JPH0154759B2 - - Google Patents

Description

[Detailed description of the invention]

The present invention relates to an apparatus for measuring the types of moving objects such as automobiles and the traffic volume, such as the number of passing vehicles for each type of vehicle. The prior art detects the presence or absence of a moving object by emitting ultrasonic waves. Although this prior art can detect the presence or absence of a moving object, it cannot detect the speed of the moving object alone. Furthermore, when the density of vehicles is high, such as during traffic jams, it becomes difficult to detect each vehicle individually, resulting in large errors and poor vehicle detection resolution. Another prior art is the so-called Doppler radar, which uses the Doppler effect by emitting microwaves or the like to a moving object. Although this prior art technology can detect the speed of a moving object, the measured value is identified based on the level of the received signal, so false detection may occur depending on the reflection cross section of the moving object or the direction of the microwave beam. may occur. Furthermore, it is difficult to accurately adjust the sensitivity. Moreover, like the prior art technology using ultrasound described above, the resolution of vehicle detection is poor. An object of the present invention is to provide a traffic volume measuring device that can accurately measure the traffic volume of moving bodies. Another object of the present invention is to provide a traffic volume measuring device with improved vehicle detection resolution. The present invention comprises: a support 5; a detection means 6 supported by the support 5; a means 8 for processing the output from the detection means 6;
31, the support 5 supports the detection means 6 above the moving object 4 on the road on which the moving object 4 travels, and the detection means 6 detects an image of the moving object 4 within the detection area 9 of the road. An image is formed on the first spatial filter 13 from the spectroscopic means 11 via the first condensing lens 12, and an image of the separated light from the spectroscopic means 11 is formed on the second spatial filter 16 via the second condensing lens 15. and from the detection area 9 to the first and second spatial filters 1
The optical path lengths up to 3 and 16 are different from each other, and the first and second spatial filters 13 and 16 have a plurality of photoelectric conversion elements 18a and 18b adjacent to each other along the moving direction of the image of the moving body 4, Multiple viewing-limiting parts 19a, 1 that do not detect light
9b are arranged alternately, and each output selected from every other photoelectric conversion element 18a, 18b is given to a differential amplifier circuit 20 that derives the difference between the two outputs. The arithmetic processing means 8, 31 include first and second
This is a traffic volume measuring device characterized in that it performs arithmetic processing on the outputs of the respective amplifier circuits 20 of the spatial filters 13 and 16. In a preferred embodiment, the arithmetic processing means 8, 31
When the output frequency of the first spatial filter 13 is 1, the output frequency of the second spatial filter 16 is 2, and the difference in optical path length is l, the difference between the first condensing lens 12 and the moving body 4 is distance d
21 is calculated as d21=2/1-2·l to obtain the height H of the moving body 4. Further, in a preferred embodiment, the arithmetic processing means 8,
31, when the output frequency of the first spatial filter 13 is 1, the output frequency of the second spatial filter 16 is 2, the difference in the optical path length is l, and K1 is a constant, the velocity V of the moving body 4 is , V=1・2/1−2・K1・l. Furthermore, in a preferred embodiment, the arithmetic processing means 8, 31, when the output frequency of the first spatial filter 13 is 1, the output frequency of the second spatial filter 16 is 2, and the difference in the optical path length is l, Distance d between the first condensing lens 12 and the moving body 4
21 as d21=2/1-2・l to find the height H of the moving body 4, and based on the height H of the moving body 4, the height H
It is characterized in that the length L of the moving body 4 determined in advance is estimated and determined in accordance with the above. One embodiment of the invention is shown in FIG. A moving object 4 such as a car runs on a road 3 having driving lanes 1 and 2. Above the road 3 is a support 5.
Accordingly, detection means 6, 7 according to the invention are provided for each driving lane 1, 2, respectively. The outputs from the detection means 6 and 7 are calculated by a calculation device 8. In this way, the traffic volume for each driving lane 1 and 2 is measured. The detection areas 9 and 10 by the detection means 6 and 7 are optimally set according to the widths of the driving lanes 1 and 2. If the detection areas 9 and 10 are too narrow, errors in counting the moving objects 4 will occur. On the contrary, if the detection areas 9 and 10 are too wide, the moving object 4 will be detected redundantly by the detection means 6 and 7 corresponding to both the driving lanes 1 and 2. FIG. 2 is a system diagram showing a specific configuration of the detection means 6 or 7. An image of the support body 4 is formed from a half mirror 11 through a condensing lens 12 onto a spatial filter 13 . Further, the light reflected by the half mirror 11 passes through another mirror 14, a condensing lens 15, and is imaged on a spatial filter 16. Condensing lenses 12 and 15 have the same configuration. Spatial filters 13 and 16 have the same configuration. FIG. 3 is a plan view of the spatial filter 13.
This spatial filter 13 is arranged in the moving direction 1 of the moving body 4.
Photoelectric conversion elements 18a, 18b adjacent along 7
and limited field of view parts 19a and 19b that do not detect light.
are arranged alternately to form one set, and a plurality of such sets are arranged. These photoelectric conversion elements 1
8a, 18b and limited field of view parts 19a, 19b
The length along the running direction 17 is as shown in the third figure, and each set has a total length 2a. Differential amplifier circuit 2
