vehicle braking control are performed.

  The stop line detection device 1 is provided with an imaging unit 2. The imaging unit 2 functions as an imaging unit that images the surroundings of the vehicle, and is installed, for example, so as to capture the front of the vehicle. For example, a CCD camera or a C-MOS camera is used as the imaging unit 2.

  The imaging unit 2 is connected to an ECU (Electronic Control Unit) 3. The ECU 3 controls the entire apparatus, and includes, for example, a CPU, a ROM, a RAM, an input signal circuit, an output signal circuit, a power circuit, and the like. The ECU 3 has an image processing function and a stop line detection function for inputting a captured image from the imaging unit 2 and performing image processing to detect a stop line drawn on the road. As the image processing unit and the stop line detection unit, Function.

  A navigation system 4 is connected to the ECU 3. The navigation system 4 inputs information related to the stop line to the ECU 3. Thereby, the stop line detection apparatus 1 can acquire the position information with the possibility of displaying the stop line such as intersection information, red signal information, and stop sign information through the navigation system 4, for example.

  A notification unit 5 is connected to the ECU 3. The notification unit 5 functions as a notification unit that notifies the driver of the vehicle when there is a possibility that the vehicle cannot stop at the position of the stop line. As this alerting | reporting part 5, what alert | reports through visuals, such as a lamp | ramp, a liquid crystal, LED, or what alert | reports through hearing, such as a buzzer and a speaker, etc. are used, for example.

  A brake unit 6 is connected to the ECU 3. The braking unit 6 performs braking control of the vehicle, and corresponds to, for example, a brake ECU. When the braking command signal is output from the ECU 3 to the braking unit 6, the vehicle is braked.

  Next, the operation of the stop line detection device 1 according to this embodiment will be described.

  FIG. 2 is a flowchart of the driving support process in the driving support system to which the stop line detection device 1 is applied. The series of control processes in FIG. 2 are repeatedly executed by the ECU 3 at a predetermined cycle, for example.

  First, as shown in S10 of FIG. 2, distance information acquisition processing is performed. This distance information acquisition process is a process of reading information on the distance D to a point where there is a stop line such as a stop sign or a red signal. For example, the distance D is read based on an information signal from the navigation system 4 or the like. Is called. Then, the process proceeds to S12, and it is determined whether or not there is a preceding vehicle within the distance D. For example, the position information of the preceding vehicle is acquired based on vehicle information by external communication or sensor information such as a radar sensor, and it is determined whether or not the preceding vehicle exists within the distance D based on the position information of the preceding vehicle. Is done.

  If it is determined in S12 that there is a preceding vehicle within the distance D, braking control is performed so as to stop the vehicle in response to the stop of the preceding vehicle (S14). On the other hand, if it is determined in S12 that there is no preceding vehicle within the distance D, stop line recognition processing is performed (S16). The stop line recognition process is a process of detecting a stop line drawn on the road and measuring the distance to the stop line. Details of the stop line recognition process will be described later.

  And it transfers to S18 and an arrival time calculation process is performed. The arrival time calculation process is a process of calculating the arrival time T to the stop line of the vehicle. For example, the time T until the vehicle reaches the stop line is calculated based on the vehicle speed, the deceleration, and the like. Then, the process proceeds to S20, and it is determined whether or not the vehicle can be stopped before the stop line. Whether or not the vehicle can be stopped before the stop line is determined based on the distance D, the arrival time T, the braking performance of the vehicle, and the like. If it is determined in S20 that the vehicle can be stopped before the stop line, the driving support process is terminated.

  On the other hand, if it is determined in S20 that the vehicle cannot be stopped before the stop line, a warning process is performed (S22). The warning process is a process for warning the driver of the vehicle through the notification unit 5 that the vehicle will not stop before the stop line. And it transfers to S24 and a braking process is performed. The braking process is a process for forcibly applying a braking force to the vehicle through the braking unit 6. By this braking process, the vehicle can be stopped before the stop line. And a series of control processing of driving support processing is ended.

  According to such a driving support process, the stop line can be recognized and the vehicle can be stopped before the stop line. For this reason, it is possible to compensate for a lack of braking force or a driving error of the vehicle driver, and to improve safety during vehicle braking.

  FIG. 3 is a flowchart of stop line recognition processing in the travel support system to which the stop line detection device 1 is applied.

  This stop line recognition process is a process executed in S16 of FIG. As shown in FIG. 3, in the stop line recognition process, a stop line detection process is performed (S30), and a stop line ranging process is performed (S32). The stop line detection process is a process for detecting a stop line based on the image captured by the imaging unit 2. Details of this processing will be described later. The stop line ranging process is a process for measuring the distance from the vehicle to the stop line. For example, the distance from the vehicle to the stop line is calculated based on the display position of the stop line on the captured image and the mounting angle of the camera that is the imaging unit 2.

