JP2002092797A - Traffic information providing system - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a traffic information providing system for determining the abnormal approach of a moving vehicle and informing the same, capable of being easily installed and operated, and improving the reliability in determination. SOLUTION: A wide angle distance measuring unit 2 includes a conical mirror 30 and a focus variable lens unit 34 in an optical system 20. A circumferential area is projected as one reflection image by the conical mirror 30. A distance of each point is determined from a condition of proper focus by changing a focal length of the focus variable lens unit 34. A control part 4 compares a distance image and a reference image of only existing matters to detect a specified substance, and detects the approaching of the vehicle and a walker, or the like by the movement between frames. The detected information is transmitted to a driver of the vehicle and the walker by a communicating part and a warning part.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

The present invention relates to a traffic information providing system for notifying a passerby that an obstacle or the like is approaching.

[0002]

2. Description of the Related Art Conventionally, there has been known a system which detects a moving object on a traffic route such as a road and provides information such as an alarm in response to the detection. In such a system, for example,
A sensor such as an ultrasonic sensor is installed above the road surface, and a vehicle passing below the sensor is detected. In addition, sensing a steel vehicle using a loop coil as a sensor is also performed.

[0003]

A conventionally used sensor such as an ultrasonic sensor detects whether or not an object exists between a road and the sensor. To detect an object occupying a predetermined size or more. The threshold value of the size is generally set for a car. In this case, a small object such as a pedestrian is not detected. On the other hand, if a small object can be detected, the vehicle cannot be distinguished from the car. was there. The loop coil has a problem that only an iron vehicle can be detected, and a pedestrian cannot be detected.

Further, since the monitoring area of these sensors is a relatively narrow spot-shaped area, there is a problem that sensors must be arranged at a plurality of locations in order to monitor a wide area, which is costly. Further, there is a problem that processing for processing information from a large number of sensors is complicated. For example, changing the location or number of sensors changes the dependency between sensor information,
Requires modification of processing software.

[0005] On the other hand, sensing using an image sensor is characterized by being capable of non-contact and monitoring over a wide range. However, it is difficult to recognize an object based on changes in brightness or color in a screen. Robustness was low and reliability was low. For example, there is a case where a contrast between a figure drawn on a road surface or a shadow and a sun is erroneously recognized as unevenness, and an erroneous determination is made that there is an obstacle or the like.

SUMMARY OF THE INVENTION The present invention has been made to solve the above-described problem. The present invention monitors a wide area by installing it at one place, and accurately notifies a pedestrian that an obstacle or the like is approaching. The purpose is to provide a traffic information providing system.

[0007]

A traffic information providing system according to the present invention measures a distance to an object in the peripheral area based on a reflection image of the peripheral area acquired using a conical mirror, A wide-angle distance measuring device that generates a distance image corresponding to the peripheral region reflected in an image; a target detection unit that detects a predetermined target based on the distance image; And a pedestrian notification means for notifying a pedestrian approaching to.

According to the present invention, the wide-angle distance measuring device can basically monitor the entire azimuth by using a conical mirror, and can monitor a wide-angle elevation angle range in each azimuth. The bus of the conical mirror can also be defined, and a wide range of surrounding areas can be monitored with one measuring instrument. The measuring device measures a distance to an object reflected at each point in the reflection image based on the reflection image of the conical mirror, and generates a distance image in which the distance is associated with each coordinate of the reflection image as image data. Output. The target detecting means detects the presence of a predetermined target such as an automobile, a pedestrian or an obstacle on a road surface, based on the distance image. Then, the pedestrian notification means notifies the pedestrian or pedestrian of another vehicle approaching the detected target object of the detected information of the target object.
This pedestrian notification means may be a display device or an audio alarm device provided on a traffic road, or may be a device that wirelessly transmits to a device that moves with a passing vehicle or a pedestrian. The position of the pedestrian is not necessarily limited to the area monitored by the wide-angle distance measuring device, and a pedestrian outside the area may be notified. Further, the passer-by notification means may display and transmit the detection information without necessarily confirming whether there is a passer-by who can receive the detection information to be notified.

The traffic information providing system according to another aspect of the present invention includes a passer detecting means for detecting the passer based on the distance image, and the passer notifying means notifies the passer of the passer. The detection information is generated according to the positional relationship between the target object and the pedestrian.

According to the present invention, not only a target object but also a pedestrian is detected based on a distance image. The detection information is generated according to the positional relationship between the target and the passerby, and the detection information is notified to the corresponding passerby.

