JPH08120618A - Road marking detecting and repairing car - Google Patents

Abstract

PURPOSE: To enhance an execution accuracy by providing a television camera, image memory, marking detecting means, start and end point detecting means, drive mechanism, and movement detecting sensor. CONSTITUTION: An image picked up by a first television camera 3a by illuminating a road by means of a lighting is stored in a first image memory 8a and the position of a road marking 11 is detected at a first marking detecting means 9a, and start and end points of a marking 11, in a broken line, are detected at a start and end point detecting sensor 10. Next, the amount of movement of a repairing car is detected by a car movement detecting sensor so that the marking 11 is caused to come to the center of a field of vision of a second television camera 3b by means of a drive mechanism 5 when the second camera 3b has reached a position of the marking 11 estimated based on an image of the camera 3a. Thereafter an image which is picked up by the camera 3b is stored in a second image memory 8b to detect a precise position of the marking 11 by means of a second marking detecting means 9b and the drive mechanism 5 is finely adjusted to position a paint nozzle 7 to the existing road marking.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a road display detecting and repairing vehicle.

[0002]

2. Description of the Related Art On a paved road, the traffic division of vehicles is indicated and a marking such as a center line or a roadside line is provided to ensure safe traffic. However, since these road markings are worn and gradually thinned by passing vehicles, repair work must be performed regularly or irregularly to always provide good road markings. As shown in the perspective view of FIG. 19, the most commonly performed conventional repair work is one in which a plurality of workers perform repairs manually, and in this case, all of them are carried out by workers.
No road marking detection device is required. However, on a highway or the like, conventionally, as shown in the perspective view of FIG. 20, a distance sensor 15 mounted on the repair vehicle 1 measures a distance L to the wall 14 beside the side road, and the distance L and the distance L are measured. The distance L2 between the wall 14 and the paint nozzle 7 is calculated from the distance L1 between the sensor 15 and the paint nozzle 7, and the paint nozzle 7 is controlled so that the distance L2 becomes a preset value, and repair work is performed. .

Further, as shown in the perspective view of FIG. 21, the television camera 3 fixed to the repair vehicle 1, an image memory 8 for storing images taken by the television camera 3, and a road from the images stored in the image memory 8 are stored. A road marking detecting device including a marking detecting means 9 for detecting the marking 11 is also known. In this type of detection device, the television camera 3 is fixed to the repair vehicle 1, an image is captured in front of the repair vehicle 1 so as to include the road marking 11, and the captured image is stored in the image memory 8
Is stored. The sign detection means 9 reads the image stored in the image memory 8 and performs a binarization process for extracting a region having a density equal to or higher than a preset threshold value. Since the road marking 11 is usually drawn in white or yellow,
Since the density is high with respect to the black or gray of the pavement surface, only the area of the road marking 11 is extracted by the binarization processing (FIG. 3). Since the television camera 3 faces the front of the repair vehicle 1 to capture an image, the area of the road marking 11 extracted in the image is a vertically long strip-shaped area (hatched portion) on the image. Therefore, as shown in FIG. 22, the center position of each scanning line of the road marking area extracted from the image is detected, and the center positions of the scanning lines are connected to each other to detect the road marking 11. Here, as shown in FIG.
The TV camera 3 is fixedly installed on the repair vehicle 1, and the range captured by the TV camera 3 can be calibrated in advance by measuring the height of the TV camera 3 and the distance D from the TV camera 3 to the visual field. Therefore, if the position of the road marking 11 on the image can be detected, the actual position of the road marking 11 with respect to the repair vehicle 1 can also be detected, and the drive mechanism 5 is controlled based on this result to position the paint nozzle 7. .

[0004]

However, since all the manual work of the worker is performed manually, the work is performed in an environment in which a general vehicle runs just beside, which is extremely dangerous. Further, in the above repair vehicle, since the position of the road marking is estimated by measuring the distance to the wall beside the road, it cannot be applied in a situation where there is no wall beside the road.
Furthermore, even if there is a wall, the road markings are not directly detected, so it cannot be accurately aligned with the existing road markings to be repaired, resulting in misalignment and duplication of road markings. , Its quality degradation occurs. Further, in the road marking detection means by the TV camera, although the road marking is directly detected, since it is detected from the image captured in front of the vehicle, from the point where the image is captured due to slight ups and downs of the road, There is a gap between the positions of the paint nozzles that actually apply the paint, and the deterioration of quality such as duplication of road markings still occurs. Also, in this case, when only the image near the paint nozzle is captured by the TV camera, the overall continuity of the road marking cannot be detected, so it is not possible to distinguish between the blurred portion and the beginning and end of the road marking, and in repairing However, there is also a deterioration in quality such as broken lines.

The present invention has been proposed in view of the above circumstances, and it is possible to improve work accuracy and work safety, and save labor, labor, and efficiency in daytime or daytime work. An object is to provide an economical road marking detection device.

[0006]

Therefore, the invention of claim 1 is a road marking repair vehicle for automatically repairing markings such as a center line and a roadside line drawn on a road.
A first TV camera fixed to the repair vehicle for imaging the road surface from immediately below to the front, a second TV camera movable with the paint nozzle and imaging just before the paint nozzle, and the front of each TV camera The first and second yellow optical bandpass filters respectively attached to
First and second image memories for respectively storing images captured by the respective television cameras, first display detecting means for detecting road display from the images stored in the first image memory, and first A start / end detecting means for detecting the start / end of the road display from the image stored in the image memory, a second display detecting means for detecting the road display from the image stored in the second image memory, A drive mechanism for moving the second television camera and the paint nozzle attached thereto in the vehicle width direction based on the outputs of the first and second display detection means, and a movement amount detection for detecting the movement amount of the repair vehicle. A second television camera based on the information of the start and end of the road display and the continuity from the image captured by the first television camera and the information of the movement amount detection sensor. Positioned and same second TV And detecting the precise position of a road displayed from image captured by the camera.

