JP2010190647A - Vehicle position measuring instrument and vehicle position measuring program - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To determine the position of a vehicle with high accuracy. <P>SOLUTION: A shape/color detector 19 and a positional information acquirer 20 are provided. The detector 19 is for determining the position of a feature on a camera image with the distance thereto measured by a laser range finder 3 to detect the shape and color of the feature from the position of the feature on the camera image. The acquirer 20 is for acquiring accurate positional information on a reference feature from a positional information recorder 11 in cases where the shape and color of the feature detected by the detector 19 coincide with the shape and color of the reference feature. From the accurate position information on the reference feature acquired by the acquirer 20, the position of the vehicle 1 is calculated by a vehicle position calculator 21. <P>COPYRIGHT: (C)2010,JPO&INPIT

Description

  The present invention relates to a vehicle position measuring apparatus and a vehicle position measuring program capable of measuring the position of a vehicle with high accuracy.

Currently, for outdoor positioning, a method of receiving GPS signals transmitted from GPS satellites and performing positioning has become the mainstream.
However, in urban areas where there are many high-rise buildings, GPS satellites cannot be seen from the measurement location (for example, if the measurement location is in a valley of a building, the GPS satellite may be hidden in the building), and the measurement environment is poor. May be.
When the measurement environment is bad, there are situations where the position cannot be measured at all, or even if the measurement is possible, the error becomes large.

As a method of eliminating this measurement error, there is a method of correcting the measurement position using the distance between the map and the feature.
For example, in the following Patent Document 1, the distance from the vehicle to the feature is measured using a radar, and in the following Patent Document 2, the distance from the vehicle to the feature is measured using a stereo camera. A vehicle position measurement device that corrects a measurement position according to a distance from a vehicle to a feature is disclosed.
However, in the vehicle position measurement devices disclosed in Patent Documents 1 and 2, the current position is estimated based on road map data and the position is corrected, so the position is corrected. However, the accuracy of the position depends on the accuracy of the map.
The accuracy of the map here includes not only the accuracy of the road on the map but also the level of accuracy with which the feature is described on the map.

The vehicle position measurement devices disclosed in Patent Documents 1 and 2 mainly specify the lane on which the vehicle currently exists, and specify the exact position of the vehicle. Not what you want.
In addition, in the vehicle position measurement devices disclosed in Patent Documents 1 and 2, since there is no means for grasping the shape of the feature, the distance from the vehicle to the feature is determined using a radar or a stereo camera. Even if it is measured, it is not known which part of the feature the measured distance is measured.
For this reason, even if the distance measured by the radar or the stereo camera is used, the position cannot be accurately corrected.

JP 2007-218848 A (paragraph number [0016], FIG. 1) JP 2007-232690 A (paragraph number [0029], FIG. 1)

  Since the conventional vehicle position measuring device is configured as described above, even if the distance from the vehicle to the feature is measured by using a radar or a stereo camera, the measured distance may be any part of the feature. I don't know if it was measured. For this reason, there has been a problem that even if the measured position of the vehicle is corrected using the distance measured by the radar or the stereo camera, the exact position of the vehicle cannot be specified.

  The present invention has been made to solve the above-described problems, and an object of the present invention is to obtain a vehicle position measuring device and a vehicle position measuring program capable of specifying the position of the vehicle with high accuracy.

  A vehicle position measurement device according to the present invention includes an image acquisition unit that captures an image of the surroundings of the vehicle and acquires an image of the surroundings, a distance measurement unit that measures a distance from the vehicle to a surrounding feature, and a distance A shape color detecting means for identifying a position on the image of the feature whose distance is measured by the measuring means and detecting the shape and color of the feature from the position on the image of the feature; and a shape color detecting means Position information acquisition means for acquiring accurate position information of the reference feature from the position information recording means when the shape and color of the feature detected by the above match the shape and color of the reference feature; The calculating means calculates the position of the vehicle from the accurate position information of the reference feature acquired by the position information acquiring means and the distance to the feature measured by the distance measuring means.

