JP2606409B2 - Vehicle monitoring method and device - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION [Industrial Application Field] The present invention relates to a vehicle monitoring technique necessary for notifying a driver of a position of an empty parking area in a parking lot or the like.

[Conventional technology]

Processing of parking lot images to determine the presence or absence of a car for each parking area is described in a document [Shin Nagao (author): Image Recognition Theory, Corona, P.105, First Edition on February 15, 1983]. 2
There is a technique using a difference S S = Σ | f (x, y) −a (x, y) | (1) between two images f (x, y) and a (x, y).

For example, FIG. 2B shows a scene in which a car is present in a parking area 101 separated by a white line illustrated in FIG. In order to find a visible car 104 by processing the image 103, in the prior art, as a first step, FIG. 2 (C) in which the value is 1 in the parking area 101 and 0 otherwise. The mask image (m (x,
y)) 105 is created in advance. Second stage parking area 10
Assuming that an image when there is no car in 1 is f ₀ (x, y), a model image a (x, y) of the parking area 101 is synthesized in advance by the following equation.

a (x, y) = f ₀ (x, y) · m (x, y) (2) When the image 103 is expressed as i (x, y) as the third stage, the parking area 101 is parked by the following equation. The area image f (x, y) is synthesized.

f (x, y) = i (x, y) · m (x, y) (3) As the fourth step, a (x, y) and f (x,
Calculate the difference S in y). Larger than the threshold S _0, for example, the difference S is set in advance as the fifth step, the parking Elijah 101 had determined that there is a car.

[Problems to be solved by the invention]

However, using the above-described conventional technique, for example, it is determined whether or not there is a car in a parking area 108 in which a large portion of the car 107 is hidden, for example, in a parking lot image 106 illustrated in FIG. 3A. In this case, for example, the mask image 109 illustrated in FIG. 3B showing the parking area 108 synthesized in the same manner as above.
To the parking area image synthesized by Equation (3) using
Since a part of 107 is captured, the difference S calculated in the same manner as above
There is larger than the threshold value S _0. As a result, parking area 108
There was a problem of making an erroneous determination that there was a car.

SUMMARY OF THE INVENTION An object of the present invention is to provide a vehicle monitoring method and apparatus for correctly determining the presence or absence of a vehicle in a parking area hidden in the vehicle.

[Means for solving the problem]

According to the vehicle monitoring method of the present invention, a plurality of parking areas for one vehicle are defined by a white line group including a plurality of vertical white lines and a plurality of horizontal white lines corresponding to a vehicle length and a vehicle width, and are obtained by observing fixed points. In monitoring the presence or absence of a parked vehicle in each parking area by imaging, based on each position of the white line group when the plurality of parking areas stored in advance is in the non-parking state, scan the image and scan the vertical white line. It detects all the edges that are the vehicle outlines that intersect with the vehicle, and determines the presence or absence of a vehicle parked in the parking area based on the number of detected edges.

According to the vehicle monitoring device of the present invention, a plurality of parking areas for one vehicle are defined by a white line group including a plurality of vertical white lines and a plurality of horizontal white lines corresponding to the vehicle length and the vehicle width, and are obtained by observing fixed points. A distant white line that monitors the presence or absence of a parked vehicle in each parking area by imaging and stores the position of the vertical white line when there is no parking in each parking area in advance in the imaging range based on the fixed point observation position Storage means, edge calculation means for scanning an image to calculate the number of edges which are vehicle contour lines intersecting the vertical white line, edge storage means for storing the edges, and the edge and the distant white line storage means. Intersection detection means for obtaining an intersection on the pixel with the stored vertical white line and processing the pixel value as a pixel value; and determining presence or absence of a vehicle in each parking area based on a processing signal from the intersection detection means. Characterized by comprising control means for knowledge, the.

[Action]

The present invention, for example, among the white line group surrounding the parking area determined to have a car in the prior art, when all of the vertical white line located far from the fixed point shooting position is hidden on the image,
It is determined that the vehicle is parked in the parking area, and when all of the vertical white lines appear on the image, it is determined that the vehicle is not parked in the parking area.