According to a second aspect of the present invention, in a road marking repair vehicle that automatically repairs markings such as a center line and a roadside line drawn on a road, a television camera movable along with a paint nozzle and a television camera A fisheye lens mounted on the front surface, a yellow optical bandpass filter mounted on the front surface of the fisheye lens, a side light-shielding plate mounted in front of the yellow optical bandpass filter, and an image stored by the TV camera. Memory, sign detection means for detecting road markings from images stored in the image memory, start / end detection means for detecting start / end of the markings from images stored in the image memory, and output of the sign detection means Based on the above, the TV camera and a drive mechanism for driving the paint nozzle are provided, and a road marker is displayed from an image captured by the TV camera. And detecting the precise position of the road sign starting and end and while detecting the continuity of.

According to a third aspect of the present invention, in a road marking repair vehicle for automatically repairing markings such as a center line and a roadside line drawn on a road, a television camera movable along with a paint nozzle, and a television camera Yellow optical bandpass filter mounted on the front side, mirror mounted in front of the yellow optical bandpass filter, scanner for driving the mirror around the horizontal axis, and image that stores the image captured by the TV camera A memory, a sign detection means for detecting road markings from the images stored in the image memory, a start / end detection means for detecting the start / end of the markings from the images stored in the image memory, and an output of the sign detection means. Based on a drive mechanism that drives the television camera and the paint nozzle based on the movement amount, and a movement amount detection sensor that detects the movement amount of the repair vehicle, And detecting the precise position of the road sign while detecting the continuity of a road marking from the image captured by Rebikamera.

According to a fourth aspect of the present invention, in a road marking repair vehicle that automatically repairs markings such as a center line and a roadside line drawn on a road, a one-dimensional line camera that can move together with a paint nozzle, The yellow optical bandpass filter mounted on the front surface of the 3D line camera, the mirror mounted on the front surface of the yellow optical bandpass filter, the scanner for driving the mirror around the horizontal axis, and the 1D line camera An image memory for accumulating and storing images to be picked up, a sign detecting means for detecting road markings from the images stored in the image memory, and a starting and ending point for detecting starting and ending points of the markings from the images stored in the image memory And a drive mechanism for driving the same one-dimensional line camera and the paint nozzle based on the output of the marking detection means.
It is characterized in that the precise position of the road marking is detected while detecting the continuity of the road marking from the image taken by the same one-dimensional line camera.

The invention of claim 5 is the invention of claims 1, 2, 3, 4
In the above, each is equipped with a lighting device for illuminating the road.

[0011]

With this structure, referring to FIG. 1, the illumination irradiates the road to stably capture an image even at night or in the shade, and in some road markings, a reflective agent is mixed and applied. It is also used to improve the signal level difference between the road marking area and other areas due to The road illuminated by the illumination is first imaged by the first television camera through the first yellow optical band pulse filter. Since the road marking is usually drawn in white or yellow, the first yellow optical band pulse filter improves the signal level difference between the road marking area and other areas. The image captured by the first TV camera is the first
The first sign detecting means detects the position of the road marking, and the start / end detecting means detects the start / end of the broken road marking and the like. Here, since the first TV camera is fixedly installed on the repair vehicle and the fixed position can be calibrated in advance, the first TV camera can be calibrated based on the position of the road marking detected on the image captured by the first TV camera. The position of the road marking within the field of view of the two TV cameras can be estimated. Therefore, the movement amount detection sensor detects the movement amount of the repair vehicle, and
When the second TV camera reaches the position of the road marking estimated from the image of the TV camera of, the driving mechanism drives the second TV camera, and the road marking is displayed in the center of the field of view of the second TV camera. I will come. Secondly, the television camera captures an image from directly above the road marking through the second yellow optical band pulse filter as with the first television camera. The image captured by the second television camera is stored in the second image memory, and the precise position of the road marking in the image is detected by the second marking detecting means. Here, since the second television camera is installed in the vicinity of the paint nozzle and is movable with the paint nozzle, the drive device is finely adjusted based on the position of the road marking detected by the second marking detection means. , Align the paint nozzle with the existing road marking.

[0012]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS Referring to the drawings, a first embodiment of an automobile road marking repair vehicle of the present invention will be described. FIG. 1 is an overall perspective view and a flow chart of main parts of the vehicle, and FIG. FIG. 3 is an explanatory view showing the procedure for extracting a road marking area from an image in FIGS. 1 and 2, FIG. 4 is a side view and a plan view of FIG. 1, and FIG. 5 is a road marking from an image in FIGS. It is explanatory drawing which shows the point which detects the start / end of.

First, in FIGS. 1 and 2, the first television camera 3a and the first illumination 4a are fixedly installed beside the repair vehicle 1 to pick up an image slightly in front of the repair vehicle 1. The optical axis of the first illumination 4a is adjusted so as to illuminate the field of view of the first television camera 3a. The first yellow optical band pulse filter 2a is mounted on the front surface of the first television camera 3a, and the image of the yellow component in the visual field is captured brightly. The output of the first TV camera 3a is the first
The image that is input to and output from the image memory 8a of
The output of the first image memory 8a is the first sign detecting means 9a.
And the start / end detection means 10. Further, the movement amount detection sensor 6 is the axle 1 of the repair vehicle 1.
3 is installed, and the amount of movement of the repair vehicle 1 is detected from the number of rotations of the wheels 12. The output of the first marking detection means 9a and the output of the movement amount detection sensor 6 are connected to the drive mechanism 5, and the second television camera 3b, the second illumination 4b, and the paint nozzle 7 are driven by the output signals of them. . A second yellow optical band pulse filter 2b is mounted on the front surface of the second television camera 3b, similarly to the first television camera 3a, and the paint nozzle 7 in the field of view irradiated by the second illumination 4b. The immediately preceding road marking 11 is imaged from directly above. The output of the second television camera 3b is connected to the second image memory 8b, the output of the second image memory 8b is connected to the second sign detecting means 9b, and the drive mechanism 5 is finely adjusted based on the output. . At this time, whether the output of the start / end detection means 10 and the output of the movement amount detection sensor 6 are input to the second sign detection means 9b, and whether or not the sign detection processing is executed in the second sign detection means 9b. Is judged.