  According to the stop line recognition process, a stop line drawn on the road can be detected based on the image captured by the imaging unit 2. Then, the distance from the vehicle to the stop line can be calculated.

  4 and 5 are flowcharts of the stop line detection process executed by the stop line detection device 1. This stop line detection process is a process executed in S30 of FIG. A series of control processing in the stop line detection processing is repeatedly executed by the ECU 3 at a predetermined cycle, for example.

  First, as shown in S40 of FIG. 4, a captured image reading process is performed. The read process of the captured image is performed by taking the image captured by the imaging unit 2 into the ECU 3. Then, the process proceeds to S42, and white line detection processing is performed. The white line detection process is a process for detecting a white line drawn on a road by performing image processing on a captured image. For example, a white line candidate is extracted by performing edge processing on the image, and it is determined whether or not a white line exists based on the extending direction or position of the line segment of the white line candidate.

  Then, the process proceeds to S44, and it is determined whether or not a white line is drawn on the road. When it is determined that no white line is drawn on the road, an initial detection area for detecting a stop line is set as a square area (rectangular area) (S46). For example, as shown in FIG. 6, a rectangular initial detection region 21 is set in the image 11. At this time, the initial detection area 21 is set at a center position in the left-right direction of the image 11 and at a position lower than the center in the up-down direction. The initial detection area 21 is set to a size that allows the stop line 13 of the road to sufficiently enter. Further, the initial detection region 21 is preferably set according to the posture state such as the curve radius of the road and the roll of the vehicle. In this case, the initial detection area 21 can be appropriately set according to the road condition and the vehicle posture.

  On the other hand, when it is determined in S44 of FIG. 4 that a white line is drawn on the road, the initial detection area for detecting the stop line is set as a trapezoidal area (S48). For example, as shown in FIG. 6, a trapezoidal initial detection region 22 is set in the image 11. At that time, the initial detection area 22 is set at a center position in the left-right direction of the image 11 and at a position lower than the center in the up-down direction. The initial detection area 22 is set so that the trapezoidal short side is on the upper side, and the left and right oblique sides are set parallel to the white lines 12 and 12. In addition, the initial detection area 22 is set to a size that allows the stop line 13 of the road to sufficiently enter. Further, the initial detection area 22 is preferably set according to the posture state such as the curve radius of the road and the roll of the vehicle. In this case, the initial detection area 22 can be appropriately set according to the road condition and the vehicle posture.

  Then, the process proceeds to S50 in FIG. 4 to perform horizontal edge extraction processing. The horizontal edge extraction process is a process for extracting a horizontal edge in the initial detection area. For example, when the change in brightness in the vertical direction (up and down direction of the image) is larger than a predetermined threshold in the image area of the initial detection area, the location is extracted as a horizontal edge point. Then, the process proceeds to S52, and a horizontal edge histogram calculation process is performed.

  The horizontal edge histogram calculation process is a process of calculating the horizontal edge frequency distribution by accumulating the horizontal edges extracted in the initial detection area for each position in the vertical direction. For example, as shown in FIG. 6, the horizontal edge points extracted in the initial detection area 22 are accumulated for each vertical position, and are calculated as the horizontal edge accumulation degree.

  Then, the process proceeds to S54, and a detection area vertical width setting process is performed. This vertical width setting process is a process for setting the vertical width of the detection area within a range including a road stop line. For example, in FIG. 6, the vertical width w1 is set with a predetermined width in the vertical direction from the vertical position centering on the vertical position where the horizontal edge accumulation level becomes the peak value.

  Then, the process proceeds to S56, and vertical edge extraction processing is performed. The vertical edge extraction process is a process for extracting a vertical edge in a detection region having a vertical width w1. For example, when the change in brightness in the horizontal direction (the left-right direction of the image) is larger than a predetermined threshold in the image area of the detection area having the vertical width w1, that place is extracted as a vertical edge point. Then, the process proceeds to S58, and a vertical edge histogram calculation process is performed.

  The vertical edge histogram calculation process is a process of calculating the vertical edge frequency distribution by accumulating the vertical edges extracted in the vertical edge extraction process for each position in the vertical direction. For example, as shown in FIG. 6, the extracted vertical edge points are accumulated for each vertical position, and are calculated as the vertical edge accumulation degree.

  Then, the process proceeds to S60 in FIG. 5 and it is determined whether or not the peak value of the histogram of the vertical edge in the detection region is equal to or greater than the threshold value A. The threshold A is a set value set in advance in the ECU 3. In S60, if the peak value of the histogram of the vertical edge is not equal to or greater than the threshold value A, it is determined that there is no pedestrian crossing, and no pedestrian crossing flag is set. (S62). Then, the process proceeds to S64, and the range of the detection region is maintained as it is without being changed. Then, the process proceeds to S70.