  According to the present invention, the image acquisition unit that captures the periphery of the vehicle and acquires the surrounding image, the distance measurement unit that measures the distance from the vehicle to the surrounding feature, and the distance measurement unit The shape color detecting means for identifying the position on the image of the feature measured and detecting the shape and color of the feature from the position on the image of the feature, and the shape color detecting means When the shape and color of the feature matches the shape and color of the reference feature, a position information acquisition unit that acquires accurate position information of the reference feature from the position information recording unit is provided, and the vehicle position calculation unit is positioned Since the position of the vehicle is calculated from the accurate position information of the reference feature acquired by the information acquisition means and the distance to the feature measured by the distance measurement means, the position of the vehicle is specified with high accuracy. There is an effect that can.

It is a block diagram which shows the vehicle from which a position is measured by the vehicle position measuring apparatus by Embodiment 1 of this invention. It is a block diagram which shows the vehicle position measuring apparatus by Embodiment 1 of this invention. It is a flowchart which shows the processing content of the vehicle position measuring apparatus by Embodiment 1 of this invention. It is a flowchart which shows the processing content of the shape color detection part 19 in the vehicle position measuring device by Embodiment 1 of this invention. It is explanatory drawing which shows a traffic light typically. It is explanatory drawing which shows the distance measurement point in case a feature is a tunnel.

Embodiment 1 FIG.
1 is a block diagram showing a vehicle whose position is measured by a vehicle position measuring apparatus according to Embodiment 1 of the present invention.
In FIG. 1, a vehicle 1 is a measurement target whose current position is measured by a vehicle position measurement device.
The camera 2 is installed in the vehicle 1 and performs a process of photographing the upper front portion (surrounding) of the vehicle 1 and outputting image data indicating an image of the upper front portion.
The laser range finder 3 is installed in the vehicle 1, measures the distance to the features (eg, traffic lights, signs, tunnels) existing at the upper front of the vehicle 1, and obtains distance data indicating the distance to the features. Perform the output process.

The GPS receiver 4 receives a GPS signal transmitted from a GPS satellite and performs a process of measuring the position of the vehicle 1 using the GPS signal.
The measurement result of the GPS receiver 4 is mainly used when the position information acquisition unit 20 described later acquires accurate position information of the reference feature, and is output as the calculation result of the vehicle position measurement device. is not.
Here, although what showed the position of the vehicle 1 using a GPS signal was shown, it uses vehicle information, such as a measurement result of angle and acceleration by a gyro, vehicle speed, the number of rotations of a wheel, and a handle angle, for example. The position of the vehicle 1 may be measured.

FIG. 2 is a block diagram showing a vehicle position measuring apparatus according to Embodiment 1 of the present invention.
In FIG. 2, a position information recording unit 11 is a recording medium such as a hard disk that records in advance accurate position information of a reference feature having a characteristic shape and color.
Here, the reference feature may be, for example, all the traffic lights that exist on the road, but it is difficult to measure accurate position information of all the traffic lights in advance. 1 assumes a part of traffic lights and the like existing on the road.
The position information recording unit 11 constitutes position information recording means.

The control unit 12 performs processing for controlling the photographing time of the camera 2, the measurement time of the laser range finder 3, the scanning angle of the laser range finder 3, and the like.
The image data acquisition unit 13 acquires image data output from the camera 2, records the image data in the image data storage unit 14, and records the shooting time of the camera 2 in the shooting time storage unit 15. To do.
The image data storage unit 14 is a recording medium such as a hard disk for storing image data.
The shooting time storage unit 15 is a recording medium such as a hard disk for storing the shooting time of the camera 2.
The camera 2 and the image data acquisition unit 13 constitute image acquisition means.

The distance data acquisition unit 16 acquires the distance data output from the laser range finder 3, records the distance data in the distance data storage unit 17, and records the measurement time of the laser range finder 3 in the measurement time storage unit 18. Perform the process.
The distance data storage unit 17 is a recording medium such as a hard disk for storing distance data.
The measurement time storage unit 18 is a recording medium such as a hard disk for storing the measurement time of the laser range finder 3.
The laser range finder 3 and the distance data acquisition unit 16 constitute a distance measuring unit.