Next, the functions of the respective parts will be described in detail.
The illuminations 4a and 4b illuminate the fields of view of the corresponding first and second television cameras 3a and 3b, respectively, thereby making it possible to stably capture images even at night or in the shade. Since the road marking 11 contains a reflecting agent, the brightness of that portion is increased by the illumination, thereby improving the signal level difference between the area of the road marking 11 and the other area of the captured image, Make it easy to extract road marking areas. Road markings 11 are usually white or yellow,
When this image is taken by a monochrome television camera, the white road marking 11 has a sufficient reflectance for all wavelengths, so there is a sufficient signal level difference from the black or gray pavement surface, Only the road marking area can be stably extracted. However, in the case of the yellow road marking 11, the reflectance other than yellow is low, so if the pavement surface has a certain reflectance in a wide wavelength range such as gray or concrete pavement, there is no difference in the total brightness level, and The level difference becomes small, and the stable road marking area cannot be extracted. Therefore, the first and second yellow optical bandpass filters 2a and 2b which transmit only the wavelengths near the wavelength at which the yellow road marking 11 shows maximum reflection are provided on the front surfaces of the first and second television cameras 3a and 3b, respectively. By installing,
The signal level difference between the yellow road marking 11 and the pavement surface can be improved, and a stable road marking area can be extracted. The first TV camera 3a is provided in the repair vehicle 11
It is fixedly installed beside the vehicle and captures an image from just above the road marking 11 to be detected, and slightly downward, and outputs the first image.
Image memory 8a.

An image as shown in FIG. 3 is stored in the first image memory 8a, and the first marking detecting means 9a performs a binarization process for extracting a region having a density larger than a preset density. The area of the road marking 11 is extracted. The extracted road marking area is stored in the first image memory 8a.
The center position of each scanning line is calculated, and this is used for road marking 1
1 is the position on the image. First television camera 3a
4 is fixedly installed on the repair vehicle 1 as shown in FIG. 4, and the height of the first TV camera 3a, the distance D from the first TV camera 3a to the field of view thereof, and the first TV camera 3a. If the distance D1 or the like of the second television camera 3b is measured, the position on the road can be estimated by detecting the position on the image. In addition, the first image memory 8a
The image stored in is also input to the start / end detection means 10 at the same time. The start / end detection means 10 binarizes the image and detects the presence or absence of the area of the road marking 11 for each scanning line. At that time, as shown in FIG. 5, a certain number N or more of scanning lines in which the area of the road marking 11 does not exist continuously from the lower direction of the image,
After that, when the area of the road marking 11 is detected by a predetermined number N of scanning lines or more, the scanning line position at which the road area starts is set as the starting end 16, and conversely, the area of the road marking 11 is a fixed number N or more scanning lines. When the scanning lines having no road marking area continue for a fixed number N or more after they have continuously existed, the scanning line position where the road marking 11 disappears is detected as the end 17.

1 and 2, the road marking 11 detected by the first television camera 3a when the repair vehicle 1 has moved.
The arrival at the position of is detected by the movement amount detection sensor 6. The movement amount detection sensor 6 is attached to the axle 13 of the repair vehicle 1 and detects the rotation speed of the axle 13.
Since the diameter of the wheel 12 is known, the movement amount of the repair vehicle 1 can be detected when the rotation speed is detected. When the repair vehicle 1 reaches the position of the road marking 11 detected by the first TV camera 3a, the center of the second TV camera 3b is located at the center position of the road marking 11 output by the first sign detecting means 9a. Drive mechanism 5 so that the second television camera 3
Move b. In this way, when the positioning of the second television camera 3b is completed, an image is captured and the content is stored in the second image memory 8b. The contents of the second image memory 8b are processed by the second marking detection means 9b to detect the road marking 11. However, at this time, by the start / end detection means 10, when it is not between the start end 16 and the end 17 shown in FIG. 5, that is, the road marking 1
If 1 does not exist, the process is not executed.

The second sign detecting means 9b binarizes the input image and extracts the area of the road marking 11 as in the case of the first sign detecting means 9a. At this time, road marking 11
Even if the road marking 11 is not detected on some scanning lines due to blurring of the image, the continuity is guaranteed by the preceding start / end detecting means 10, so that the information of the road marking areas of the preceding and following scanning lines is used as a basis. To complement. When the position of the road marking 11 is detected from the image captured by the second television camera 3b, the drive mechanism 5 is controlled so that the center position of the road marking 11 is located in the center of the image, and the second television camera 3b is controlled. Fine tune. Since the paint nozzle 7 is driven together with the second television camera 3b, the positioning of the paint nozzle 7 is completed when the fine adjustment of the second television camera 3b by the drive mechanism 5 is completed.

According to such a first embodiment, the following effects can be obtained. (1) Based on the image captured by the first TV camera, the continuity of the road marking and the rough position are detected, and the second
Since the precise position of the road marking is detected based on the image taken by the TV camera of No. 1, the accurate positioning of the paint nozzle can be performed automatically, which saves labor and does not require manual work. Is less dangerous to (2) Since the road marking is directly detected, it can be applied even if there is no structure such as a wall beside the road, and its position detection accuracy is high, so that the road marking is duplicated and the quality is deteriorated. Don't come (3) In order to image the front with the first TV camera,
In addition to being able to detect the overall continuity of road markings,
Since the TV camera is installed in the vicinity of the paint nozzle and can be driven together with the paint nozzle, it can be repaired without being affected by slight ups and downs on the road, and the industrial application effect is extremely large.

A second embodiment of the present invention will be described with reference to the drawings. FIG. 6 is an overall perspective view and a flow chart of its main equipment, and FIG. 7 is an explanatory view showing the relationship between the visual field of the fisheye lens of FIG. 6 and a light shielding plate. FIG. 8 is an explanatory view showing a procedure for extracting a road marking area from an image in the apparatus of FIG. 6, and FIG. 9 is an enlarged view of FIG. 8 (C).

First, in FIG. 6, the same reference numerals as those in FIGS. 1 to 5 denote the same devices as those in FIGS. 1 to 5, respectively. The television camera 3 is installed in the repair vehicle 1, and the paint nozzle 7 is installed by the drive mechanism 5. Can be driven with. The television camera 3 has its optical axis adjusted in the direction of the paint nozzle 7.
The image is captured so that the vicinity is the center. A fisheye lens 6, a yellow optical bandpass filter 2 and a light shielding plate 4 are mounted on the front surface of the television camera 3 so that an image of a yellow component of only a road portion within the field of view is captured brightly. The output of the television camera 3 is connected to the image memory 8 and stores the output image. The output of the image memory 8 is connected to the sign detecting means 9 and the start / end detecting means 10. The output of the marking detection means 9 is connected to the drive mechanism 5, and the output signal thereof drives the television camera 3 and the paint nozzle 7. The output of the start / end detection means 10 is connected to the paint nozzle 7 and controls the outflow amount of the paint.

