JP3534557B2 - Vehicle entry detection device - Google Patents

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a device for detecting entry of a vehicle at a tollgate of a parking lot or a toll road.

[0002]

2. Description of the Related Art A conventional technique is shown in FIGS. Figure 4
FIG. 1 is a structural diagram of a first example of a vehicle entry detection device according to a conventional technique. FIG. 5 illustrates a first example of a vehicle approach detection device according to the related art.
Explanatory drawing of the example process.

FIG. 6 is a structural diagram of a second example of a vehicle approach detection device according to the prior art. FIG. 7 is an explanatory diagram of a process example of a second example of the vehicle entrance detection device according to the related art.

In a parking lot or toll road, it is detected that a vehicle has entered the gate, and the entrance / exit gate is opened, or a toll ticket is issued, and the operation of the fare adjustment machine is instructed. It is necessary to output information to the host device such as.

As a vehicle approach detection device for this purpose, there are conventionally a system using an ultrasonic sensor, a system using an electromagnetic wave and a photoelectric sensor, and the image processing system will be described here.

FIG. 4 is a block diagram of a first example of a vehicle approach detection device according to the prior art, in which the image pickup device 1 is approaching the vehicle 10 in the approach direction of the lane 12 and the background 11.
Image. The captured image is digitized and stored in the image memory 2.

Further, the reference image memory 21 is stored in the vehicle 1 in advance.
An image in which only the background 11 when 0 does not exist is registered. The contents of the image memory 2 and the reference memory 21 are input to the difference calculation circuit 22, and the absolute value of the difference is calculated for each pixel of the two input images.

The output of the difference calculation circuit 22 is input to the center detection circuit 23, and the coordinates of the center of the aggregate area of pixels whose absolute value of the difference is larger than a preset threshold value are detected. The output of the center detection circuit 23 is input to the movement state determination circuit 24.

The moving state judging circuit 24 judges whether the input center position changes according to the traveling direction of the vehicle on the image, and passes one vehicle or newly enters another vehicle. To detect.

The output of the moving state determination circuit 24 is input to the host device 9 via the communication circuit 8 to notify the host device 9 that the entry of the vehicle has been detected. The operation of the device will be described in more detail with reference to FIG.

When the image of the vehicle is picked up by the image pickup device 1 from the diagonally right front side of the vehicle 10 as shown in FIG. 4, an image including the vehicle 10 and the background 11 is picked up as shown in FIG. 5A. At this time, on the image, the vehicle 10 advances from left to right. This input image is stored in the image memory 2.

Further, in the reference memory 21, an image in which only the background 11 as shown in FIG. 5B is picked up is registered in advance. The difference calculation circuit 22 obtains the absolute value of the difference between the contents of the image memory 2 and the contents of the reference memory 21 for each pixel.

Since the background 11 on the image has the same image, the difference becomes zero. However,
When the vehicle 10 is approaching, the densities of the pixels in that portion are changing, and when the difference is obtained, some value is output.

Here, when the coating color of the vehicle 10 is white,
In the case of black, the sign of the difference value is reversed. Therefore, by obtaining the absolute value, only the region of the vehicle 10 shows an output above a certain value, and only the region of the vehicle 10 is obtained as shown in FIG. 5C.

The center detection circuit 23 determines the center position of the vehicle 10 area as shown in FIG. 5 (c). Figure 5 shows the center position
Upper and lower ends j _min in the height direction of the vehicle 10 area shown in (c),
The midpoint may be obtained from j _max and the lateral left and right edges i _min and i _max .

The moving state determination circuit 24 detects passage of the vehicle from the water level state at the center position. That is, since the vehicle 10 travels from left to right on the image, the center position sequentially moves to the right from the time when it is first detected.

Therefore, it is assumed that the same vehicle 10 is in progress while the center position is moving to the right, and the passing is ended when the vehicle moves to the right end. In addition, when returning to the left, it is determined that a new vehicle 10 has entered.

