Vehicle detection device

Abstract

The present invention provides a vehicle measuring device that can convert a video signal input from an imaging unit into an on / off signal indicating the passage of a vehicle and output it. [Solution] A planarization process is performed in real time over time to flatten a vehicle image F with a width w that occupies a portion of the road width, and at the same time, the total vehicle width ws of the vehicle image F is aggregated. Simultaneously with these planarization and aggregation processes, the total vehicle width ws, that is, the area of ​​the vehicle image F, is integrated. When the integrated value g exceeds a preset threshold d1, it is determined that one vehicle has been detected, and a pulse signal p consisting of an on / off waveform is output. When the integrated value g exceeds the threshold d1, the integrated value g is reset to zero, and the integration process of the area of ​​the vehicle image F is repeated.

Description

Technical Field

[0001] The present invention relates to a vehicle detection device that detects a traveling vehicle traveling on a road.

Background Art

[0002] Patent Document 1 discloses a road traffic information providing system including a plurality of road sensors and a road traffic control center connected to the road sensors via a network.

[0003] The road sensor is an ultrasonic vehicle detector, and transmits road traffic information such as the traffic volume of vehicles traveling on the road, that is, the number of vehicles passing per unit time, the average traveling speed, and the time occupancy, to the road traffic control center.

[0004] The ultrasonic vehicle detector calculates the traffic volume and road traffic information by an ultrasonic detector processing unit that receives a pulse signal composed of an on-off waveform from the head unit. Then, the central processing unit that has acquired the traffic volume and the road traffic information can set various constants for lighting control of traffic signal lights based on these pieces of information.

[0005] In recent years, an image detector having an imaging unit near the road is installed as a road sensor, and the traffic volume and average traveling speed of traveling vehicles are calculated by performing image analysis.

Prior Art Documents

Patent Document

[0006]

Patent Document 1

Summary of the Invention

Problem to be Solved by the Invention

[0007] However, if the aforementioned image detector is newly installed in place of conventional ultrasonic vehicle detectors, it is not possible to output a pulse signal consisting of on / off waveforms to detect vehicle passage, making it impossible to connect to the existing ultrasonic detector processing unit. Consequently, it is difficult to transmit information such as traffic volume to the central processing unit.

[0008] The object of the present invention is to solve the above-mentioned problems and provide a vehicle detection device that converts the video signal obtained from the imaging unit into an on / off signal indicating the passage of a vehicle, and that can be connected to an existing ultrasonic sensor processing unit. [Means for solving the problem]

[0009] A vehicle detection device according to the present invention for achieving the above objective comprises an imaging unit that photographs vehicles traveling in the same direction in one or more lanes, and a calculation unit that outputs a pulse signal consisting of an on / off waveform based on image information obtained from the imaging unit, wherein the calculation unit performs image processing on the image information, with the vertical axis being the road width of the lane and the horizontal axis being the passage of time, and flattens the vehicle image of the vehicle width of the vehicle occupying a part of the road width over time, integrates the total vehicle width of the flattened vehicle image over time, determines that one vehicle has been detected when the integrated value exceeds a predetermined threshold, outputs the pulse signal, resets the integrated value to zero, and repeats the integration process. [Effects of the Invention]

[0010] According to the vehicle detection device of the present invention, it is possible to convert the video signal obtained by the imaging unit into an on / off signal indicating the passage of a vehicle and output it to a signal processing unit including an ultrasonic sensor processing unit. By using the video signal obtained by the imaging unit, it is possible to prevent missed detections of moving vehicles, and even for vehicles traveling across lanes, the system can process them by dividing them according to the proportion of the vehicle width for each lane, so that vehicles traveling in a lane are not counted extra. [Brief explanation of the drawing]

[0011] [Figure 1] This is a block circuit diagram of the vehicle detection device. [Figure 2] This is an explanatory diagram of an existing ultrasonic vehicle detector. [Figure 3] This is an explanatory diagram of an existing ultrasonic vehicle detector. [Figure 4] This is a diagram illustrating the imaging area of ​​the imaging unit. [Figure 5] This is an explanatory diagram showing how the video signal from the imaging unit is planarized as a vehicle image when the vehicle is traveling at high speeds. [Figure 6] This is an explanatory diagram of the conversion process that converts a vehicle image into a pulse signal consisting of on / off waveforms. [Figure 7] This is an explanatory diagram showing how the video signal from the imaging unit is flattened into a vehicle image when the vehicle is traveling at a slow speed. [Figure 8] This is an explanatory diagram of the conversion process that converts a vehicle image into a pulse signal. [Modes for carrying out the invention]

[0012] The present invention will be described in detail based on the illustrated embodiments. Figure 1 is a block circuit diagram of the vehicle detection device, which consists of an imaging unit 1 installed above the vicinity of intersection K, and a processing unit 2 connected to the imaging unit 1 via wireless or wired connection.