In a traffic information providing system according to another aspect of the present invention, the wide-angle distance measuring device is arranged so that a central axis of the conical mirror is directed to a collision caution point, and is provided on the reflected image viewed from the central axis. The distance image is continuously generated based on the detection result of the target object detection means and the pedestrian detection means with respect to the time-series distance image, based on the detection results of the target object and the passerby. A collision danger detecting means for detecting that the passerby and the passerby approach each other at the center of the distance image and notifying the passerby of a possibility of collision;

According to the present invention, distance images are generated in chronological order, and the movement of an object and a pedestrian is detected from a temporal change. By arranging the central axis of the conical mirror of the wide-angle distance measuring instrument toward the collision caution point, the center of the reflected image corresponds to the collision caution point. For example, an intersection or a corner is set as a collision caution point. The fact that the target and the pedestrian are heading toward the collision attention point is detected from the fact that the distance between each position in the distance image and the center of the distance image decreases with time.

The traffic information providing system according to another aspect of the present invention includes reference image holding means for holding, as a reference distance image, the distance image generated in a state where the object does not exist, and detects the target object. The means detects the target object by comparing the distance image measured as needed with the reference distance image.

According to the present invention, the object detecting means detects a change in the shape of an object in the peripheral area of the wide-angle distance measuring device by comparing the distance image measured as needed with the reference image. , And those that match the target are detected from the changed shapes.

[0015] Still another traffic information providing system according to the present invention includes reference image holding means for holding the distance image generated in a state where the object and the pedestrian do not exist as a reference distance image, The target object detection unit detects the target object by comparing the distance image measured at any time with the reference distance image, and the pedestrian detection unit detects the distance image measured at any time as the reference distance image. The passerby is detected by comparison.

According to the present invention, the object detecting means detects a change in the shape of an object in the peripheral area of the wide-angle distance measuring device by comparing the distance image measured as needed with the reference image. , And those that match the target and those that match the pedestrian are detected from the changed shapes.

[0017]

Next, embodiments of the present invention will be described with reference to the drawings.

FIG. 1 is a schematic block diagram showing the overall configuration of a traffic information providing system according to an embodiment of the present invention. This system detects moving vehicles and obstacles at intersections and corners where visibility is poor and notifies other vehicles and pedestrians and other pedestrians. Can be.

The present system includes a wide-angle distance measuring device 2, a control unit 4, a communication unit 6, and a warning unit 8. FIG.
1 is a schematic overall configuration diagram of a wide-angle distance measuring device 2. FIG.
The wide angle distance measuring device 2 includes an optical system 20, a variable focus lens unit control unit 22, and an image processing unit 24. The optical system 20 includes a conical mirror 30, a main lens 32, a variable focus lens unit 34, and an electronic camera 36.

The outer surface of the conical mirror 30 is a mirror, and can reflect light from all directions around the central axis in the direction of the central axis. By making the conical surface convex in the elevation angle direction, the angle of view in the elevation angle direction reflected on the reflection image of the conical mirror 30 as viewed from the central axis direction can be enlarged. The closer to the circumference of the reflection image, the more the actual position of the object reflected at that point and the conical mirror 30
An angle (hereinafter, referred to as an elevation angle す) formed by a straight line connecting to the central axis of the conical mirror 30 becomes large.

Main lens 32 and variable focus lens unit 3
Reference numeral 4 denotes an imaging unit that forms a reflection image of the conical mirror 30 as viewed from the center axis direction of the conical mirror 30. The variable focus lens unit controller 22 changes the focal length of the variable focus lens unit 34. The image processing unit 24 detects an in-focus area of the reflected image picked up by the electronic camera 36 at each focal length of the focus variable lens unit 34, and detects an in-focus area based on the optical conditions at that time. Find the distance to the object reflected in. The measurement of the distance is performed for each point of the reflected image by changing the focal length of the focus variable lens unit 34, and as a result, an image corresponding to the reflected image is obtained at each point of the reflected image. A distance image in which the distance to the object to be reflected is the value of each pixel is generated.

The control unit 4 controls the wide-angle distance measuring device 2, the communication unit 6, and the warning unit 8. The control performed by the control unit 4 includes control of starting, stopping, and initialization processing of each unit.
Specifically, for the wide-angle distance measuring device 2, while outputting a control instruction to the focus variable lens unit control unit 22 and the image processing unit 24, the focus variable lens unit control unit 22
Based on the control state signal from the control unit, whether the unit can be controlled is detected and the system operation state is monitored. Further, the control unit 4 fetches a distance image from the image processing unit 24, and obtains information such as a position, a distance, and a size of an obstacle, a vehicle, and a pedestrian in the monitoring target area based on the distance image. Further, the control unit 4 generates a communication message transmitted from the communication unit 6 and a warning message notified from the warning unit 8 based on the obtained information, and outputs them to the communication unit 6 and the warning unit 8, respectively. Also, it controls operations such as information transmission / reception operations in the communication unit 6 and display / voice output of a warning message in the warning unit 8.