When a new vehicle 10 is input, the communication circuit 8 outputs the information to the host device 9. The host device 9 opens and closes the gate, issues a pass ticket, and starts the operation of the fare adjustment machine. In the first conventional example, the difference calculation circuit 22 is used to detect the vehicle 10. However, in the case of the difference,
When the density of the entire image changes due to changes in sunshine conditions, all areas may be determined to be changed areas, that is, the vehicle 10.

Therefore, in the second conventional example shown in FIG. 6, the correlation calculation circuit 5 is used instead of the difference calculation circuit 22. Since other parts are the same as those in the first conventional example, description thereof will be omitted, and only the operation of the correlation calculation circuit 5 will be described with reference to FIG.

The correlation calculation circuit 5 divides the image memory 2 and the reference memory 21 into several rectangular areas in the same manner as shown in FIGS. 7 (a) and 7 (b). For each corresponding rectangular area of the image memory 2 and the reference memory 21, the normalized correlation value represented by the following equation is calculated.

Normalized correlation value = Σ (fg−μν) / σπN Equation (1) where, f: density data of each pixel in each rectangular area of the image memory 2, g: each density in each rectangular area of the reference memory 21 Pixel density data μ: average value of density data in each rectangular area of the image memory 2 ν: average value of density data in each rectangular area of the reference memory 21 σ: of density data in each rectangular area of the image memory 2 Standard deviation π: Standard deviation of density data in each rectangular area of the reference memory 21 N: Number of data in each rectangular area This is, for each rectangular area, the density is normalized by the standard deviation of the density to obtain a difference. Even if there is a change in the density of the entire image due to a change in the sunshine conditions, the area where the change in density distribution is small is a large correlation value, and the area where the change in density distribution is large is not affected by the change. Indicates a small correlation value. As a result, as shown in FIG.
Only the rectangular area including the 0 area shows a small correlation value, and the area of the vehicle 10 can be detected.

[0022]

However, the conventional techniques have the following problems. (1) In the vehicle approach detection device according to the related art, it is necessary to obtain the absolute value of the difference between all areas of the image memory or the normalized correlation value in order to detect the approach of the vehicle, which requires processing time. (2) Further, in order to grasp the progress state of the vehicle, the center position of the vehicle region on the image is confirmed by sequentially shifting from left to right. Incorrect detection may prevent accurate judgment. (3) Further, when a pedestrian or a vehicle in the adjacent lane enters the field of view of the image pickup device, that portion is also detected as a vehicle, and there is a possibility that malfunction may occur, which is unreliable. . The present invention aims to provide a device capable of solving these problems.

[0023]

[Means for Solving the Problems] (First Means) A vehicle entry detecting device according to the present invention is a device for detecting entry of a vehicle into a gate at a tollgate of a parking lot or a toll road. Image pickup device 1 for picking up an image in a direction, and (B) an image memory 2 for storing an image output from the image pickup device
And (C) a rectangular area cutout circuit 3 for cutting out a plurality of vertical or horizontal rectangular areas in the image memory, and (D) a rectangular area reference memory 4 for registering an image of the rectangular area when no vehicle is present. (E) Correlation calculation circuit 5 for obtaining the normalized correlation value of the rectangular area reference memory and the rectangular area of the original image output by the rectangular area cutout circuit; (F) For each rectangular area from the output of the correlation calculation circuit 5. 0 or 1
Bit pattern generation circuit 6 for obtaining the bit pattern of
And (G) an approach passage determination circuit 7 that determines the passage of the vehicle by considering one bit pattern string for each rectangular area output from the bit pattern generation circuit 6 and (H) the rectangular area. The cutout circuit 3 sets a plurality of rectangular areas substantially perpendicular to the passing direction of the vehicle on the image memory. Is 0 for no change,
If it is small, it is determined that there is a change, and 1 is given as a bit pattern. By treating as
It is characterized in that the approach and the passage of the vehicle are detected at high speed from the approach and passage pattern which is assigned in advance for each data. (Second Means) The vehicle entrance detection device according to the present invention is the first means. A rectangular area is set as the preliminary detection area RX, and (2) when a vehicle is detected only in the preliminary detection area RX, the entry detection signal is not output to the host device, and 2
It is characterized in that the reliability is improved by outputting the approach detection signal of the vehicle after the vehicle is detected also in the rectangular area after the second one.