0 inputs are given outputs selected from every other photoelectric conversion element 18a, 18b. The differential amplifier circuit 20 derives a signal having a level obtained by subtracting the signal level applied to one input from the signal level applied to the other input. Such a spatial filter 13 is well known to those skilled in the art as a binary polar fence-like reticle having a transmittance distribution shown in FIG. Such a spatial filter 13 has a one-dimensional narrow band pass characteristic with respect to the traveling direction 17, and can extract a specific spatial frequency component from the irregular radiance distribution of the moving body 4. This spatial frequency component is converted into the frequency of an electric signal by the photoelectric conversion elements 18a and 18b, and a signal proportional to the traveling speed of the moving body 4 is obtained as the output of the differential amplifier circuit 20. As the photoelectric conversion elements 18a and 18b, photodiodes made of silicon or the like are suitable, for example. Instead of the rectangular reticle pattern shown in FIG. 3, it may be a double parallelogram reticle pattern as shown in FIG. 5, or a parallelogram reticle pattern as shown in FIG. It may be a reticle pattern. According to the reticle pattern shown in FIGS. 5 and 6,
This has the advantage that low-frequency sidelobe components can be reduced. Another spatial filter 16 also has a similar configuration. The frequency of the electrical signals output from the spatial filters 13 and 16 is set to 1+2, and the distances between the condenser lenses 12 and 15 and the image plane of the spatial filter are set to d11 and d12 as shown in FIG. The distance between the optical lens 12 and the moving body 4 is d21, the optical path length between the other condensing lens 15 and the moving body 4 is d22, the speed of the moving body 4 is V, and the distance between the spatial filters 13 and 16 is The width of the paired reticle is 2
a (see FIG. 3), the first and second equations hold true. 1=1/2a・d1/d21・V ...(1) 2=1/2a・d1/d22・V ...(2) In this example, d11=d12≡d1 ...(3) and d22 -d21=l...(4). According to another idea of the present invention, d11
≠d12. Therefore, d21=2/1-2・l...(5) V=1・2/1-2・l/k...(6) However, k=d1/2a...(7) Fifth According to the equation and the sixth equation, it can be seen that the distance d21 to the moving object 4 and the speed V of the moving object 4 can be obtained. Further, the height of the moving body 4, that is, the vehicle height H, is determined from FIGS. 1, 2, and Equation 5. H=h-2/1-2l (8) Here, h is the distance from the surface of the driving lanes 1 and 2 to the condenser lens 12. FIG. 7 is a block diagram of the electrical circuitry associated with the detection means 13 and 16. The output from the differential amplifier 20 included in the detection means 13 is transmitted to the amplifier circuit 22.
The signal is amplified by a bandpass filter 23, noise is removed by a bandpass filter 23, and the signal is applied from a line 24 to a pulse shaping circuit 25 that shapes it into a rectangular wave and a circuit 26 that performs level discrimination. Similarly, the output from the detection means 16 is amplified by an amplifier circuit 27, noise is removed by a band-pass filter 28, and is applied from a line 29 to a pulse shaping circuit 30 for shaping into a rectangular wave. Outputs from the pulse shaping circuits 25, 30 and the level discrimination circuit 26 are given to a processing circuit 31 including a microprocessor, where information on traffic volume is processed. FIG. 8 is a block diagram showing a specific configuration of the level discrimination circuit 26. The signal from bandpass filter 23 via line 24 has, for example, the waveform of FIG. 91. This signal is transmitted to the detection circuit 3
2 performs envelope detection of only one polarity, either positive or negative, for example, only of positive polarity in this embodiment. This envelope detection output is given to one input of the comparison circuit 33. Comparison circuit 3
The other input of No. 3 is given an output from a setting circuit 34 that sets the discrimination level. Comparison circuit 33
derives a high level signal as shown in FIG. 92 only during the period in which the output from the detection circuit 32 exceeds the discrimination level set by the setting circuit 34. The operation shown in FIG. 7 will be explained in more detail with reference to FIG. 10. From the spatial filter 13 to the amplifier circuit 2
2, the noise is removed by a bandpass filter 23, and the signal is led out to a line 24, which has a frequency 1 and has a waveform, for example, as shown in FIG. 10. The pulse shaping circuit 25
4 is shaped into a rectangular wave to derive the rectangular wave shown in FIG. Period W1 of this square wave
corresponds to the time it takes for the image of the moving body 4 to pass through the length 2a of the spatial filter. Similarly, one more
The signal derived from the two bandpass filters 28 onto the line 29 has a frequency of 2 and a waveform as shown in FIG. 10(3). The pulse shaping circuit 30 shapes the signal on the line 29 into a rectangular wave to derive a signal having the waveform shown in FIG. This period W2 corresponds to the time required for the image of the moving body 4 to pass through the length 2a of the spatial filter included in the detection means 16. The reference clock signal used in processing circuit 31 has a high frequency, as shown in FIG. line 24
The output from the detection circuit 32 to which the signal is input is as shown in FIG.
This detected output is subjected to level discrimination based on the reference level l1 set in the setting circuit 34. As a result, a signal shown in FIG. 10 is derived from the comparison circuit 33. As a result, the output from the comparison circuit 33 is delayed from the detection signal derived from the detection circuit 32 by the times Td1 and Td2. In the processing circuit 31, pulse shaping circuits 25, 30
2, and receives the level-discriminated signal from the level discrimination circuit 26, and calculates the vehicle speed V of the moving body 4 based on the sixth formula,
Further, the height, that is, the vehicle height H of the automobile is determined based on the eighth equation. Corresponding to this vehicle height H, the vehicle length L of the mobile object 4 is estimated based on, for example, Table 1.