The operation of each device and member will be described in detail below. The road marking 11 is usually white or yellow. When this image is picked up by a monochrome television camera, the white road marking 11 shows all wavelengths. Since it has a sufficient reflectance, there is a sufficient signal level difference with the black or gray color of the pavement surface, and only the road marking area can be stably extracted.
However, in the case of the yellow road marking 11, since the reflectance other than yellow is low, if the pavement surface has a certain reflectance in a wide wavelength range such as gray or concrete pavement, there is no difference in the total brightness level, The signal level difference becomes small, and stable road marking areas cannot be extracted.
Therefore, in the present embodiment, the yellow optical bandpass filter 2 that transmits only the wavelengths near the wavelength at which the yellow road marking 11 shows maximum reflection is mounted on the front surface of the television camera 3 so that the yellow road marking 11 becomes The signal level difference on the pavement surface can be improved, and stable road marking areas can be extracted.

The fish-eye lens 6 has a hemispherical field of view from the front surface thereof, and as shown in FIG. 7, it is possible to capture an image of a wide range from the vicinity of the paint nozzle 7 to the front of the repair vehicle 1. At this time, since it has a hemispherical field of view, naturally, an area other than the road such as the vehicle body of the repair vehicle 1 or the situation of the opposite lane of the repair vehicle 1 is also imaged. Disturbance elements such as road markings and lighting of the opposite lane often exist in areas other than these roads, which lowers the detection accuracy of the target road marking 11. Therefore, in this embodiment, the light shielding plate 4 is attached to shield the portion other than the road region from light,
It facilitates detection of the target road marking 11. An image as shown in FIG. 8 is stored in the image memory 8, and the marking detection means 9 extracts the area of the road marking 11 by the binarization process of extracting the area having a density larger than the preset density. . The center position of the extracted road marking area is calculated for each scanning line of the first image memory 8, and this is used as the position of the road marking 11 on the image. At this time, since the image is captured using the fisheye lens 6, the road marking 1
1 can be detected over a wide range from the vicinity of the paint nozzle 7 to the front of the repair vehicle 1. The images stored in the image memory 8 are also input to the start / end detection means 10 at the same time. The start / end detection means 10 binarizes the image and detects the presence or absence of the area of the road marking 11 for each scanning line.

Here, as shown in FIG. 9, a certain number N or more of scanning lines in which the area of the road marking 11 does not exist continue from the lower direction of the image, and thereafter, the area of the road marking 11 has a certain number N.
When more than the scanning lines of the above are detected, the scanning line position at which the road area starts is set as the start end 16, and conversely, after the area of the road marking 11 is continuously present for a certain number N of scanning lines or more, the road marking area When there are a certain number N or more continuous scanning lines, the scanning line position where the road marking 11 disappears is detected as the end 17. By changing the constant N according to the position of the scanning line to be processed, even if the fisheye lens 6 is distorted, it can be detected with a length corresponding to the actual distance. Movable mechanism 5
Is the television camera 3 and the paint nozzle 7 so that the road marking position detected by the marking detection means 9 is located at the center of the image.
Drive. The paint nozzle 7 and the television camera 3 are adjusted so that their central axes coincide with the traveling direction.
Therefore, the paint nozzle 7 can be positioned by adjusting the position of the television camera 3 by the movable mechanism 5. In the paint nozzle 7, the paint is made to flow out only when the output of the start / end detection means 10 is between the start point 16 and the end point 17, that is, when the paint nozzle 7 is on the road marking 11, and otherwise the paint is made to flow out. To stop drawing the road marking 11.

According to the second embodiment, the road is imaged by the television camera through the fisheye lens and the yellow optical bandpass filter. Since road markings are usually drawn in white or yellow, the yellow optical bandpass filter improves the signal level difference between road marking areas and other areas. The light-shielding plate installed in front of the yellow optical bandpass filter is adjusted so that a portion other than the road in the image to be captured is not captured, and acts to remove the region other than the road from the processing target. The image picked up by the television camera is stored in the image memory, and in the image, the sign detection means detects the overall continuity of the road sign and the position near the paint nozzle.
At the same time, the start / end detection means detects the start / end of the broken road marking. Since the television camera is arranged in the vicinity of the paint nozzle and is movable together with the paint nozzle, the position and continuity of the road marking detected by the marking detection means and the start / end information detected by the start / end detection means are displayed. Based on this, the drive is adjusted to align the paint nozzle with the existing road marking.

According to such a second embodiment, the following effects can be obtained. (1) Since the continuity of the road marking and the precise position of the road marking can be detected at the same time based on the image taken by the TV camera, accurate positioning of the paint nozzle can be performed automatically, which saves labor. Since it is designed and does not require manual work, there is little danger. (2) Since the road marking is directly detected, it can be applied even if there is no structure such as a wall beside the road, and its position detection accuracy is high. Don't come (3) Since the TV camera is installed near the paint nozzle and can be driven together with the paint nozzle, it can be repaired without being affected by the slight undulations or steps on the road. Extremely large.

Referring to the drawings, a third embodiment of the present invention will be described. FIG. 10 is an overall perspective view and a flowchart of its main equipment, and FIG. 11 is an enlarged view showing the main parts of FIG.
12 is a side view and a plan view of FIG. 10, and FIG. 13 is FIGS.
FIG. 12 is an explanatory diagram showing the procedure for extracting the road marking area from the image, and FIG. 14 is an image diagram showing a modified example of the image of FIG.

In the above drawing, the same reference numerals as those in FIGS. 1 to 9 indicate the same devices as those in the same drawings, and the third embodiment is the second one.
Similar to the embodiment, the television camera 3 is installed in the repair vehicle 1 and can be driven together with the paint nozzle 7 by the drive mechanism 5. Here, in this embodiment, on the front surface of the television camera 3,
The yellow optical bandpass filter 2 and the mirror 4 are attached, and the image of the yellow component of the road is captured brightly. The mirror 4 is connected to the scanner 18, and as shown in FIG. 12, a first imaging position 19 a in which the imaging field of view of the TV camera 3 is in front of the repair vehicle 1 and a second imaging position 19 a in the vicinity of the paint nozzle 7. The imaging position 19b can be changed alternately. The output of the television camera 3 is connected to the image memory 8 and stores the output image. The output of the image memory 8 is connected to the sign detecting means 9 and the start / end detecting means 10.