  The shape color detection unit 19 includes image data stored in the image data storage unit 14, shooting time of the camera 2 stored in the shooting time storage unit 15, distance data and measurement time stored in the distance data acquisition unit 16. Using the measurement time of the laser range finder 3 stored by the storage unit 18, the position of the feature whose distance is measured by the laser range finder 3 is specified on the image, and the feature on the image is displayed. A process for detecting the shape and color of the feature from the position is performed. The shape color detecting unit 19 constitutes a shape color detecting unit.

The position information acquisition unit 20 refers to the position measured by the GPS receiver 4 when the shape and color of the feature detected by the shape color detection unit 19 match the shape and color of the reference feature. From the position information of the reference feature recorded by the recording unit 11, a process of acquiring the position information of the reference feature corresponding to the feature is performed. The position information acquisition unit 20 constitutes position information acquisition means.
The vehicle position calculation unit 21 performs processing for calculating the position of the vehicle 1 from the accurate position information of the reference feature acquired by the position information acquisition unit 20 and the distance data stored by the distance data storage unit 17. The vehicle position calculation unit 21 constitutes vehicle position calculation means.

  In the example of FIG. 2, each of the control unit 12, the image data acquisition unit 13, the distance data acquisition unit 16, the shape color detection unit 19, the position information acquisition unit 20, and the vehicle position calculation unit 21 that are components of the vehicle position measurement device. Is configured with dedicated hardware (for example, a semiconductor integrated circuit on which a CPU is mounted). However, when the vehicle position measuring device is configured with a computer, the control unit 12 acquires image data. Unit 13, distance data acquisition unit 16, shape color detection unit 19, position information acquisition unit 20, and vehicle position calculation unit 21, a vehicle position measurement program is stored in the memory of the computer. The CPU may execute a vehicle position measurement program stored in the memory.

  FIG. 3 is a flowchart showing the processing contents of the vehicle position measuring device according to the first embodiment of the present invention. FIG. 4 shows the processing contents of the shape color detecting unit 19 in the vehicle position measuring device according to the first embodiment of the present invention. It is a flowchart which shows.

Next, the operation will be described.
Under the instruction of the control unit 12, the camera 2 periodically captures the upper front portion of the vehicle 1 and outputs image data indicating an upper front image to the image data acquisition unit 13.
When receiving image data from the camera 2, the image data acquisition unit 13 records the image data in the image data storage unit 14 and records the shooting time of the camera 2 in the shooting time storage unit 15.

The laser range finder 3 measures the distance to a feature (for example, a traffic light, a sign, a tunnel) existing at the front upper part of the vehicle 1 under the instruction of the control unit 12, and determines the distance to the feature. The indicated distance data is output to the distance data acquisition unit 16.
Here, the control unit 12 periodically changes the angle of the laser transmitted from the laser range finder 3 so that the laser range finder 3 scans the front upper part of the vehicle 1 (see FIG. 1). .
The control unit 12 controls the shooting time of the camera 2 and the measurement time of the laser range finder 3 so that the measurement frequency of the laser range finder 3 is higher than the shooting frequency of the camera 2.
When the distance data acquisition unit 16 receives the distance data from the laser range finder 3, the distance data acquisition unit 16 records the distance data in the distance data storage unit 17 and records the measurement time of the laser range finder 3 in the measurement time storage unit 18.

Here, the reason why the camera 2 captures the upper front part of the vehicle 1 and the laser range finder 3 measures the distance to the feature existing in the upper front part of the vehicle 1 is as follows.
Many features such as buildings, roadside trees, and signs are arranged in the left-right direction with respect to the traveling direction of the vehicle 1.
On the other hand, since the front side of the vehicle is a passage, the feature exists only in the upward direction, and the types of the existing feature are also limited. The existing features are limited to traffic lights, overpasses and tunnels, and signs.
For this reason, it is better to detect the feature existing in the upper front part of the vehicle than to detect the feature present in the left-right direction of the vehicle (there is more likely to be a detection error such as an oversight). Therefore, the camera 2 takes an image of the front upper part of the vehicle 1 and the laser range finder 3 measures the distance to the feature existing in the upper front part of the vehicle 1.