That is, in order to solve the problems of the prior art, the present invention, in the present invention, an image pickup device 1 for picking up an image in the entrance direction, an image memory 2 for storing an image output by the image pickup device, A rectangular area cutout circuit 3 for cutting out a plurality of vertical or horizontal rectangular areas in the image memory, a rectangular area reference memory 4 for registering an image of the rectangular area when there is no vehicle, and contents of the rectangular area reference memory And a correlation calculation circuit 5 for obtaining a normalized correlation value of a rectangular area of the original image output from the rectangular area cutout circuit, and a bit pattern generation circuit 6 for obtaining a 0 or 1 bit pattern for each rectangular area from the output of the correlation operation circuit. The bit pattern generation circuit outputs the bit pattern sequence for each rectangular area, and the approach pass determination circuit 7 is used to determine one data and determine the approach pass of the vehicle.

Therefore, it operates as follows. With the above configuration, a plurality of rectangular areas are set substantially perpendicular to the vehicle passing direction on the image memory, and the rectangular areas are cut out from the image.

When the normalized correlation value in each rectangular area is large, 0 is set as no change, and when it is small, 1 is set as change and a bit pattern is created by the bit pattern generation circuit. The pattern string is treated as data having a bit length corresponding to the number of rectangular areas.

The approach passage determination circuit 7 can detect the entry and passage of the vehicle at high speed by allocating advancement passage patterns for each data in advance. Also, of the multiple rectangular areas, the rectangular area closest to the vehicle approach direction on the image is set as the preliminary detection area, and when a vehicle is detected only in the preliminary detection area, an entry detection signal is output to the host device. However, the reliability can be improved by outputting the vehicle entry detection signal when the vehicle is detected in the second and subsequent rectangular areas.

[0028]

DETAILED DESCRIPTION OF THE INVENTION

(First Embodiment) FIG. 1 shows a first embodiment of the present invention.
~ Shown in FIG. The components having the same functions as those of the conventional technique are designated by the same reference numerals, and the description thereof will be omitted.

FIG. 1 is a block diagram of a vehicle entry detection device according to a first embodiment of the present invention. FIG. 2 is an explanatory diagram of an image processing state according to a traveling state of the vehicle according to the first embodiment.

FIG. 3 is an explanatory diagram of the relationship between the detection bit pattern of the change area on the image and the traveling state of the vehicle according to the first embodiment. As shown in FIG. 1, the image including the vehicle 10 and the background 11 imaged by the imaging device 1 is digitized and stored in the image memory 2.

The contents of the image memory 2 are cut out into a plurality of rectangular areas by the rectangular area cutting circuit 3. The rectangular region to be cut out at this time is a vertically long region that is substantially perpendicular to the left-right direction, which is the traveling direction of the vehicle 10 on the image, as shown in FIG. 2, and a plurality of rectangular regions 13 are present. The width, that is, the width from the R ₀ area to the R ₄ area is set to be slightly smaller than the size of the vehicle on the image.

Further, the height of the rectangular area 13 may be as high as the height on the image of the passing vehicle 10, but in FIG. 2, it is set to the height of the image memory 2. The width of the rectangular area 13 may be determined in consideration of the target processing speed, and may be about several pixels.

Each rectangular area R ₀ , R ₁ , R ₂ , R ₃ , R ₄
Since the position and the size can be determined in advance, the rectangular area cutout circuit 3 is configured to read the contents of the pixel position stored in the storage element in advance.

The rectangular area reference memory 4 stores the rectangular area 1 when the vehicle 10 is not present, as shown in FIG.
Collect and memorize only 3. The output of the rectangular area cutout circuit 3 and the output of the rectangular area reference memory 4 are input to the correlation calculation circuit 5, and the respective rectangular areas R ₀ , R ₁ , R ₂ , R ₃ , R ₃ are input.
_The normalized correlation value is calculated for each ₄ .