The output of the marking detection means 9 is connected to the drive mechanism 5, and the output signal from the drive mechanism 5 drives the television camera 3 and the paint nozzle 7. The output of the start / end detecting means 10 is connected to the paint nozzle 7, controls the outflow amount of the paint and is also input to the sign detecting means 9, and based on the continuity of the road marking 11 at the second imaging position 19b. It is determined whether or not the road marking 11 is detected. The output of the moving amount detection sensor 6 is input to the sign detection means 9, and the television camera 3
Determines whether the start and end positions obtained by imaging the front of the repair vehicle 1 are within the field of view when the television camera 3 images the vicinity of the paint nozzle 7, and detects the road marking 11. It is determined whether or not.

The operation of each device and member will be described in detail below. The road marking 11 is usually white or yellow. When this image is taken by a monochrome television camera, the white road marking 11 shows all wavelengths. Since it has a sufficient reflectance, there is a sufficient signal level difference with the black or gray color of the pavement surface, and only the road marking area can be stably extracted.
However, in the case of the yellow road marking 11, since the reflectances other than yellow are low, if the pavement surface has a certain reflectance in a wide wavelength range such as gray or concrete pavement, there is no difference in the total brightness level, The signal level difference becomes small, and stable road marking areas cannot be extracted.
Therefore, in the present embodiment, the yellow optical bandpass filter 2 that transmits only the wavelengths near the wavelength at which the yellow road marking 11 shows maximum reflection is mounted on the front surface of the television camera 3 so that the yellow road marking 11 becomes The signal level difference on the pavement surface can be improved, and stable road marking areas can be extracted.

First, the angle of the mirror 4 is adjusted by the scanner 18 so that the visual field of the television camera 3 is at the first image pickup position 19a. In this state, the television camera 3 captures an image via the yellow optical bandpass filter 2. An image as shown in FIG. 13 is stored in the image memory 8, and the marking detection means 9 extracts the area of the road marking 11 by a binarization process of extracting an area having a density larger than a preset density. . The center position of the extracted road marking area is obtained for each scanning line of the image memory 8, and this is set as the position of the road marking 11 on the image. The images stored in the image memory 8 are also input to the start / end detection means 10 at the same time. The start / end detection means 10 binarizes the image and detects the presence or absence of the area of the road marking 11 for each scanning line.

As shown in FIG. 14, a certain number N or more of scanning lines in which the area of the road marking 11 does not exist are continuously detected from the lower direction of the image, and thereafter, the area of the road marking 11 is detected by a certain number N or more scanning lines. Then, the scanning line position at which the road area starts is set as the start end 16, and conversely, after the area of the road marking 11 is continuously present for a certain number N of scanning lines or more, a scanning line having no road marking area is detected. When the number of roads continues for a certain number N or more, the scanning line position where the road marking 11 disappears is detected as the end 17. Next, the mirror 18 is driven by the scanner 18 so that the field of view of the television camera 3 is adjusted to the second imaging position 19b. The movement amount detection sensor 6 detects that the repair vehicle 1 has moved and that the television camera 3 has reached the position imaged at the first imaging position 19a.

The movement amount detection sensor 6 is attached to the axle 13 of the repair vehicle 1 and detects the rotation speed of the axle 13.
Since the diameter of the wheel 12 is known, the movement amount of the repair vehicle 1 can be detected when the rotation speed is detected. When the repair vehicle 1 reaches the position of the road marking 11 detected at the first imaging position 19a, the center of the television camera 3 is positioned at the center position of the road marking 11 output by the sign detecting means 9 at that time. Then, the TV camera 3 is moved by the drive mechanism 5.
When the positioning of the television camera 3 is completed, an image is taken and the content is stored in the image memory 8. Image memory 8
The contents of (1) are processed by the marking detection means 9 similarly to the previous image pickup, and the road marking 11 is detected. However, at this time, when the start / end detection means 10 does not exist between the start end 16 and the end 17, that is, when the road marking 11 does not exist, the process is not executed.

At this time, even if the road markings 11 are not detected on some scanning lines due to the fading of the road markings 11 and the like, the continuity is guaranteed by the preceding start / end detecting means 10, so that the continuity of Complement based on the information of the road marking area of the scanning line. When the position of the road marking 11 is detected from the image captured by the TV camera 3 at the second imaging position 19b, the drive mechanism 5 is controlled so that the center position of the road marking 11 is located at the center of the image, and the TV camera 3 Fine-tune. Since the paint nozzle 7 is driven together with the television camera 3, when the fine adjustment of the television camera 3 by the drive mechanism 5 is completed, the positioning of the paint nozzle 7 is also completed. In the paint nozzle 7, when the output of the start / end detection means 10 is between the start point 16 and the end point 17, that is,
The paint is made to flow out only when the paint nozzle 7 is on the road marking 11, and at other times, the paint is stopped from flowing so that the road marking 11 is not drawn. By repeating the above operation, the road marking 11 is detected and the repair work is executed.

According to the third embodiment, the road is
It is reflected by a mirror and imaged by a television camera through a yellow optical bandpass filter. Since the road markings are usually drawn in white or yellow, the yellow optical bandpass filter improves the signal level difference between the road marking area and other areas. The angle of the mirror installed in front of the yellow optical bandpass filter is changed by the scanner, and the mirror is controlled so as to alternately capture the image in front of the repair vehicle and the image in the vicinity of the paint nozzle. The image taken by the television camera is stored in the image memory, and when the television camera takes an image of the front of the repair vehicle, the sign detection means detects the overall continuity of the road marking and the start / end detection means. The start and end of the broken road marking are detected. Further, the position near the paint nozzle is detected when the television camera images the vicinity of the paint nozzle. The movement amount detection sensor detects the movement amount of the repaired vehicle, and the correspondence between the position of the road marking detected by the TV camera capturing the front of the vehicle and the position of the road marking detected by the television camera capturing the vicinity of the paint nozzle Make a mark. The television camera is installed in the vicinity of the paint nozzle and is movable together with the paint nozzle, based on the position and continuity of the road marking detected by the marking detection means and the start / end information detected by the start / end detection means. Adjust the drive to align the paint nozzle with the existing road marking to be repaired.