[Table] Furthermore, the number of passing moving objects 4 is counted based on the vehicle speed V and the estimated vehicle length L. The measurement accuracy of frequencies 1 and 2 is important because it determines the accuracy of the traffic measurement device. The detected frequencies 1 and 2 depend on the shape of the reticle pattern of the detection means 13 and 16, the dimensions of the optical system including the half mirror 11, the mirror 14, and the condensing lenses 12 and 15, and the vehicle speed of the moving body 4. Determined.
Since the outputs from the detection means 13 and 16 are transient signals that are generated only when the moving object 4 passes through the detection areas 9 and 10 of the driving lanes 1 and 2, the wave number is limited. Therefore, the cycles of the signals from the detection means 13 and 16, and therefore the outputs from the pulse shaping circuits 25 and 30, are measured over a plurality of n cycles, the average value thereof is determined, and the reciprocal of the average value of the cycles is calculated. A method that calculates the frequency with high accuracy is suitable. The inventor determined that, for example, about 10 would be appropriate for n.
In the illustrated embodiment, 1>2. The output from the level discrimination circuit 26 is the detection area 9
is a logic "1" or logic "0" pulse signal generated while the moving body 4 is moving. As another embodiment, the output from the bandpass filter 23 may be analog/digital converted and provided to the processing circuit 31, and the level may be discriminated by executing a program. Regarding the processing timing in the processing circuit 31, frequencies 1 and 2 are measured alternately. Parallel processing may be performed to shorten processing time. Note that frequencies 1 and 2 are multiple n as described above.
Assuming that the average value of the period is 1 and 2, the following equations are shown. 1=1/T1...(9) 2=1/T2...(10) Bandpass filter 23 shown in FIG.
11 and the waveform shown in FIG. 11 obtained by level discrimination by the level discrimination circuit 26 are input to the processing circuit 31. Frequencies 1 and 2 are measured during the first time Ta during which the moving object 4 passes through the detection area 9 or 10, and the vehicle speed V and vehicle height H are determined. There are two methods for the measurement timing of frequencies 1 and 2: The difference between these two methods lies in the timing at which the vehicle height H is determined, which is a problem mainly when the moving object 4 is a passenger car. FIG. 12 shows an enlarged view of the output waveform from the bandpass filter 23 at time Ta. As shown in FIG. 13, in a passenger car, the bonnet section 35, roof section 36, and rear section 37 have different vehicle heights Ha, Hb, and Hc. 37 and have lengths La, Lb, and Lc, respectively. One measurement timing method is to calculate the frequency based on the received signal from the tip of the moving body 4 to the length La.
Measure 1 and 2 and find the average height of the bonnet part 35.
Find Ha. In this case, the frequency n of the received signal is determined so as to fall within the transit time of the length La. Another measurement timing method calculates the average height Hb of the roof portion 36 based on the outputs from the band pass filters 23, 28 corresponding to the roof portion 36. In this case, the vehicle speed V is determined during the first period Ta1 (see FIG. 12) corresponding to the bonnet portion 35,
The passage time r of the bonnet part 36 is estimated and the time r is set, and after this time r has elapsed, the roof part 36 is
Measurement calculations for frequencies 1 and 2 are performed during a period of time Ta2 corresponding to . It is possible to identify passenger cars, buses, and trucks from each other using either of these measurement timing methods for frequencies 1 and 2, and
By using the two methods in combination, it is possible to identify vehicle types in even more detail. For example, the length La,
By measuring Lb and Lc, the cross-sectional shape of the moving body 4 can be recognized. The vehicle length L is calculated by comparing the vehicle height H calculated at time Ta with a predetermined threshold value.