By using the normalized correlation value, the change in the density of the entire image due to the change in sunshine is not affected. If the normalized correlation value of each of the rectangular regions R ₀ , R ₁ , R ₂ , R ₃ , and R ₄ output from the correlation calculation circuit 5 is larger than the threshold value set in advance, the bit pattern generation circuit 6 outputs the image. 0 is generated if there is no change, and 1 is generated if the image is changed if the change is small.

As a result, as shown in FIGS. 2B to 2D, a pattern in which only the rectangular region 13 becomes 1 when the vehicle 10 is on the road is obtained. In FIG. 2, the rectangular area 13 determined to be 1 is hatched.

The output of the bit pattern generation circuit 6 is input to the entry passage determination circuit 7. As shown in FIG. 3, the approach passage determination circuit 7 includes rectangular areas R ₀ , R ₁ , R ₂ , R ₃ ,
In this case, the bit pattern of R ₄ is treated as 5-bit data, and a numerical value represented by 5 bits (3 from 0 to 31)
2)) to determine the progress of the vehicle.

In FIG. 3, R ₀ is the minimum digit and R ₄ is the maximum digit, and each bit pattern is digitized. Numerical value 0 indicates that there is no vehicle 10, and numerical values 1, 3,
1 for bits from 7, 15, 31 and R ₀ to R _{4 in} order
The vehicle 10 enters in such a manner as to stand.

On the contrary, when the vehicle 10 passes, the numerical value is 30,
In order from 28,24,16 and R ₀ bits up R ₄ is back to zero. Here, since the width from the R ₀ area to the R ₄ area is set to be slightly smaller than the size of the vehicle on the image, the rectangular areas at both ends of R ₀ and R ₄ are ₀ as in the case of the numerical value 2 or the like. Despite being R ₁ to R ₃
There should be no state in which 1 is set to 1, and if such a bit pattern appears, it should be ignored because it is an erroneous detection.

When 1 stands from both ends of the rectangular area 13 as in the case of the numerical value 17, it is determined that the vehicle 10 traveling in front passes and the vehicle 10 following is entering. As described above, since a maximum of 32 driving states can be easily determined by considering a simple bit pattern, if the approach passage determination circuit 7 generates a bit pattern if it is registered in the storage element in advance, the bit pattern is generated. The traveling state of the vehicle 10 can be easily determined by reading the contents of the storage element as the address.

As a result of the determination, when the entry of the vehicle 10 is detected, the entry information is output to the host device 9 through the communication circuit 8. Further, in the passage determination circuit 7, 1 is set in the rectangular area R ₀ .
If the vehicle 10 is immediately determined to enter the vehicle and output the entry information, a problem will occur if the rectangular area R ₀ is erroneously detected. Therefore, the rectangular area R ₀ is set as the preliminary detection area. in only one stood in the rectangular area R ₀ is not determined that the entry of the vehicle 10, in both of the rectangular region R ₀ and the rectangular region R ₁ 1
The reliability can be improved by determining that the vehicle 10 is approaching only when the vehicle stands.

[0042]

Since the present invention is constructed as described above, it has the following effects. (1) By using the present invention, the area where the image processing for detecting the entry of the vehicle is performed is only a few rectangular areas, and the amount of data is reduced. Therefore, the speed can be increased. (2) Further, since the processing is based on the normalized correlation calculation,
False detection due to changes in sunshine can be reduced. (3) Further, since the progress of the vehicle is performed only by reading the storage element, the speed is high, and the bit pattern that is not actually possible can be registered in advance depending on the generation status of the bit pattern determined to be the vehicle. Even if an erroneous detection occurs in about one rectangular area, it can be easily detected, and the reliability can be improved.

[Brief description of drawings]

FIG. 1 is a configuration diagram of a vehicle entry detection device according to a first embodiment of the present invention.

FIG. 2 is an explanatory diagram of an image processing state according to a traveling state of the vehicle according to the first embodiment.