According to the third embodiment, the following effects can be obtained. (1) Since the continuity of the road marking and the precise position of the road marking can be detected at the same time based on the image taken by the TV camera, accurate positioning of the paint nozzle can be performed automatically, which saves labor. There is little danger because it can be done manually. (2) Since the road marking can be directly detected, it can be applied even if there is no structure such as a wall beside the road, and its position detection accuracy is high. Don't come (3) Since the TV camera is installed near the paint nozzle and can be driven together with the paint nozzle, it can be repaired without being affected by the slight undulations or steps on the road. Extremely large.

A fourth embodiment of the present invention will be described with reference to the drawings. FIG. 15 is an overall perspective view thereof and a flowchart of its main equipment, FIG. 16 is a side view and a plan view of FIG.
FIG. 7 is an explanatory view showing the procedure for extracting the road marking area from the images of FIGS. 15 to 16, and FIG. 18 is an enlarged view showing the procedure for detecting the start and end of the road marking in the modification of FIG. 17 (C). .

In the above drawing, the same reference numerals as those in FIGS. 1 to 14 indicate the same members as those in the drawings. First, in FIGS. 15 to 16, the one-dimensional line camera 3 is installed in the repair vehicle 1 and driven. It can be driven together with the paint nozzle 7 by the mechanism 5. A yellow optical bandpass filter 2 and a mirror 4 are mounted on the front surface of the one-dimensional line camera 3, and a yellow component image of a road is captured brightly. The mirror 4 is connected to a scanner 18, and as shown in FIG. 16, the imaging field of view of the one-dimensional line camera 3 from the vicinity of the paint nozzle 7,
A wide range is scanned up to the front of the repair vehicle 1. 1
The output of the dimensional line camera 3 is connected to the image memory 8,
The output image is sequentially accumulated and stored for each scanning line according to the angle of the scanner 18. The output of the image memory 8 is connected to the sign detecting means 9 and the start / end detecting means 10. The output of the marking detection means 9 is connected to the driving mechanism 5, and the one-dimensional line camera 3 and the paint nozzle 7 are driven by the output signal thereof. Start-end detection means 1
The output of 0 is connected to the paint nozzle 7 and controls the outflow amount of the paint.

The operation of each part will be described in detail below. The road marking 11 is usually white or yellow. When this image is picked up by the monochrome one-dimensional line camera 3, the white road marking 11 shows all wavelengths. Since it has a sufficient reflectance, there is a sufficient signal level difference with the black or gray color of the pavement surface, and only the road marking area can be stably extracted. However, in the case of the yellow road marking 11, since the reflectance other than yellow is low, if the pavement surface has a certain reflectance in a wide wavelength range like gray or concrete pavement,
There is no difference in total brightness level, the signal level difference is small, and stable road marking areas cannot be extracted. Therefore, in this embodiment, the yellow optical bandpass filter 2 that transmits only the wavelengths near the wavelength at which the yellow road marking 11 shows maximum reflection is mounted on the front surface of the one-dimensional line camera 3 to make the yellow road It is possible to improve the signal level difference between the marking 11 and the pavement surface, and it becomes possible to stably extract the road marking area.

First, the angle of the mirror 4 is adjusted by the scanner 18 so that the field of view of the one-dimensional line camera 3 scans a wide range from the vicinity of the paint nozzle 7 to the front of the repair vehicle 1. At this time, the one-dimensional line camera 3 captures an image through the yellow optical bandpass filter 2,
Each one-dimensional image is accumulated and stored in the scanning line on the image memory 8 corresponding to each scanning line according to the angle of the scanner 18. As a result, FIG.
The image as shown in FIG. 7 is accumulated and stored, and the road marking 11 is subjected to the binarization processing for extracting the region having the density larger than the preset density in the road marking detection means 9.
Area is extracted. In the extracted road marking area, the center position is obtained for each scanning line of the image memory 8 and the center position is calculated.
1 is the position on the image. The images stored in the image memory 8 are also input to the start / end detection means 10 at the same time. The start / end detection means 10 binarizes the image and detects the presence or absence of the area of the road marking 11 for each scanning line.

As shown in FIG. 18, scanning lines in which the area of the road marking 11 does not exist are continuous for a certain number N or more from the lower direction of the image, and thereafter, the area of the road marking 11 is detected for a certain number N or more scanning lines. Then, the scanning line position where the road area starts is set as the start end 16, and conversely, after the area of the road marking 11 continues for a certain number N of scanning lines or more, the scanning line having no road marking area becomes constant. Road marking 1 when continuous for several N or more
The scanning line position where 1 disappears is detected as the terminal end 17.
1 at the center position of the road marking 11 output by the marking detection means 9.
The one-dimensional line camera 3 is moved by the drive mechanism 5 so that the center of the three-dimensional line camera 3 comes to the center. Paint nozzle 7
Is driven together with the one-dimensional line camera 3, the positioning of the paint nozzle 7 is also completed when the fine adjustment of the one-dimensional line camera 3 by the drive mechanism 5 is completed. In the paint nozzle 7, the paint is caused to flow out only when the output of the start / end detection means 10 is between the start point 16 and the end point 17, that is, when the paint nozzle 7 is on the road marking 11, and otherwise the paint is discharged. Stop the outflow and mark the road 11
Do not draw. By repeating the above operation, the road marking 11 can be detected and the repair work is executed.

According to the fourth embodiment, the road is reflected by the mirror and is imaged by the one-dimensional line camera through the yellow optical bandpass filter. Since road markings are usually drawn in white or yellow, the yellow optical bandpass filter improves the signal level difference between road marking areas and other areas. The angle of the mirror installed on the front surface of the yellow optical bandpass filter is sequentially changed by the scanner, and the mirror is controlled to sequentially capture images from the vicinity of the paint nozzle to the front of the repair vehicle. The images captured by the one-dimensional line camera are sequentially accumulated in the order in which they were captured and stored in the image memory, so that an image in which a wide range from the vicinity of the paint nozzle to the front of the repair vehicle is moved is generated. Based on the image accumulated and stored in the image memory, the overall continuity of the road marking and the precise position of the road marking near the paint nozzle are detected by the marking detecting means,
The start / end detection means detects the start / end of the dashed road marking. The one-dimensional line camera is installed in the vicinity of the paint nozzle and is movable together with the paint nozzle, and is based on the position and continuity of the road marking detected by the marking detection means and the start / end information detected by the start / end detection means. Adjust the drive to align the paint nozzle with the existing road marking to be repaired.