In this Embodiment 1, it considers making detection easier by making the feature which has a characteristic shape among these features into a detection target.
In the first embodiment, for example, a traffic signal having the characteristics that “three objects having the same shape are arranged” and “the arrangement order and the color order are determined” is set as a detection target feature.
FIG. 5 is an explanatory view schematically showing a traffic light.
By arranging the color information of the laser that hits the traffic light from the distance measured by the laser range finder 3 and the color relationship of the camera image, almost the same number of “red / yellow / blue (or green)” pairs are arranged. . The shape color detection unit 19 described later detects the shape and color of the traffic light by detecting this.

The shape color detection unit 19 includes image data stored in the image data storage unit 14, shooting time of the camera 2 stored in the shooting time storage unit 15, distance data stored in the distance data acquisition unit 16, and measurement. Using the measurement time of the laser range finder 3 stored by the time storage unit 18, the position of the feature whose distance is measured by the laser range finder 3 is specified on the camera image, and on the camera image of the feature The shape and color of the feature are detected from the position at (ST1 in FIG. 3).
Hereinafter, the processing content of the shape color detection part 19 is demonstrated concretely.

First, the shape color detection unit 19 differs in the installation position and installation angle of the camera 2 in the vehicle 1 from the installation position and installation angle of the laser range finder 3 in the vehicle 1 (however, the camera 2 and the laser range finder 3 The installation position and installation angle are known), and distance data, which is a measurement result of the laser range finder 3, is converted into relative coordinates of the camera 2 (step ST11 in FIG. 4).
Since the process itself for converting the measurement data of the laser range finder 3 into the relative coordinates of the camera 2 is a known technique, a detailed description thereof is omitted. For example, the measurement result of the laser range finder 3 is three-dimensionally rotated and translated. Can be calculated.

When the shape color detection unit 19 converts the measurement data of the laser range finder 3 into the relative coordinates of the camera 2, the shape color detection unit 19 specifies the position of the converted measurement data on the camera image.
However, when the position of the converted measurement data on the camera image is specified, the distance data measurement frequency by the laser range finder 3 is higher than the imaging frequency of the image data by the camera 2, so the distance data is converted to the camera 2. It is necessary to correct the distance and position from the vehicle position at the time of shooting.
Therefore, the shape color detection unit 19 actually refers to the shooting time of the camera 2 stored in the shooting time storage unit 15 and the measurement time of the laser range finder 3 stored in the measurement time storage unit 18. Furthermore, the difference between the vehicle position at the time when the distance was measured by the laser range finder 3 and the vehicle position at the time when the camera 2 was photographed before that (the relative position of the vehicle position at the time of distance measurement to the vehicle position at the time of image capturing) (Position) is calculated (step ST12).
For example, the measurement result of the GPS receiver 4 may be used as the vehicle position when the distance is measured by the laser range finder 3 and the vehicle position when the distance is measured by the camera 2. You may make it utilize the past calculation result by the vehicle position calculation part 21 mentioned later.

When the shape color detection unit 19 calculates the difference between the vehicle position at the time of image capturing and the vehicle position at the time of distance measurement, the difference between the vehicle positions is converted into distance data that has been converted into relative coordinates of the camera 2 first. By adding, the position on the camera image of the feature whose distance is measured by the laser range finder 3 is calculated (step ST13).
When calculating the position of the feature on the camera image, it is assumed that the specifications of the camera 2 (eg, image size, focal length, etc.) are known.

When the shape color detection unit 19 calculates the position of the feature whose distance is measured by the laser range finder 3 on the camera image, the shape color detection unit 19 confirms the color of the image at the position and specifies the color of the feature. The shape color detection unit 19 sets the color of the pixel at the position on the camera image where the feature exists as the color of the feature (step ST14).
The shape color detection unit 19 repeatedly performs the processes of steps ST11 to ST14 on the distance data within a certain period of time, thereby obtaining the positions of a plurality of points on the feature, thereby determining the shape of the feature and the ground. Know the color inside an object.