The communication section 6 communicates the status in the monitoring area to the information terminal of a vehicle or a pedestrian in or near the monitoring area of the present system by radio. The information to be transmitted has appropriate contents depending on the situation where the vehicle or pedestrian as the receiver is placed, and is transmitted to the corresponding receiver.

The warning unit 8 reports the situation in the monitoring area using a display device or a voice utterance device. Consideration is given to the position and orientation of the display device and sound generator so that appropriate information is transmitted according to the situation where each vehicle and each pedestrian is placed, and the content of the message is also transmitted to the other party. It is assumed that it has been accepted.

Further, the function, operation and processing of the present system will be described in detail. In the present system, as described above, the area around the place where the wide-angle distance measuring device 2 is arranged is monitored based on the distance image. FIG. 3 and FIG. 4 are schematic diagrams for explaining a reflection image by the conical mirror 30 of the wide-angle distance measuring device 2 arranged above the road surface. FIG. It shows an object point and an image point of the object on a conical mirror. FIG. 4 shows an object point viewed from the side of the conical mirror 30 and an image point on the conical mirror.

As shown in FIG. 3, as a feature of the cone mirror image, the orientation of the object point with respect to the cone mirror coincides with the orientation of the image point in the reflection image of the cone mirror. That is, in the example shown in FIG. 3, the object point A and its image point A ', and the object point B and its image point B' are located in the same direction as viewed from the central axis O of the conical mirror. Thus, the actual azimuth of the object can be obtained from the azimuth of the image in the reflection image.

As another characteristic of the conical mirror image,
If the conical mirror image reflects a plane, the closer to the periphery of the reflected image, the farther it points. That is, when monitoring a vehicle running on a road or a pedestrian on a sidewalk, they are basically on a plane, and the distance from the wide-angle distance measuring device 2 can be obtained from the distance from the center of the reflected image. it can.

FIG. 5 is a schematic diagram showing an example of the relationship between the actual position of the object on the road surface and the position of the image on the reflection image of the conical mirror. FIG. 5A is a schematic view of the intersection as viewed from above, and shows a car traveling from south to north approaching the intersection. In the figure, the car is at position 40 at time t ₁ and at position 42 at time t ₂ (t ₂ > t ₁ ). FIG. 5B is a schematic diagram of a reflection image of a conical mirror of the wide-angle distance measuring device 2 installed above the center of the intersection. If the upper part of the screen of the reflection image is east, the left side of the screen is south. That is, the vehicle shown in FIG. 5A is reflected on the left side of the screen, and it is detected that the vehicle exists on the road extending south from the intersection. Also, if the image of this vehicle is projected at the position 44 at time t _{1 and} at the position 46 at time t ₂ , that is, as the vehicle approaches the center of the reflected image as the image frame is followed, It is detected that the vehicle is approaching the intersection. Conversely, if the vehicle moves toward the outer periphery of the reflection image, it is detected that the vehicle is moving away from the intersection.

For example, in this system, the screen of the reflection image is divided into four areas 50 to 56 in the east, west, north and south as shown in FIG. Watch the image. At the same time, when an image moving in the image center direction is detected in two or more areas, a vehicle approaching the intersection in the corresponding area is directed to an electronic bulletin board or a communication device installed in front of the intersection in each area. Information can be transmitted to prevent a head-on collision at an intersection.

FIG. 6 is a schematic processing flowchart of the present system. Since so-called existing objects such as road surfaces and buildings are fixed objects, their distance images do not change. That is, it becomes the background. This system generates a distance image (reference image) of only the existing object in advance, stores the distance image in the storage device of the control unit 4, holds it (S60), and operates on it. In the operation state, moving objects on the road surface are detected (S62), and whether or not they may collide is determined (S64). Notifying the approach of a moving object to prevent an accident (S66)
The operation is repeated.

The detection of a moving object during operation is obtained by comparing a distance image acquired during operation with a reference image. In operation, when a moving object such as a vehicle or a pedestrian is within the monitoring area of the wide-angle distance measuring device 2, the distance image differs from the reference image which is a distance image only of the existing object. That is, if the difference between the distance image generated as needed and the reference image is calculated, only the moving object portion remains without being zero. Therefore, a distance image of only the moving object can be obtained. Further, since the reflection image of the conical mirror is divided for each road direction extending from the intersection, if the coordinates in the image are determined, it is possible to determine on which road the moving object exists. Even if there are a plurality of moving objects, if the coordinates in the screen are uniquely determined, the location of the moving objects is also uniquely determined.