According to the fourth embodiment, the following effects can be obtained. (1) Since the continuity of road markings and the precise position of road markings can be detected simultaneously based on the images taken by the one-dimensional line camera, mirror, and scanner, accurate positioning of paint nozzles can be performed automatically. Therefore, labor saving is achieved, and there is little danger because the work is not manually performed. (2) By detecting the road marking directly, it can be applied even if there is no structure such as a wall beside the road, and its position detection accuracy is high, so there is a reduction in quality such as duplication of road marking. Don't come (3) A one-dimensional line camera is installed in the vicinity of the paint nozzle and can be driven together with the paint nozzle, so repairs can be performed without being affected by slight ups and downs on the road, and industrial use. The effect is extremely large.

[0043]

According to the invention of claim 1, in a road marking repair vehicle for automatically repairing markings of a center line, a roadside line, etc. drawn on a road, from a portion directly below the vehicle fixed to the repair vehicle. A first television camera that captures an image of the road surface over the front, a second television camera that is movable with the paint nozzle and captures an image immediately in front of the paint nozzle, and first and second devices that are respectively mounted on the front faces of the television cameras. Yellow optical band pass filter, first and second image memories for respectively storing images captured by the respective television cameras, and first for detecting road display from the images stored in the first image memory. The display detection means, the start / end detection means for detecting the start / end of the road display from the image stored in the first image memory, and the road display from the image stored in the second image memory. A second display detecting means to be known, and a drive mechanism for moving the second television camera and the paint nozzle attached thereto in the vehicle width direction based on the outputs of the first and second display detecting means, A movement amount detection sensor for detecting the movement amount of the repair vehicle, detecting the start and end and continuity of the road display from the image taken by the first TV camera, and the information and the movement amount detection sensor. By positioning the second TV camera based on the information of the above, and detecting the precise position of the road display from the image captured by the second TV camera, the construction accuracy and work safety are improved and labor saving is performed. INDUSTRIAL APPLICABILITY The present invention is extremely useful industrially because an economical road marking detection device that can perform labor-saving and highly efficient work is obtained.

According to the invention of claim 2, in a road marking repair vehicle for automatically repairing markings of a center line, a roadside line, etc. drawn on a road, a television camera movable along with a paint nozzle, and the same television. A fisheye lens mounted on the front of the camera, a yellow optical bandpass filter mounted on the front of the fisheye lens, a side light shield mounted in front of the yellow optical bandpass filter, and an image captured by the TV camera are stored. Image memory, sign detection means for detecting road markings from the images stored in the image memory, start / end detection means for detecting the start / end of the markings from the images stored in the image memory, and sign detection means Based on the output of the same TV camera and a drive mechanism that drives the same paint nozzle, By detecting the precise position of the road marking while detecting the start and end and continuity of the road marking, the construction accuracy and work safety are improved, and labor is saved, labor is saved, and operation is highly efficient even if there are no side walls on the road. However, the present invention is extremely useful industrially because there is no fear of duplication of road markings and an economical road marking repair vehicle is obtained.

According to the invention of claim 3, in a road marking repair vehicle for automatically repairing markings of a center line, a roadside line, etc. drawn on a road, a television camera movable along with a paint nozzle, and the television. A yellow optical bandpass filter mounted on the front of the camera, a mirror mounted in front of the yellow optical bandpass filter, a scanner for driving the mirror around the horizontal axis, and an image captured by the TV camera are stored. Image memory, a sign detecting means for detecting a road sign from the image stored in the image memory, a start / end detecting means for detecting the start / end of the sign from the image stored in the image memory, and a sign detecting means A drive mechanism that drives the TV camera and paint nozzle based on the output, and a movement amount detection sensor that detects the movement amount of the repair vehicle. By detecting the precise position of the road marking while detecting the continuity of the road marking from the image captured by the TV camera, the road marking area can be stably extracted in addition to the effect of the invention of claim 2. In addition to facilitating the detection of the target road marking, the paint economy can be achieved by allowing the paint to flow out only when the paint is present on the road marking by the action of the start / end detection means. It is useful.

According to the invention of claim 4, in a road marking repair vehicle for automatically repairing markings such as a center line and a roadside line drawn on a road, a one-dimensional line camera movable with a paint nozzle, A yellow optical bandpass filter mounted on the front surface of the same one-dimensional line camera, a mirror mounted on the front surface of the same yellow optical bandpass filter, a scanner for driving the mirror around the horizontal axis, and the same one-dimensional line An image memory for accumulating and storing images captured by the camera, a sign detecting means for detecting a road sign from the images stored in the image memory, and a start and end of the sign from the images stored in the image memory. The same one-dimensional line camera is provided with start and end detection means and a drive mechanism for driving the same one-dimensional line camera and the paint nozzle based on the output of the sign detection means. By detecting the precise position of the road marking while detecting the continuity of the road marking from the image picked up by the video camera, the effect of the invention of claim 3 is obtained, and by shielding the portion other than the road area, the noise image INDUSTRIAL APPLICABILITY The present invention is extremely useful industrially because it has an effect of preventing imaging and facilitating detection of a target road marking.

According to the invention of claim 5, claims 1, 2,
In 3 and 4, since the lighting devices for illuminating the road are provided, respectively, the effects according to claims 1, 2, 3 and 4 can be obtained regardless of day and night, and particularly at night when the traffic volume is low. INDUSTRIAL APPLICABILITY The present invention is extremely useful industrially because an economical road marking detection / repair vehicle that maximizes utilization and operates efficiently can be obtained.

[Brief description of drawings]

FIG. 1 is an overall perspective view showing a first embodiment of an automobile road marking repair vehicle of the present invention and a flow chart of its main equipment.

FIG. 2 is an enlarged view showing a main part of FIG.

FIG. 3 is an explanatory diagram showing a procedure for extracting a road marking area from an image in FIGS. 1 and 2.