[0013] The imaging unit 1, which is installed above the lane L by a pillar or the like, uses a digital camera or the like to take pictures of one or more lanes L. Vehicles a are traveling in the same direction in lane L, and these vehicles a include not only automobiles but also motorcycles.

[0014] The calculation unit 2 performs image processing on the video signal from the imaging unit 1, and sequentially outputs a pulse signal p consisting of an on-off waveform to the conventional ultrasonic sensing processing unit M1 in the housing M. The ultrasonic sensing processing unit M1 transmits road traffic information such as the traffic volume of traveling vehicles a and the calculated average traveling speed va to the central processing unit C1 of the traffic control center C at predetermined time intervals, for example, every 5 minutes.

[0015] In addition to the existing ultrasonic vehicle detector D, the central processing unit C1 can be connected to a plurality of vehicle detection devices and the ultrasonic sensing processing unit M1 via a telephone line or the like as illustrated.

[0016] Figure 2 is an explanatory diagram for a case using the existing ultrasonic vehicle detector D, and the housing M accommodating the ultrasonic sensing processing unit M1 is fixed to a nearby utility pole or the like. The ultrasonic vehicle detector D emits ultrasonic waves b from the transmitting portion H1 toward the road surface S of the road from the head portion H installed directly above the road, and the presence or absence of the traveling vehicle a on the road can be determined based on the response time of the reflected wave of the ultrasonic wave b from the road surface S at the receiving portion H2.

[0017] That is, when the response time of the reflected wave is equal to or greater than a predetermined value, it is determined as "no detection" that the traveling vehicle a is not present below the head portion H, and when the response time is less than the predetermined value, it is determined as "detection present" that the traveling vehicle a is present below the head portion H, so that the horizontal axis represents time, and the presence or absence of the traveling vehicle a can be detected as the pulse signal p consisting of an on-off waveform.

[0018] The head portion H is arranged for each lane L, outputs a pulse signal p indicating the presence or absence of the traveling vehicle a for each lane L to the ultrasonic sensing processing unit M1, and the ultrasonic sensing processing unit M1 can calculate the traffic volume of traveling vehicles a corresponding to the detected number of pulse signals p and the average traveling speed va based on the received pulse signals p. Then, the ultrasonic sensing processing unit M1 outputs the calculated traffic volume and average traveling speed va to the central processing unit C1.

[0019] As shown in FIG. 3, the ultrasonic sensing processing unit M1 cannot acquire the pulse signal p indicating "sensed" via the head unit H for a traveling vehicle a1 traveling between the first lane L1 and the second lane L2, or a motorcycle a2 traveling along the roadside.

[0020] The ultrasonic sensing processing unit M1 can calculate the traveling speed v of the traveling vehicle a based on the traffic volume that is the number of passing vehicles in the lane L, the detection time of the pulse signal p, and the average vehicle length na. Then, by averaging the traveling speeds v of the traveling vehicles a within a predetermined period, for example 5 minutes, the average traveling speed va in the lane L can be calculated, and the traffic volume and the calculated average traveling speed va are transmitted to the central processing unit C1 at a predetermined time interval, for example every 5 minutes. Note that the ultrasonic sensing processing unit M1 may sequentially output the pulse signal p to the central processing unit C1, and the central processing unit C1 may calculate the traffic volume and the average traveling speed va.

[0021] In some cases, the traveling vehicle a may travel not along a plurality of lanes L but regardless of the lanes L, and further, when the proportion of motorcycles among the traveling vehicles a is high, the detection accuracy of the traveling vehicle a tends to decrease when the conventional ultrasonic vehicle detector D is installed.

[0022] FIG. 4 is an explanatory diagram of an imaging area E which is a part of the screen captured by the imaging unit 1, and the arithmetic unit 2 performs planarization image processing on the traveling vehicle a passing through this imaging area E, and generates a pulse signal P composed of an on-off waveform to be described later from this image-processed vehicle image F. The arithmetic unit 2 transmits this pulse signal to the ultrasonic sensing processing unit M1.

[0023] The set imaging area E is a region along the road width wr orthogonal to the lane L, and when the traveling vehicle a passes as illustrated, the vehicle width w of the traveling vehicle a at the dot portions in the imaging area E can be detected.

[0024] Figure 5 is an explanatory diagram showing how the video signal from the imaging area E of the imaging unit 1 is planarized as a vehicle image F by the calculation unit 2. The horizontal axis represents elapsed time t, and the vertical axis represents the road width wr of the lane L. The vehicle image F of the vehicle width w of the vehicle a occupying the road width wr is planarized over time.