The distance from the center of the reflection image and the actual distance depend on a certain function, and since a distance image is obtained, the apparent size (area) of the object on the screen is obtained.
Can be used to estimate the actual size. In addition, by classifying the vertical and horizontal lengths of the object, it is estimated and identified whether the object is an automobile, a motorcycle or a bicycle, a pedestrian, or the like.

FIG. 7 is a schematic diagram showing an example of the relationship between the actual position of the object on the road surface and the position of the image on the reflection image of the conical mirror for explaining the outline of the determination of the possibility of collision. is there. FIG. 7A is a schematic view of the intersection viewed from above, and shows a car traveling from east to west and a car traveling from south to north approaching the intersection. In the figure, the car from the east is at position 70 at time t ₁ and at position 72 at time t ₂ (t ₂ > t ₁ ). The automobile from the south is in the position 74 at time t _1,
At position 76 at time t _2. FIG. 7B is a schematic diagram of a reflection image of a conical mirror of the wide-angle distance measuring device 2 installed above the center of the intersection. In the present system, two vehicles traveling on roads in different directions are photographed by the electronic camera 36 of one wide-angle distance measuring device 2 and monitored on one reflected image. Here, each time the image frame is followed, if the two approach the center of the reflected image, it can be said that the two are approaching the intersection, so that a collision between the encounters may occur. Now, assuming that the eastern road is the priority road, vehicles traveling on the southern road must stop temporarily before the intersection. In addition, a vehicle running on priority roads should be slowed down just in case. Therefore, the control unit 4 of the present system notifies the communication unit 6 and the warning unit 8 of the two when the two approach each other and there is a possibility of a collision from the distance image to prevent the collision accident. To prevent.

For example, the information to be notified to the vehicles can be generated according to the approaching state of each vehicle to the intersection. For example, different information can be generated according to the approach direction to the intersection. Specifically, in the state where the wide-angle distance measuring device 2 is installed at the intersection as described above, the direction in the screen always coincides with the actual azimuth centering on the intersection. A vehicle approaching from above the screen is always a vehicle approaching from the intersection east road, and a vehicle approaching from the left side of the screen is always a vehicle approaching from the intersection south road. The control unit 4 detects a moving object based on the distance image, and detects a priority road traveling vehicle from the upper quarter region 50 and a non-priority road traveling vehicle from the left quarter region 54. The controller 4 determines that the vehicle is approaching the intersection if the positional relationship between the two vehicles approaches the center of the reflected image and the distance between the two vehicles also decreases each time the vehicle follows the frame. Then, "attention" information is notified to vehicles traveling on the priority road, and "stop" information is reported to vehicles traveling on the non-priority road.

In order to notify the generated information to a vehicle or a pedestrian, the information can be displayed as visual information such as character information on an electric bulletin board or the like, or transmitted as voice information such as a loudspeaker or a siren. The communication unit 6 transmits and transmits information to vehicles and pedestrians using, for example, a road-vehicle communication beacon or a pedestrian beacon. FIG. 8 is a schematic diagram illustrating a road-to-vehicle beacon arranged at an intersection, and is a schematic diagram when the intersection is viewed from above. Vehicle 80 traveling from east to west
And the vehicle 82 traveling from the south to the north respectively have reached a position just before the intersection. The road-to-vehicle beacon is installed in front of an intersection or the like, and transmits information to a nearby vehicle. For example, in the figure, the beacon 84 provided in front of the intersection of the eastern road transmits information to the vehicle 80 coming from the east side when approaching the vehicle, while the beacon 84 provided in front of the intersection of the southern road. 86 transmits information to the vehicle 82 coming from the south side when approaching the vehicle. Thus, the beacon can transmit information to a nearby vehicle or pedestrian transceiver. Also, for example, the transceiver may transmit its own identification information or the like in response to an inquiry from the beacon, and the beacon may receive the information.

As a method of distributing the information generated for each direction, the position of the vehicle traveling on the road is obtained from the coordinates in the image. Therefore, for example, for vehicles traveling on the eastern road,
Information such as the position and speed of the vehicle traveling on the south road is transmitted to the vehicle traveling on the south road, and information such as the position and speed of the vehicle traveling on the east road is transmitted to the vehicle traveling on the south road. If there is a pedestrian approaching the intersection, the pedestrian is sent information on the position and speed of the vehicle from each direction approaching the intersection. Send

Next, each processing will be described in detail. FIG.
FIG. 9 is a process flowchart of a reference image generation process S60. The operator selects a state in which there is no movable object in the monitoring area of the wide-angle distance measuring device 2, that is, a state in which only the existing object is reflected on the conical mirror, and starts the reference image generation processing. Thus, a distance image in a state where no moving object is present is captured (S90). The captured distance image is stored in the control unit 4 as a reference image representing a known state (S9).
2).