4 is a side view and a plan view of FIG. 1. FIG.

FIG. 5 is an explanatory diagram showing a procedure for detecting the start and end points of the road marking from the images in FIGS.

FIG. 6 is an overall perspective view showing a second embodiment of the present invention and a flow chart of its main equipment.

7 is an explanatory diagram showing a relationship between a field of view of the fish-eye lens of FIG. 6 and a light blocking plate.

8 is an explanatory diagram showing a procedure for extracting a road marking area from an image in the apparatus shown in FIG.

9 is an enlarged view of FIG. 8 (C).

FIG. 10 is an overall perspective view showing a third embodiment of the present invention and a flow chart of its main equipment.

11 is an enlarged view showing a main part of FIG.

12 is a side view and a plan view of FIG.

FIG. 13 is an explanatory diagram showing a procedure for extracting a road marking area from an image in FIGS. 11 to 12;

FIG. 14 is an image diagram showing a modified example of FIG.

FIG. 15 is an overall perspective view showing a fourth embodiment of the present invention and a flowchart of its main equipment.

16 is a side view and a plan view of FIG. 15. FIG.

FIG. 17 is an explanatory diagram showing a procedure for extracting a road marking area from the image in FIG.

FIG. 18 is an enlarged view showing a procedure for detecting the start and end points of a road marking which is a modified example of FIG. 17 (C).

FIG. 19 is a perspective view showing a work procedure of conventional road marking repair.

FIG. 20 is a perspective view showing a work procedure of conventional road marking repair on an expressway.

FIG. 21 is a flowchart of main devices in FIG.

22 is a diagram showing a procedure for detecting the road marking by connecting the center positions of the respective scanning lines from the image in FIG. 20. FIG.

FIG. 23 is a side view and a plan view showing a procedure for positioning the paint nozzle in FIG. 21.

[Explanation of symbols]

 1 Repair Vehicle 2 Bandpass Filter 2a First Yellow Optical Bandpass Filter 2b Second Yellow Optical Bandpass Filter 3 Television Camera 3a First Television Camera 3b Second Television Camera 3c One-dimensional Television Camera 4a, 4b Lighting 5 Drive mechanism 6 Moving amount detection sensor 7 Paint nozzle 8 Image memory 8a First image memory 8b Second image memory 8c Image memory 9 Mark detection means 9a First sign detection means 9b Second sign detection means 10 Start / end detection Means 11 Road marking 12 Wheels 13 Axle 14 Wall 15 Distance sensor 16 Start end 17 End 19a Imaging position 19b Imaging position 20 Shading plate 21 Fisheye lens 24 Mirror 25 Scanner

Claims (5)

[Claims] 1. A road marking repair vehicle that automatically repairs markings such as a center line and a roadside line drawn on a road. A first vehicle image fixedly fixed to the repair vehicle to image the road surface from a portion directly below to the front. A television camera, a second television camera that can move together with the paint nozzle and captures an image immediately in front of the paint nozzle, first and second yellow optical bandpass filters mounted on the front faces of the television cameras, respectively. First and second image memories for respectively storing images captured by the respective television cameras, first display detecting means for detecting road display from the images stored in the first image memory, and first and second image memories. A start / end detecting means for detecting the start / end of the road display from the image stored in the image memory;
Second display detecting means for detecting a road display from the image stored in the image memory, and the second television camera based on the outputs of the first and second display detecting means and the paint attached to the second television camera. A drive mechanism that moves the nozzle in the vehicle width direction and a movement amount detection sensor that detects the movement amount of the repair vehicle are provided, and the start and end and continuity of the road display can be determined from the image captured by the first TV camera. The second television camera is positioned based on the detected information and the information of the movement amount detection sensor, and the precise position of the road display is detected from the image captured by the second television camera. A featured road display detection repair vehicle. 2. In a road marking repair vehicle for automatically repairing markings such as a center line and a roadside line drawn on a road, a television camera movable along with a paint nozzle and mounted on the front surface of the television camera. With a fisheye lens,
A yellow optical bandpass filter that is attached to the front of the fisheye lens, a side light-shielding plate that is attached to the front of the yellow optical bandpass filter, an image memory that stores the image captured by the TV camera, and the same image memory Marking detection means for detecting road markings from the captured image,
The television camera includes start / end detection means for detecting the start / end of the marking from the image stored in the image memory, and a drive mechanism for driving the television camera and the paint nozzle based on the output of the marking detection means. A road display detection / repair vehicle characterized by detecting a precise position of a road marking while detecting the start and end and continuity of the road marking from an image captured by. 3. In a road marking repair vehicle for automatically repairing markings such as a center line and a roadside line drawn on a road, a television camera movable along with a paint nozzle and a yellow color attached to the front of the television camera. An optical bandpass filter, a mirror mounted in front of the yellow optical bandpass filter, a scanner for driving the mirror around the horizontal axis, an image memory for storing images taken by a TV camera, and an image memory A sign detecting means for detecting a road sign from the image stored in the image, a start / end detecting means for detecting the start / end of the sign from the image stored in the image memory, and a television camera based on the output of the sign detecting means. It is equipped with a drive mechanism that drives the paint nozzle and a movement amount detection sensor that detects the movement amount of the repair vehicle. A road display detection / repair vehicle characterized by detecting a precise position of a road marking while detecting continuity of the road marking from an imaged image. 4. A road marking repair vehicle for automatically repairing markings such as a center line and a roadside drawn on a road, and a one-dimensional line camera movable along with a paint nozzle and a front surface of the one-dimensional line camera. The yellow optical bandpass filter attached to the, the mirror attached to the front surface of the yellow optical bandpass filter, the scanner for driving the mirror around the horizontal axis, and the image captured by the same one-dimensional line camera are accumulated. An image memory for storing the image, a sign detecting means for detecting a road sign from the image stored in the image memory, a start / end detecting means for detecting the start / end of the sign from the image stored in the image memory, The one-dimensional line camera and the drive mechanism for driving the paint nozzle based on the output of the marking detection means are provided, and the one-dimensional line camera takes an image. A road display detection / repair vehicle characterized by detecting a precise position of a road marking while detecting continuity of the road marking from an image. 5. A road display detection / repair vehicle according to any one of claims 1, 2, 3 and 4, further comprising an illuminating device for illuminating a road.