For example, in the case of the parking area 111 where the car 107 illustrated in FIG. 3 (A) is parked, among the white line group surrounding the parking area 111, a vertical white line located far away from the fixed point photographing position as a base point. One white line 113 is hidden with the outline 112 constituting the car 107 as a boundary. On the other hand, in the case of the parking area 108 in which the vehicle illustrated in FIG. 3A is not stopped, the white line 110 located far from the viewpoint among the white lines surrounding the parking area 108 is seen without being hidden by the car 107. I have. Therefore, first, for example, when determining the presence or absence of a car in the parking area 111, to find an edge that is not parallel to the white lines 110 and 113 from the parking lot image 106, a differential mask of the angle θ formed by the white lines 110 and 113 with the x-axis direction Is binarized from the image of the parking lot image 106 to detect the edge group 114 illustrated in FIG. 3C.
Next, the car 107 illustrated in FIG.
Car 107 crossing white line 115 corresponding to white line 113 when there is no
By examining the contour, ie, the edge of
2 is found, and as a result, it is determined that the car is parked in the parking area 111.

When the same processing is performed on the white line 110, the edge group 1
Of 14, there is no edge that intersects the white line 110, so
It is considered that there is no contour line that hides the white line 110, and as a result, it is determined that there is no car in the parking area.

〔Example〕

 Hereinafter, embodiments of the present invention will be described in detail.

FIG. 1 is a block diagram showing one embodiment of a vehicle monitoring device of the present invention.

The vehicle monitoring device includes an image storage unit 1 that stores an image of a parking lot input by a TV camera or the like, and the vertical white line when no parking is performed in each parking area in advance in an imaging range based on a fixed point observation position. Distant white line storage means 4 for storing the position of
An edge calculation unit 2 that scans an image and calculates the number of edges that are vehicle contour lines that intersect the vertical white line, an edge storage unit 3 that stores the edge, and an edge and the far white line storage unit. Intersection detection means 5 for finding an intersection on the pixel with the stored vertical white line and processing it as a pixel value;
An intersection table storage means 6 for storing an intersection table described later, a display means 7, and a control means 8 for controlling the edge calculation means 2, the intersection detection means 5, and the display means 7 are provided. The control means 8 can be realized by a memory and a microprocessor.
The edge storage means 3, the far white line storage means 4, and the intersection table storage means 6 can be constituted by a memory, and the display means 7 can be realized by a CRT and a memory.

A case will be described in detail in which the vehicle monitoring device having the above configuration determines whether or not there are vehicles in the parking areas 111 and 108 illustrated in FIG. 3A, for example.

In this case, the image storage unit 1 stores an image 106 of the parking lot illustrated in FIG. 3A input by a TV camera or the like.
The intersection table storage means 6 stores the intersection table in which the column of the number of edges is reset to 0 in the intersection table 120 of the format illustrated in FIG. 4A. The distant white line storage unit 4 stores, for example, a group of pixels through which the white line 110 passes and a group of pixels through which the white line 115 passes as shown in FIG. Are filled with, for example, 1 and 2, respectively, and distant white line pixels in which 0 is substituted for other pixel values are stored. FIG. 4B shows the distant white line image 116, wherein 117 denotes a pixel group through which the white line 110 passes, and 118 denotes a pixel group through which the white line 115 passes. The distant white line storage unit 4 further stores the pixel groups 117 and 11 illustrated in FIG.
The eight directions θ ₁ and θ ₂ are stored in the format illustrated in FIG. 4 (C).

The operation of the vehicle monitoring apparatus shown in FIG. 1 starts when the control means 8 activates the edge calculation means 2. The started edge calculation means 2 calculates the average value of the directions θ ₁ and θ ₂ of the pixel group stored in the far white line storage means 4 as the first stage. Is calculated.

As a second step, when the image 106 of the parking lot stored in the image storage means 1 is represented by f (x, y), f (x, y) is expressed by the following equation (4).
The following intensity image a (x, y;) is calculated using a differential mask d (x, y;) (6) in which the average value shown in Expression (5) is substituted for the value of θ in.

a (x, y;) = | ∫∫f (x−u, y−v) · d (u, v;) dudv |
(7) As a third step, a binary image illustrated in FIG. 3C in which the intensity image a (x, y;) is binarized by, for example, an appropriately determined threshold value T Is calculated.