FIG. 3 is an explanatory diagram of a relationship between a detection bit pattern of a change area on an image and a traveling state of the vehicle according to the first embodiment of the present invention.

FIG. 4 is a structural diagram of a first example of a vehicle entry detection device according to a conventional technique.

FIG. 5 is an explanatory diagram of a processing process of a first example of a vehicle entry detection device according to a conventional technique.

FIG. 6 is a structural diagram of a second example of a vehicle approach detection device according to a conventional technique.

FIG. 7 is an explanatory view of a processing process of a second example of the vehicle entry detection device according to the related art;

[Explanation of symbols]

DESCRIPTION OF SYMBOLS 1 ... Imaging device 2 ... Image memory 3 ... Rectangular area cutout circuit 4 ... Rectangular area reference memory 5 ... Correlation calculation circuit 6 ... Bit pattern generation circuit 7 ... Entry passage determination circuit 8 ... Communication circuit 9 ... High-order apparatus 10 ... Vehicle 11 ... Background 12 ... Lane 13 ... Rectangular area, R ₀ , R ₁ , R ₂ , R ₃ , R ₄ 14 ... Preliminary detection area, R _X 21 ... Reference memory 22 ... Difference calculation circuit 23 ... Center detection circuit 24 ... Movement state determination circuit

─────────────────────────────────────────────────── ─── Continuation of the front page (56) References JP-A-8-293090 (JP, A) JP-A-8-44994 (JP, A) JP-A-8-123935 (JP, A) Iida et al. Road charging Development of vehicle detection device for system, Mitsubishi Heavy Industries Technical Report, Japan, Mitsubishi Heavy Industries, Ltd., November 30, 1996, Vol. 33, No. 6, pp. 432-435 (58) Fields investigated (Int.Cl. ⁷ , DB name) G06T ⁷ /00-7/60 G01B 11/00 G06T 1/00 280 JISST file (JOIS)

Claims (2)

(57) [Claims] 1. A device for detecting entry of a vehicle into a gate at a tollgate of a parking lot or a toll road, comprising: (A) an image pickup device (1) for picking up an image in the entrance direction.
(B) an image memory (2) for storing an image output from the image pickup device, and (C) a rectangular area cutout circuit (3) for cutting out a plurality of vertical or horizontal rectangular areas in the image memory, (D) Rectangular area reference memory (4) for registering an image of a rectangular area when there is no vehicle, and (E) Content of the rectangular area reference memory and normalization of the rectangular area of the original image output by the rectangular area cutout circuit. A correlation calculation circuit (5) for obtaining a correlation value; (F) a bit pattern generation circuit (6) for obtaining a bit pattern of 0 or 1 for each rectangular area from the output of the correlation calculation circuit (5); (H) The rectangular area cutout circuit, which has an approach passage judging circuit (7) for judging whether the vehicle is approaching or not by considering one data as a bit pattern string for each rectangular area output from the pattern generation circuit (6). (3) is an image memo A plurality of rectangular areas are set substantially perpendicular to the passing direction of the vehicle above, and (I) the bit pattern generation circuit (6) determines that there is no change when the normalized correlation value in each rectangular area is large. 0
If it is small, 1 is given as a bit pattern because there is a change, and (J) the approach passage determination circuit (7) outputs the bit pattern sequence obtained from the bit pattern generation circuit (6) by the number of rectangular regions. The vehicle entry detection device is characterized by detecting entry and passage of a vehicle at high speed from an entry passage pattern which is assigned in advance for each data by handling as data having a bit length of. 2. A rectangular area which is closest to the vehicle approach direction on an image among a plurality of rectangular areas cut out by a rectangular area cutting circuit (3) is set as a preliminary detection area (R _X ), (2) When the vehicle is detected only in the preliminary detection area (R _X ), the approach detection signal is not output to the host device, and the vehicle is also detected in the second and subsequent rectangular areas. The vehicle entry detection device according to claim 1, wherein the vehicle entry detection signal is output after the vehicle entry.