When the shape color detection unit 19 detects the shape and color of the feature, the position information acquisition unit 20 determines whether or not the shape and color of the feature matches the shape and color of the reference feature (FIG. 3). Step ST2).
That is, for example, if the reference feature is a traffic light, the position information acquisition unit 20 may continuously display “red / yellow / blue (green)” as the feature color as shown in FIG. If “red / yellow / blue (green)” appears continuously, it is confirmed whether or not the shape of “red / yellow / blue (green)” appears as a circle.
When confirming whether or not the shape is circular, a series of “red, yellow, and blue (green)” appears in the same direction in the vertical direction (the number of consecutive occurrences is around To see if it appears).

The position information acquisition unit 20 refers to the position measured by the GPS receiver 4 when the shape and color of the feature detected by the shape color detection unit 19 match the shape and color of the reference feature. It is confirmed whether or not the position information of the reference feature corresponding to the feature is recorded by the position information recording unit 11 (step ST3).
For example, the distance data obtained by the laser range finder 3 is added to the position measured by the GPS receiver 4 to obtain an approximate position of the feature, and the position information indicating the position close to the approximate position is obtained from the reference feature. Judged as position information.
The position measured by the GPS receiver 4 includes an error, and the exact position of the feature cannot be obtained. However, the number of traffic lights as reference features is not set between 1 and 2 m. Therefore, some errors are allowed.
Here, the approximate position of the feature is obtained using the position measured by the GPS receiver 4. For example, the traffic signal that is the reference feature is referred to with reference to the information on the road on which the vehicle 1 is running. You may make it guess.

  If the position information recording unit 11 records the accurate position information of the reference feature corresponding to the feature whose shape and color have been detected by the shape color detection unit 19, the position information acquisition unit 20 can detect the reference feature. Is output to the vehicle position calculation unit 21.

When the vehicle position calculation unit 21 receives the accurate position information of the traffic signal as the reference feature from the position information acquisition unit 20, the vehicle position calculation unit 21 calculates the vehicle 1 from the accurate position information and the distance data stored in the distance data storage unit 17. Calculate the position.
Hereinafter, the position calculation process of the vehicle 1 in the vehicle position calculation part 21 is demonstrated concretely.

The vehicle position calculation unit 21 calculates the relative position between the traffic light that is the feature and the vehicle 1 using the distance data that is the measurement result of the laser range finder 3 (step ST4).
For example, when the position of the center of the yellow signal of the traffic light is measured as the position information of the traffic light that is the reference feature, the position of the center of the yellow signal of the traffic light and the relative position of the vehicle 1 are calculated.
When the vehicle position calculation unit 21 calculates the relative position of the traffic signal that is the feature and the vehicle 1, the vehicle position calculation unit 21 grasps the true position of the traffic signal indicated by the accurate position information of the traffic signal that is the reference feature, and the coordinate system of the position of the traffic signal Then, after the coordinate system of the position of the vehicle 1 is matched, the position of the vehicle 1 is calculated by translating the true position of the traffic light by the relative position (step ST5).

As apparent from the above, according to the first embodiment, the position of the feature whose distance is measured by the laser range finder 3 is specified on the camera image, and the position of the feature on the camera image is determined. If the shape and color of the feature detected by the shape color detection unit 19 and the shape and color of the feature detected by the shape color detection unit 19 coincide with the shape and color of the reference feature, the position information recording unit 11 The position information acquisition unit 20 that acquires the accurate position information of the reference feature from the position information acquisition unit 20 is provided, and the vehicle position calculation unit 21 acquires the accurate position information of the reference feature acquired by the position information acquisition unit 20 and the laser range finder 3. Since the position of the vehicle 1 is calculated from the distance to the feature measured by the above, there is an effect that the position of the vehicle 1 can be specified with high accuracy.
In this case, the position error calculated by the vehicle position calculation unit 21 depends on the error of the laser range finder 3 and the error when measuring the position of the traffic light. However, the positioning error is more accurate than when positioning by receiving a GPS signal. Is possible.