[0025] The illustrated lane L is, for example, a three-lane system consisting of the first lane L1 to the third lane L3, and the rectangular vehicle image F is an individual visualization of a moving vehicle a with a vehicle width w. Note that while the vehicle image F is explained as rectangular for clarity, in reality it changes according to the vehicle width w.

[0026] In Figure 5, which has been processed by the calculation unit 2 with an elapsed time t set to, for example, 1 minute, there are multiple vehicle images F of moving vehicles a. Vehicle image F1, which has a large vehicle width w, represents an automobile, and vehicle image F2, which has a small vehicle width w, represents a motorcycle.

[0027] Figure 6 is an explanatory diagram illustrating the process of converting an image-processed vehicle image F into a pulse signal p consisting of on / off waveforms. To explain the case where only the central second lane L2 shown in Figure 5 is extracted and converted into a pulse signal p, the planarization process shown in the upper diagram, which flattens the vehicle image F with a width w occupying a portion of the road width wr of the second lane L2, is created in real time over time, while simultaneously accumulating the total vehicle width ws of the vehicle image F shown in the middle diagram.

[0028] Simultaneously with these planarization and aggregation processes, the total vehicle width ws is integrated, that is, the area of ​​the vehicle image F is integrated, as shown in the second diagram. Then, as shown in the third diagram, when the integrated value g exceeds a preset threshold d1, it is determined that one vehicle a has passed, and a pulse signal p consisting of an on / off waveform is output, as shown in the lower diagram.

[0029] When the integral value g exceeds the threshold d1, the integral value g is reset to zero as shown in the figure, and the integration process of the area of ​​the vehicle image F is repeated. Note that a pulse signal p1 is output only at the start of the first integration when the calculation unit 2 is started, regardless of the integral value g. Note that this initial pulse signal p1 can be ignored when calculating traffic volume over a long period of time.

[0030] In the upper diagram of Figure 6, vehicle image F detects 8 vehicles. However, for a vehicle a traveling across the gap between the first lane L1 and the second lane L2, or between the second lane L2 and the third lane L3, the integration is performed only over the vehicle width w included in the second lane L2. Therefore, instead of detecting 8 vehicles, the lower diagram shows 7 pulse signals p, and the output indicates that 7 vehicles have traveled.

[0031] By processing based on the proportion of vehicle width w included in lane L, vehicles a traveling across lanes L are not counted twice.

[0032] The calculation unit 2 is capable of calculating the pulse width pw of the pulse signal p. As a method for calculating this pulse width pw, it uses the travel speed v, which is calculated from the movement speed of the outer frame of the traveling vehicle a on the screen, which is detected at any time based on image information from the imaging unit 1, and is a well-known technique.

[0033] The calculation unit 2 calculates the pulse width pw in the ultrasonic sensing processing unit M1 by dividing the average vehicle length na used by the average driving speed va of multiple driving vehicles a calculated on the screen over a predetermined period of time. These calculated pulse widths pw are inversely proportional to the average driving speed va; they become shorter when the average driving speed va is fast and longer when the average driving speed va is slow.

[0034] The conversion process to convert to such a pulse signal p is explained using a vehicle image F with a predetermined time of 1 minute, but in reality, the conversion process is repeated continuously over a long period of time, and a pulse signal p consisting of on / off waveforms is output from the calculation unit 2 to the ultrasonic sensing processing unit M1 connected to it.

[0035] The above explanation describes the conversion process for the second lane, L2, but it can be similarly applied to the first lane, L1, and the third lane, L3. Furthermore, it is possible to apply the above conversion process to the total lane L, which is the sum of lanes L1 through L3.

[0036] Furthermore, by calculating traffic volume individually for each lane (L1 to L3), it becomes possible to obtain information such as the congestion level of the right-turn lane, and to set the illumination time of the right-turn arrow signal in the right-turn lane with greater precision.

[0037] Furthermore, the threshold d is set according to the average vehicle length na of vehicles a, including motorcycles. On roads where there is a high proportion of motorcycles and other vehicles with small vehicle lengths n and widths w, the average vehicle length na is shorter and the threshold d is set lower. Conversely, on roads where there is a low proportion of motorcycles and other vehicles, the average vehicle length na is longer and the threshold d is set higher.

[0038] Conversely, if there is a large proportion of trucks and other vehicles with large vehicle lengths n and widths w, the average vehicle length na will increase and the threshold d will be set higher. Conversely, if there is a small proportion of trucks and other vehicles, the average vehicle length na will decrease and the threshold d will be set lower. Instead of using a threshold d that changes according to the average vehicle length na, it is also possible to use an average vehicle width wa that correlates with the average vehicle length na, and change the threshold d according to the average vehicle width wa. These average vehicle lengths na and average vehicle widths wa are stored in advance for the roads being traveled.