FIG. 10 is a detailed processing flowchart of the moving object detection processing S62. When the processing is started, the control unit 4 reads a reference image stored in a storage device such as a magnetic disk, for example, on a memory (S100). In addition, the wide angle distance measuring device 2 is controlled to acquire a current distance image (S102), and a difference value between the distance image and the reference image is calculated (S104). FIG. 11 is a schematic diagram of an image showing an example of a process for obtaining the difference value. FIG. 11A is a reference image in which only an existing object is shown. Here, an intersection 110 is shown as the existing object. FIG. 11B shows a current distance image, in which distance images at times t ₁ and t ₂ (t ₂ > t ₁ ) are shown in an overlapping manner. The images 112 and 114 respectively show the time t _{1 of} a certain vehicle approaching the intersection from the south side,
which is an image in t _2. FIG. 11C shows a difference distance image between the current distance image and the reference image. The image of the intersection 110, which is an existing object, disappears, and only the images 112 and 114 of the vehicle remain.

The area of the moving object is detected based on the difference distance image (S120). The detected area of the moving object is sequentially labeled (S122), and the actual distance and azimuth of the area are calculated from the coordinates in the image (S12).
4). This labeling and position calculation are performed for all the detected movements. When the number of areas of the moving object that has not been labeled becomes 0 (S126), the attribute of the moving object corresponding to each moving object area is determined. The moving object type discriminating process S128 and the moving direction / speed detecting process S130 for detecting the moving direction and speed of the moving object are performed, and then the moving object detection process ends.

FIG. 12 shows a moving object type discriminating process S128.
It is a processing flow figure of. The moving object area label is initialized to 0 (S140). In the following, processing is performed in which attention is paid to the moving object area of the designated moving object area label, and the attribute of the corresponding moving object is determined (S142 to S152). This process is repeated for all moving objects (S156), starting from the initial value 0 and incrementing in order (S154).

First, the maximum width d and the maximum length 1 of the moving object area for each level number are obtained from the attributes of the area, respectively (S142, S144). FIG. 13 is a schematic diagram for explaining a method for obtaining the maximum width and the maximum length of the region, and shows a difference distance image. In this difference distance image, an image of an intersection which is an existing object is indicated by a dotted line for convenience. The moving object area 160 corresponds to the designated label number, and is located here on the eastern road.
In a simple method, the width d of the moving object region is a size along the width direction of the road where the region is located, and the length l of the moving object region is along the direction in which the road where the region is located extends. Size can be defined. Here, since the wide-angle distance measuring device 2 is arranged at the center of the intersection, the road width direction coincides with the circumferential direction around the center of the image, and the width d of the moving object region is the azimuth θ of the region. Is associated with the difference d 'between the maximum value and the minimum value, and the length l of the moving object area is associated with the difference l' between the maximum value and the minimum value of the distance of the area from the image center. The width d 'on this screen,
The length l ′ varies according to the distance r from the center of the screen due to the nature of the conical mirror even for the same moving object. further,
The conversion magnification according to the distance r differs between the azimuth direction and the radial direction of the conical mirror. This conversion magnification can be known from the shape of the conical mirror. The controller 4 corrects the width d 'and the length l' in consideration of the conversion magnification, and obtains the width d and the length l of the moving object area.

The control unit 4 compares the width d and the length l of each moving object region with a predetermined threshold value, and determines whether the moving object region corresponds to a vehicle, a motorcycle, or a pedestrian. Is determined (S146), and among the moving object attributes, the type of the moving object is defined as one of “four-wheeled vehicle”, “two-wheeled vehicle”, and “pedestrian” (S148 to S152). Discrimination processing S
In 146, for example, 2 for each of the width d and the length l
The determination is performed using the three thresholds. Here, two thresholds for the width d are α, A (α <A), and two for the length l.
Assuming that the three thresholds are β and B (β <B), d> A and l>
B is “four-wheeled vehicle”, α <d <A and β <l <
The case of B is determined as “motorcycle”, and the case of d <α and l <β is determined as “pedestrian”.