is estimated based on Table 1 above. This commander L
The signal duration time T is determined from the vehicle speed V and the vehicle speed V as follows. With reference to FIG. 14, the traveling direction 17 of the moving body 4
If the length of the detection areas 9 and 10 is LD, the time T0 for the moving object 4 to pass through the detection areas 9 and 10 is
It can be found from Equation 11. At this time, the speed V of the moving body 4
is constant. T0=L+LD/V (11) When α is the processing time from when the moving object 4 is detected at the position 38 to when the vehicle speed V is calculated, the signal duration time T is expressed by Equation 12. Here, the processing time α does not necessarily match the time Ta. T=T0-α =L+LD/V-α...(12) After finding this signal duration time T, time 40 is determined based on a predetermined value m (however, m>1), when time T/m has elapsed. (See FIG. 11), it is determined whether the output from the level discrimination circuit 26 is logic "0" or not. When the output from the level-specific flight circuit 26 is not logic "0", that is, when it is logic "1", it is determined that the moving object 4 is passing through the detection areas 9 and 10. When the output from the level discrimination circuit 26 is logic "0", it is determined that the detection of the vehicle speed V and vehicle height H is erroneous. The value m is
Affects vehicle counting error. For example, this value m
By selecting an appropriate value for , it is possible to distinguish whether the moving object 4 is a passenger car or a binary vehicle. The process returns to the first process at time 41, which has elapsed by ΔT=(1-1/m)T+β (13) from time 40. Here, β is a safety constant and is a correction term due to the fact that the vehicle length L is used to calculate the time T. (1-1/m)T alone means that the received signal continues when the vehicle length L is shorter than it actually is, and there is a risk that unnecessary processing will be performed. Although it does not directly affect the processing, the level discrimination circuit 26
The output from remains a logic "1". A value β is introduced to remove such factors. β determines the resolution of the distance between the following vehicle and the following vehicle. Also this β
is appropriately given by Equation 13 as a function of vehicle speed V. β=Ls/V...(14) Here, Ls is given as a unit of distance. The operation of the processing circuit 31 will be explained with reference to FIG. When calculating the traffic volume in step n1, the process moves to step n2, and it is determined whether or not there is an output from the detection means 13. If there is an output, the process moves to step n3, and equations 9 and 10 are used. Measure frequencies 1 and 2 based on. step n4
Now, the speed V and the vehicle height H are calculated based on the sixth and eighth equations. In steps n5, n7, and n9, the vehicle height H is set to a predetermined value H based on Table 1.
Compare with 1, H2, and H3. In steps n5 to n9, the vehicle height H is compared with predetermined values H1 to H3, and as a result, in steps n6, n8, n10, or n=11, the vehicle length L is set to the estimated value L1, L2, or L3. , L4. In step n12, the signal duration time T is determined by Equation 12 using the vehicle length L estimated in this way.
Set. In step n13, time T/
There is a delay of m, and in step n14 it is determined whether the output R of the level discrimination circuit 26 is logic "0". In this way, steps n13 and n14
Then, it is determined whether the moving object 4 is a vehicle to be detected. If the output from the level discrimination circuit 26 is logic "0", it is determined to be an erroneous detection in step n18, and the process returns to step n2. When the output from the level discrimination circuit 26 is logic "1", the process moves to step n15, and information on traffic volume is output. In step n16, the value β is calculated based on the 14th equation. In step n17, time ΔT is calculated based on equation 13 to prevent erroneous detection of the following vehicle, and the process returns to step n12. When outputting traffic information as a serial 16-bit digital quantity, the contents are expressed as shown in Table 2, for example.