[0039] Figure 7 is an explanatory diagram showing how the video signal from the imaging unit 1 to the imaging area E is planarized by the calculation unit 2 when the speed of the moving vehicle a is slower than that of the moving vehicle a shown in Figure 6. When the moving speed is slow, the vehicle image F is displayed with a longer horizontal axis, i.e., a longer time axis, when planarized.

[0040] Figure 8 is an explanatory diagram of the conversion process that converts the vehicle image F of the second lane L2 in Figure 7 into a pulse signal p consisting of on / off waveforms. The threshold d2 fluctuates inversely proportional to the driving speed v calculated from the image information from the imaging unit 1; the threshold d2 is higher when the driving speed v is slow, and lower when the driving speed v is fast.

[0041] Thus, the threshold d2 in Figure 8 is higher than the threshold d1 in Figure 6. For example, if the average driving speed va in Figure 8 is half the average driving speed va in Figure 6, then the threshold d2 will be twice the value of the threshold d1.

[0042] Furthermore, when changing the threshold d based on the travel speed v, it is preferable to use, for example, the average travel speed va of vehicles a passing over one minute, in which case the threshold d is automatically changed every minute.

[0043] In this way, the calculation unit 2 performs planarization, aggregation, and integration of the vehicle image F, and when the integrated value g exceeds a preset threshold d2, it determines that one moving vehicle a has been detected. Then, as shown in the lower figure, it outputs a pulse signal p.

[0044] As described above, since the threshold d changes depending on the average vehicle length na or average vehicle width wa traveling in lane L, it is also possible to create a matrix table with the vertical axis representing the increase in average vehicle length na or average vehicle width wa and the horizontal axis representing the decrease in average driving speed va, starting from the zero point, and extract and set the threshold d from this matrix table.

[0045] When this matrix table is adopted, the threshold d extracted will be highest when the average vehicle length na or average vehicle width wa is large and the average driving speed va is slow, and conversely, the threshold d extracted will be lowest when the average vehicle length na or average vehicle width wa is short and the average driving speed va is fast.

[0046] Thus, in the vehicle detection device according to the present invention, the image unit 1 obtains a video signal, which is then converted by the calculation unit 2 into a pulse signal p consisting of on / off waveforms indicating the passage of a vehicle, and output to the ultrasonic sensing processing unit M1.

[0047] In this way, by using the video signal obtained by the imaging unit 1, it is possible to prevent the detection of moving vehicles a from being missed. Furthermore, even for moving vehicles a traveling across lanes L, by dividing them according to the proportion of the vehicle width w for each lane L and processing them accordingly, it is possible to avoid counting extra moving vehicles a traveling in lane L. [Explanation of symbols]

[0048] 1. Imaging Unit 2 Arithmetic section a. Vehicles in motion d, d1, d2 thresholds E Imaging area F, F1, F2 vehicle images g integral value L lane M1 Ultrasonic Sensing Processing Unit na average vehicle length p, p1 pulse signal pw pulse width t elapsed time va Average driving speed w vehicle width wr road width wa average vehicle width ws Total vehicle width

Claims

1. A vehicle detection device comprising an imaging unit that photographs vehicles traveling in the same direction in one or more lanes, and a calculation unit that outputs a pulse signal consisting of on / off waveforms based on the image information obtained from the imaging unit, The calculation unit performs image processing on the image information, with the vertical axis representing the road width of the lane and the horizontal axis representing the passage of time, to flatten the vehicle image of the vehicle width of the vehicle occupying a portion of the road width as time progresses. A vehicle detection device characterized by integrating the total vehicle width of the planarized vehicle image over time, determining that one of the moving vehicles has been detected when the integrated value exceeds a predetermined threshold, outputting the pulse signal, resetting the integrated value to zero, and repeating the integration process.

2. The calculation unit calculates the average speed of the vehicles from the movement speeds of the multiple vehicles on the screen in the image information. The vehicle detection device according to claim 1, characterized in that the pulse width of the pulse signal is calculated based on the average driving speed and the average vehicle length stored in advance.

3. The vehicle detection device according to claim 1 or 2, characterized in that the threshold changes according to the average driving speed.

4. The vehicle detection device according to claim 3, characterized in that the threshold is changed according to the average vehicle length or a preset average vehicle width.

5. The vehicle detection device according to claim 4, characterized in that the threshold is extracted from a matrix table with a zero point as the starting point, where the vertical axis represents the increase in the average vehicle length or the average vehicle width, and the horizontal axis represents the decrease in the average driving speed.

6. The vehicle detection device according to claim 1 or 2, characterized in that the calculation unit outputs the pulse signal to an ultrasonic sensing processing unit that can connect to the head unit of an ultrasonic sensor.

Images