FIG. 14 shows a process of calculating the position of a moving object S12.
FIG. 4 is a processing flowchart of No. 4; The moving object label determined in the processing S122 is set, and the moving object area corresponding to the label is set as the processing target (S170). A point having the minimum angle (azimuth angle) from a predetermined reference generating line is obtained from the end of the moving object area, and the minimum angle at that time is calculated (S17).
2). In addition, a point having the maximum angle (azimuth angle) from a predetermined reference bus is obtained from the end of the moving object area, and the maximum angle at that time is calculated (S174). Then, an average value (angle average value) of these minimum angle and maximum angle is obtained (S
176). Next, a point having the minimum distance (radius) from the center of the conical mirror is obtained from the end of the moving object area, and the minimum radius at that time is calculated (S178). Further, a point having a maximum distance (radius) from the center of the conical mirror is obtained from the end of the moving object area, and the maximum radius at that time is calculated (S18).
0). Then, an average value (radius average value) of the minimum radius and the maximum radius is obtained (S182). The elevation angle ψ corresponding to the radius average value based on the conical mirror surface function of the conical mirror 30.
(S184), and the distance from the center of the intersection of the moving object on the road surface is calculated from the elevation angle ψ (S18).
6). On the other hand, the azimuth angle θ is obtained from the angle average value and the conical mirror installation condition (S188), and the azimuth of the moving object on the road surface is calculated from the azimuth angle θ (S190). The position represented by the distance and the direction determined in this way is stored as the moving object attribute (S192).

FIG. 15 shows a moving direction / speed detecting process S13.
FIG. 11 is a process flowchart of a main routine of No. 0. When the processing is started, the difference distance image calculated in the previous frame and stored in the storage device is read into a predetermined work area of the program (S200). Also, the difference distance image calculated in the current frame is stored in the predetermined work area (S20).
2). Then, a moving object region tracking matching process S204 for obtaining a moving object region having a correlation between the previous frame and the current frame, that is, a region of a certain moving object in the previous frame and a region in the current frame is performed. Based on a change in the position of the region with respect to the frame, detection of a movement direction S206 and detection of a movement amount (speed) S208 are performed.

FIG. 16 is a flowchart of the tracking matching process S204. By this processing, where a certain moving object in the differential distance image of the previous frame has moved in the differential distance image of the current frame, or a new object has occurred in the image (that is, a new monitoring of the wide-angle distance measuring device 2 has been performed). The presence or absence of a moving object that has entered the area and the presence or absence of a moving object that has disappeared from the image (that is, has moved out of the monitoring area of the wide-angle distance measuring device 2) are detected. As an initial setting, an unmatched attribute indicating that the moving object detected in the previous frame is not associated with the moving object in the current frame is added to all of the moving objects detected in the previous frame (S220). The label number n of the moving object detected in the image is set to an initial value 0 (S222). Also, the label number m of the moving object detected in the difference distance image of the previous frame is set to an initial value 0 (S224). A matching test is performed between the set moving object area with the label number m of the previous frame and the moving object area with the label number n of the current frame (S226). If they match, the label number m of the moving object of the previous frame is added to the moving object of the current frame as the temporary label number n '(S228), while the unmatched attribute of the moving object of the previous frame is deleted (S230). If the label number n of the moving object in the current frame is not the final number n _MAX (S232), it is incremented and the process returns to step S224 (S234).

If there is no match due to the matching (S226), the label number m of the moving object of the previous frame, which is mismatched here, is not the final number (S23).
6) This is incremented (S238), and matching processing with the newly set moving object of the previous frame is performed (S226).

Even if the label number m of the moving object in the previous frame reaches the final number, if there is no matching object with the moving object with the label number n in the current frame (S236), the moving of the current frame in the matching process is performed. The object is a new moving object that did not exist in the previous frame, and (n _MAX +1) is added as the provisional label number n ′ (S24).
0). Then, control goes to a decision process S232. Judgment process S
In 232, when it is detected that the last number n _MAX of the label number of the moving object in the current frame is detected, the presence or absence of the corresponding moving object in the previous frame is checked for all the moving objects in the current frame. Will be. In this case, subsequently, a process for organizing the label numbers is performed.

The label number arrangement processing is started by setting the label number m of the moving object in the previous frame to the initial value 0 (S
250). When the unmatched attribute is added to the moving object with the label number m of the set previous frame (S2
52), it means that there is no corresponding moving object in the current frame, that is, it has moved out of the monitoring area, and in this case, the label number m for the current frame is set to a missing number (S254). The inspection of the unmatched attribute and the process of missing numbers are repeated by sequentially incrementing the label number m of the moving object in the previous frame (S256). After the processing is completed for all the moving objects of the previous frame (S258), the temporary label number n ′ temporarily added to each moving object of the current frame is replaced with a regular label number n (S260). ) The tracking matching processing S204 ends. In the following processing,
Processing relating to the missing label number n is skipped.

FIG. 17 is a processing flowchart of the moving direction detection processing S206. This processing is started by setting the label number n of the moving object to be processed to the initial value 0 (S27).
0). The angle average value of the moving object having the set label number n in the previous frame and the angle average value in the current frame are respectively extracted from the storage device (S27).
2, S274), and the difference value (angle average value difference) is calculated (S276). The azimuth fluctuation amount is calculated from the angle average difference (S278), and the moving azimuth from the existing azimuth in the previous frame is obtained (S280). This processing S
The steps 272 to S280 are repeated by sequentially incrementing the label number n (S282). When the processing is completed for all the label numbers n (S284), the movement direction detection processing ends.