  On the other hand, if the peak value of the vertical edge histogram is greater than or equal to the threshold value A in S60, it is determined that there is a pedestrian crossing, and a pedestrian crossing flag is set. It is recorded (S66). Then, the process proceeds to S68, and detection area change processing is performed. This detection area changing process is a process of changing the range of the detection area so as not to include a crosswalk. For example, as shown in FIG. 6, in the detection region with the vertical width w1, the region w3 below the vertical range with the vertical width w2 centering on the peak value of the vertical edge is set as the detection region. By changing the setting of the detection area w3 in this way, the detection area of the stop line can be set so as not to include the white line constituting the pedestrian crossing.

  Then, the process proceeds to S70, and the process of calculating the line length of the horizontal edge is performed. This process is a process of extracting a line segment of a horizontal edge continuous in the horizontal direction in the detection region and calculating a line segment length of the horizontal edge. When there is a pedestrian crossing, the line segment length of the horizontal edge is calculated in the detection area w3. On the other hand, when there is no pedestrian crossing, the line segment length of the horizontal edge is calculated in the detection area w1. Then, the process proceeds to S72, and binarization processing is performed on the upper region of the horizontal edge line segment. This binarization process is a process of binarizing the luminance value of the image in the upper region of the horizontal edge line segment.

  Then, the process proceeds to S74, and it is determined whether or not the line length of the horizontal edge is greater than or equal to a predetermined value and the horizontal edge line segment in the binarized image has a vertical width greater than or equal to a predetermined value. The predetermined value of the horizontal edge line segment length and the vertical width of the horizontal edge line segment are set values set in advance in the ECU 3. If the line length of the horizontal edge is not greater than or equal to the predetermined value in S74, or the horizontal edge line segment does not have a vertical width greater than or equal to the predetermined value in the binarized image, the control process of the stop line detection process is terminated.

  On the other hand, if the line length of the horizontal edge is greater than or equal to a predetermined value in S74 and the horizontal edge line segment has a vertical width greater than or equal to a predetermined value in the binarized image, the horizontal edge line segment is set as a stop line. (S76). Then, a series of control processing of the stop line detection processing ends.

  As described above, according to the stop line detection device 1 according to the present embodiment, the stop line detection area is set so as not to include the pedestrian crossing, and the stop line is detected in the detection area, thereby the pedestrian crossing. Can be prevented from being erroneously detected as a stop line. For this reason, it is possible to improve the detection accuracy of the stop line.

  In addition, by setting the detection area below the pedestrian crossing on the image and setting the stop line detection area on the near side of the pedestrian crossing to limit the detection area, the detection processing load is reduced while improving the detection accuracy. Can be planned.

  Further, by detecting the stop line based on the continuity of the horizontal edge and its thickness, it is possible to reduce erroneous recognition of the stop line and to detect the stop line appropriately.

  Further, by detecting the pedestrian crossing based on the frequency histogram at the vertical position of the vertical edge, it is possible to accurately and easily detect the pedestrian crossing including many vertical edges.

  In addition, embodiment mentioned above shows an example of the stop line detection apparatus based on this invention. The stop line detection device according to the present invention is not limited to the above, and the stop line detection device according to the embodiment is modified or otherwise changed so as not to change the gist described in each claim. It may be applied.

1 is a schematic configuration diagram of a stop line detection device according to an embodiment of the present invention. It is a flowchart which shows the driving | running | working assistance process in the stop line detection apparatus of FIG. It is a flowchart which shows the stop line recognition process in the stop line detection apparatus of FIG. It is a flowchart which shows the stop line detection process in the stop line detection apparatus of FIG. It is a flowchart which shows the stop line detection process in the stop line detection apparatus of FIG. It is explanatory drawing of the captured image in the stop line detection apparatus of FIG. 1, a horizontal edge accumulation degree, and a vertical edge accumulation degree.

Explanation of symbols

  DESCRIPTION OF SYMBOLS 1 ... Stop line detection apparatus, 2 ... Imaging part, 3 ... ECU, 4 ... Navigation system, 5 ... Notification part, 6 ... Control part.

Claims (4)

In a stop line detection device that detects a stop line drawn on the road based on an image of the road,
A pedestrian crossing detection means for detecting a pedestrian crossing of the road;
Detection area setting means for setting a stop line detection area for detecting the stop line so as not to include the pedestrian crossing;
Stop line detection means for detecting a stop line in the stop line detection region;
A stop line detection device.   The stop line detection device according to claim 1, wherein the detection area setting unit sets the stop line detection area on the near side of the pedestrian crossing.   The stop line detection device according to claim 1, wherein the stop line detection unit detects the stop line based on continuity of horizontal edges and the thickness thereof. The stop line detection device according to claim 1, wherein the pedestrian crossing detection unit detects the pedestrian crossing based on a frequency histogram at a vertical position of a vertical edge.

Images