[Table] As described above, according to the present invention, two optical systems having different optical path lengths and a spatial filter formed by the optical systems are used, and the output of the spatial filter is used to detect a moving body. Since the traffic volume is measured by measuring the height, the traffic volume of moving objects can be accurately measured, and the resolution of vehicle detection is improved by using a spatial filter.

[Brief explanation of drawings]

FIG. 1 is an overall perspective view of an embodiment of the present invention, FIG. 2 is a system diagram showing a specific configuration of the detection means 6,
FIG. 3 is a front view of the spatial filter 13, FIG. 4 is a diagram showing the transmittance distribution of the spatial filter, and FIGS. 5 and 6 are diagrams showing the ray take pattern of the spatial filter of other embodiments of the present invention. , FIG. 7 is a block diagram of an electric circuit including the arithmetic unit 8, FIG. 8 is a block diagram showing a specific configuration of the level discrimination circuit 26,
9 is a waveform diagram for explaining the operation of the level discrimination circuit 26 shown in FIG. 8, FIG. 10 is a waveform diagram for explaining the operation of the electric circuit shown in FIG.
1 is a waveform diagram for explaining the idea of the present invention for counting the number of moving objects 4, FIG. 12 is an enlarged view of the output waveform from the bandpass filter 23, and FIG. 13 is a waveform diagram of a passenger car as the moving object 4. 14 is a side view, FIG. 14 is a plan view showing a state in which the moving body is continuously running, and FIG. 15 is a flowchart for explaining the operation of the processing circuit 31. 1, 2... Driving lane, 3... Road, 4... Moving object,
6, 7...detection means, 9,10...detection area, 11...
Half mirror, 12, 15... Condensing lens, 14...
Mirror, 13, 16... Spatial filter, 22, 27
...Amplification circuit, 23, 28...Band pass filter,
25, 30... Pulse shaping circuit, 26... Level discrimination circuit, 31... Processing circuit.

Claims (1)

[Claims] 1. A support 5, a detection means 6 supported by the support 5, and means 8 for calculating the output from the detection means 6.
31, the support 5 supports the detection means 6 above the moving object 4 on the road on which the moving object 4 travels, and the detection means 6 detects an image of the moving object 4 within the detection area 9 of the road. An image is formed on the first spatial filter 13 from the spectroscopic means 11 via the first condensing lens 12, and an image of the separated light from the spectroscopic means 11 is formed on the second spatial filter 16 via the second condensing lens 15. and from the detection area 9 to the first and second spatial filters 1
The optical path lengths up to 3 and 16 are different from each other, and the first and second spatial filters 13 and 16 have a plurality of photoelectric conversion elements 18a and 18b adjacent to each other along the moving direction of the image of the moving body 4, Multiple viewing-limiting parts 19a, 1 that do not detect light
9b are arranged alternately, and each output selected from every other photoelectric conversion element 18a, 18b is given to a differential amplifier circuit 20 that derives the difference between the two outputs. The arithmetic processing means 8, 31 include first and second
A traffic volume measuring device characterized in that it performs arithmetic processing on the outputs of the respective amplifier circuits 20 of the spatial filters 13 and 16. 2 The arithmetic processing means 8 and 31 calculate the first condensing lens 12 when the output frequency of the first spatial filter 13 is 1, the output frequency of the second spatial filter 16 is 2, and the difference in optical path length is l. distance d between and the moving body 4
21. The traffic volume measuring device according to claim 1, wherein the height H of the moving body 4 is determined by calculating d21=2/1-2·l. 3 The arithmetic processing means 8 and 31 move when the output frequency of the first spatial filter 13 is 1, the output frequency of the second spatial filter 16 is 2, the difference in the optical path length is l, and K1 is a constant. The traffic volume measuring device according to claim 1, characterized in that the speed V of the body 4 is calculated as follows: V=1.2/1-2.K1.l. 4 The arithmetic processing means 8 and 31 calculate the first condensing lens 12 when the output frequency of the first spatial filter 13 is 1, the output frequency of the second spatial filter 16 is 2, and the difference in optical path length is l. distance d between and the moving body 4
21 as d21=2/1-2・l to find the height H of the moving body 4, and based on the height H of the moving body 4, the height H
2. The traffic volume measuring device according to claim 1, wherein the length L of the moving object 4 is estimated and determined in advance according to the length L of the moving body 4.