FIG. 18 shows the movement amount (speed) detection processing S208.
It is a processing flow figure of. This processing is started by setting the label number n of the moving object to be processed to the initial value 0 (S
290). The radius average value of the moving object having the set label number n in the previous frame and the radius average value in the current frame are respectively fetched from the storage device (S2).
92, S294), their difference value (radius average value difference)
Is calculated (S296). The moving distance is calculated from the radius average value difference (S298), and the moving speed is calculated based on the inter-frame time difference (S300). This processing S292
Steps S300 to S300 are repeated by sequentially incrementing the label number n (S302). When the processing is completed for all the label numbers n (S304), the movement amount (speed) detection processing ends.

FIG. 19 is a processing flow chart of the collision possibility determination processing S64. With this processing, it is determined whether to transmit the warning information. In order to perform processing by focusing on a moving object near an intersection, a label number of a moving object within a predetermined radius from the center of the intersection is searched for all moving objects, and position information of the moving object is read from a storage device. (S310 to S316). The label number i of the moving object to be compared is set to the initial value 0 (S318). Further, the label number k of the moving object to be compared is set to the initial value (i + 1) (S320). The distance between the set moving object to be compared and the moving object to be compared is calculated (S322), and the distance in this frame and the distance between the same two moving objects calculated in the previous frame and stored in the storage device are calculated. Are compared with each other (S324). As a result of this comparison, if the current distance is smaller than the previous distance (S326), it is determined that there is a danger of collision (S328), and an instruction to give a collision prevention warning is issued (S330). The above processing S322 to S330
Are sequentially repeated by incrementing the label number k of the moving object to be compared (S332). Also, the label number k
Reaches the maximum value (S334), the label number i of the moving object to be compared is sequentially incremented (S33).
6) When the label number i reaches the maximum value (S338), the collision possibility determination processing S64 ends.

The case where this system is installed at an intersection with poor visibility has been described above. In this case, the present system monitors a wide area from the information on the intersection, collects information on approaching vehicles, pedestrians, and bicycles in the monitored area, and predicts a danger such as a collision and issues a warning. This prevents a collision accident that occurs due to a lack of a traffic light at an intersection or a stoppage violation, and a collision accident that occurs due to a blind spot from the driver's seat.

As another installation example of the present system,
Curved roads in the mountains can be mentioned. FIG. 20 is a schematic diagram of an embodiment in which the present system is installed on a curve in a mountainous area. On the inner side 402 of the curved road 400, there are objects such as cliffs and mountains that obstruct the view, and two oncoming vehicles 404, 40
The six cannot see each other until they approach each other. In such a case, the wide-angle distance measuring device 2 is installed at the center of the curve where both sides of the curve can be seen.
And the vehicle 4 just before the curve reflected on the wide-angle distance measuring device 2
04, 406, and the presence thereof is detected, and a warning device 408, 410 such as an electric bulletin board provided near the entrance of the curve.
To notify the other driver. For example,
In this system, it is possible to detect and notify not only that the oncoming vehicle is approaching, but also what kind of vehicle is the oncoming vehicle. That is, it is possible to determine whether the vehicle is a small vehicle or a large vehicle based on the distance and the apparent size, and it is also possible to determine the moving speed. Thus, for example, the warning device 410 can display and warn the vehicle 406 that a large vehicle is approaching, and the warning device 408 can provide the vehicle 404 with information about the approach of a high-speed small vehicle as information. .

Another example of installation of the present system is a three-dimensional self-propelled parking lot. Collisions between running vehicles or between running vehicles and pedestrians are likely to occur in the multi-story parking lot due to blind spots caused by parked vehicles and low illuminance due to being indoors. Can monitor a wide area and prevent accidents.

[0055]

According to the traffic information providing system of the present invention, it is possible to acquire the traffic information in the entire space area by the wide-angle distance measuring device using the conical mirror and to transmit it to the pedestrian. A single installation can cover a wide area, and is easy to install and operate. The measuring device measures the distance to each point, and based on the measured value, the unevenness of the monitoring area is directly grasped, and the target object can be detected with high accuracy, thereby providing highly reliable traffic information. In addition, the peripheral area of the entire circumference of the measuring instrument is defined as one image, and the processing is performed based on the image, whereby an effect of facilitating the approach determination processing for the vehicle, the pedestrian, and the like can be obtained.

[Brief description of the drawings]

FIG. 1 is a schematic block diagram showing the overall configuration of a traffic information providing system according to an embodiment of the present invention.