Note that since the traffic light does not light two colors at the same time, only one color may be obtained depending on the situation.
In that case, the same place is acquired many times by passing through the same place multiple times, and all the colors of the signal are acquired. In this case, there is a high possibility that the positional relationship is shifted according to the acquisition time, but the object can be achieved by allowing a slight shift in the signal recognition stage.

Embodiment 2. FIG.
In the first embodiment, an example in which the reference feature is a traffic signal is shown. However, the reference feature is not limited to a traffic signal. For example, the reference feature may be a sign.
Like the traffic light, the sign has a characteristic shape and color.
Unlike traffic lights, signs do not flicker and have little color change. Moreover, like a traffic light, it may be arrange | positioned above a road and a sign can be made into the target object of a detection.

Embodiment 3 FIG.
In the first embodiment, an example in which the reference feature is a traffic signal is shown. However, the reference feature is not limited to a traffic signal. For example, a location such as a tunnel entrance may be used as the reference feature. .
In the case of a tunnel, unlike a traffic light or a sign, it can be recognized by confirming that “the vehicle is vacant (the measured distance data does not exist)” (see FIG. 6).

  DESCRIPTION OF SYMBOLS 1 Vehicle, 2 Cameras (image acquisition means), 3 Laser range finder (distance measurement means), 4 GPS receiver, 11 Position information recording part (position information recording means), 12 Control part, 13 Image data acquisition part (Image acquisition) Means), 14 image data storage section, 15 photographing time storage section, 16 distance data acquisition section (distance measurement means), 17 distance data storage section, 18 measurement time storage section, 19 shape color detection section (shape color detection means), 20 position information acquisition part (position information acquisition means), 21 vehicle position calculation part (vehicle position calculation means).

Claims (6)

  A position information recording unit that records accurate position information of a reference feature having a characteristic shape and color in advance, an image acquisition unit that captures a surrounding image of the vehicle, A distance measuring means for measuring a distance from the vehicle to a surrounding feature, a position on the image of the feature whose distance is measured by the distance measuring means, and an image of the feature When the shape color detection means for detecting the shape and color of the feature from the position above, and the shape and color of the feature detected by the shape color detection means match the shape and color of the reference feature, Position information acquisition means for acquiring accurate position information of the reference feature from the position information recording means, accurate position information of the reference feature acquired by the position information acquisition means, and the ground measured by the distance measurement means The position of the vehicle from the distance to the object Vehicle position measuring device and a vehicle position calculating means for output.   2. The vehicle position measuring apparatus according to claim 1, wherein the image acquiring means photographs the upper front part of the vehicle, and the distance measuring means measures the distance to the feature existing in the upper front part of the vehicle.   3. The vehicle position measuring apparatus according to claim 1, wherein the feature is a traffic signal on a road.   3. The vehicle position measuring device according to claim 1, wherein the feature is a road sign.   3. The vehicle position measuring apparatus according to claim 1, wherein the feature is a tunnel.   A position information recording processing procedure for recording accurate positional information of a reference feature having a characteristic shape and color in advance, an image acquisition processing procedure for capturing a surrounding image by photographing the surroundings of the vehicle, A distance measurement processing procedure for measuring a distance from the vehicle to the surrounding feature, and a position on the image of the feature whose distance is measured by the distance measurement processing procedure is specified; and The shape color detection processing procedure for detecting the shape and color of the feature from the position on the image of the image, and the shape and color of the feature detected by the shape color detection processing procedure match the shape and color of the reference feature The position information acquisition processing procedure for acquiring the accurate position information of the reference feature from the position information recording processing procedure, the accurate position information of the reference feature acquired by the position information acquisition processing procedure, and the distance. Measured by the measurement procedure Vehicle position measuring program for executing a vehicle position calculation processing procedure in a computer for calculating the distance from the vehicle position to the feature.

Images