FIG. 2 is a schematic overall configuration diagram of a wide angle distance measuring device.

FIG. 3 is a schematic view of an object point on a road surface and an image point on a conical mirror as viewed from above.

FIG. 4 is a schematic view of an object point on a road surface and an image point on a conical mirror as viewed from the side.

FIG. 5 is a schematic diagram illustrating an example of a relationship between an actual position of an object on a road surface and an image position on a reflection image of a conical mirror;

FIG. 6 is a schematic processing flowchart of the present system.

FIG. 7 is a schematic diagram illustrating an example of a relationship between an actual position of an object on a road surface and an image position on a reflection image of a conical mirror, for explaining an outline of determination of a possibility of collision.

FIG. 8 is a schematic diagram illustrating a road-to-vehicle beacon arranged at an intersection, and is a schematic diagram when the intersection is viewed from above.

FIG. 9 is a process flowchart of a reference image generation process.

FIG. 10 is a process flowchart of a moving object detection process.

FIG. 11 is an explanatory diagram showing an example of a process for obtaining a difference distance image.

FIG. 12 is a processing flowchart of a moving object type determination process.

FIG. 13 is a schematic diagram illustrating a method for obtaining a maximum width and a maximum length of a moving object area.

FIG. 14 is a processing flowchart of a position calculation processing of a moving object.

FIG. 15 is a processing flowchart of a main routine of movement direction / speed detection processing.

FIG. 16 is a flowchart illustrating a tracking matching process.

FIG. 17 is a processing flowchart of a moving azimuth detecting process.

FIG. 18 is a processing flowchart of a movement amount (speed) detection process.

FIG. 19 is a processing flowchart of a collision possibility determining process.

FIG. 20 is a schematic view of an embodiment in which the present system is installed on a curve in a mountainous area.

[Explanation of symbols]

2 Wide angle distance measuring device, 4 control unit, 6 communication unit, 8
Warning unit, 20 optical system, 22 variable focus lens unit control unit, 24 image processing unit, 30 conical mirror, 32 main lens, 34 variable focus lens unit, 36 electronic camera.

──────────────────────────────────────────────────の Continued on the front page (51) Int.Cl. ⁷ Identification symbol FI Theme coat ゛ (Reference) G08G 1/04 G08G 1/04 D // G01C 3/06 G01C 3/06 PF term (Reference) 2F065 AA03 AA06 BB05 BB15 CC11 CC16 FF10 JJ03 JJ09 JJ26 LL10 LL19 QQ13 QQ24 QQ25 QQ36 QQ39 RR07 SS09 SS13 UU02 UU05 2F112 AB05 AB10 BA03 BA18 CA20 DA32 DA40 FA03 FA21 FA36 FA38 FA45 GA10 5H180 EA04 CC07 CB DA02 DA03 FA32 FA53 FA66 FA67 FA69 GA08 GA34 HA03

Claims (5)

[Claims] 1. A method of measuring a distance to an object in the peripheral area based on a reflection image of the peripheral area acquired using a conical mirror, and generating a distance image corresponding to the peripheral area reflected in the reflection image. A wide-angle distance measuring device, a target detecting means for detecting a predetermined target based on the distance image, a passer notifying means for notifying the passer approaching the target of the detection information of the target, A traffic information providing system comprising: 2. The traffic information providing system according to claim 1, further comprising a pedestrian detection unit that detects the pedestrian based on the distance image, wherein the pedestrian notification unit is notified to the pedestrian. A traffic information providing system, wherein the detection information is generated according to a positional relationship between the object and the pedestrian. 3. The traffic information providing system according to claim 2, wherein the wide-angle distance measuring device is arranged so that a central axis of the conical mirror is directed to a collision attention point, and the wide-angle distance measuring device is provided on the reflected image viewed from the central axis. Based on the detection result of the target object detection unit and the passerby detection unit for the time series of the distance image, the passerby notification means, based on the detection result of the target object and the A traffic information providing system, comprising: a collision danger detecting unit that detects that a passer-by and a passer-by approach each other at the center of the distance image and notifies the passer-by of a fear of a collision. 4. The traffic information providing system according to claim 1, wherein the distance image generated in a state where the target object does not exist is stored as a reference distance image. The traffic information providing system, further comprising: comparing the distance image measured as needed with the reference distance image to detect the target object. 5. The traffic information providing system according to claim 2, wherein the distance image generated in a state where the object and the pedestrian do not exist is held as a reference image. The target object detection means detects the target object by comparing the distance image measured as needed with the reference distance image, and the pedestrian detection means calculates the distance image measured as needed. Detecting the pedestrian by comparing with the reference distance image.