As a fourth step, for example, by using a labeling technique which can be easily realized by a conventional technique, for example, a number exemplified in FIG. 118 is stored in the edge storage means 3 and the process is terminated.

When the edge calculation means 2 ends the processing, the control means 8 activates the intersection detection means 5. Activated intersection detection means 5
In the first stage, for example, the distant white line image stored in the distant white line storage unit 4 is represented by w (x, y), and the edge storage unit 3
When the edge image 118 stored by is represented by e (x, y), w
Each pixel value of (x, y) and e (x, y) is sequentially scanned, for example, with the x direction as a main scanning direction and the y direction as a sub scanning direction. Filled pixel P ₁
(X ₁ , y ₁ ) w (x ₁ , y ₁ )> 0 and e (x ₁ , y ₁ )> 0 (9) Get. Of cross table 120 for storing cross-table storage means 6 as the second stage, after adding 1 to the number of edges parking Elijah 111 found as second line pointed to pixel values L _m, first
Stores the value 7 of the pixel value L _e in section edge.

As the third step, the processing is continued in the same manner as the procedure described in the first step, whereby the position (x ₂ , y ₂ ) of the pixel P ₂ and the pixel value illustrated in FIG. Get. If the value in the same edge number of a pixel value L _e as the fourth stage is not stored already in the line of parking Elijah 111,
After adding 1 to the number of edges about parking Elijah 111 cross table 120 operates similarly to the procedure described in the second stage, stores the values 8 pixel values L _e in the section second edge, the above process Is repeated until scanning of all pixels is completed, and then the processing is terminated. At the time when the intersection detecting means 5 ends the processing, the intersection table 120 stores the intersection table 121 illustrated in FIG. 4 (F).

When the intersection detecting means 5 ends the processing, the control means 8 activates the display means 7. The activated display means 7 scans each row of the intersection table 121 stored by the intersection table storage means 6 and, for example, the parking area 108 represented by the first row has 0 edges, so that no car is stored in the parking area 108. Judge that there is no such as CRT
The character string "parking area 108 ... no cars" is output above, and the number of edges of the parking area 111 represented by the second line is 2
Is greater than or equal to one, it is determined that there is a car in the parking area 111, and the character string "parking area 111 ... there is a car" is output on the CRT, for example, and all the processing is completed.

〔The invention's effect〕

ADVANTAGE OF THE INVENTION According to this invention, the effect of being able to correctly determine the presence or absence of the car of each parking area by processing the image of the parking lot imaged from the line of sight direction in which a part of the car in front is hidden.

[Brief description of the drawings]

FIG. 1 is a block diagram showing an embodiment of a vehicle monitoring device according to the present invention, FIG. 2 is a diagram for explaining the prior art, FIG. 3 is a diagram for explaining the principle of the present invention, FIG. FIG. 4 is a diagram for explaining the operation of the embodiment. 2 ... Edge calculation means 3 ... Edge storage means 4 ... Distance white line storage means 5 ... Intersection detection means 106 ... Parking lot image 118 ... Detected edge image 121 ... Calculated intersection table

Claims (2)

(57) [Claims] A plurality of vertical parking lines corresponding to a vehicle length and a vehicle width define a plurality of vertical white lines and a plurality of horizontal white lines to define a plurality of parking areas for one vehicle. A vehicle monitoring method for monitoring the presence or absence of a parked vehicle in a parking area, wherein the plurality of parking areas stored in advance are scanned based on each position of the white line group when there is no parking, and A vehicle monitoring method comprising: detecting all edges that are vehicle contour lines intersecting a vertical white line, and determining whether there is a vehicle parked in a parking area based on the number of detected edges. 2. A plurality of parking areas for one vehicle are defined by a white line group including a plurality of vertical white lines and a plurality of horizontal white lines corresponding to a vehicle length and a vehicle width. A vehicle monitoring device for monitoring the presence or absence of a parked vehicle in a parking area, wherein a far white line storage that stores a position of the vertical white line in a non-parking state in each parking area in advance in an imaging range based on a fixed point observation position. Means, an image scanning means, an edge calculating means for calculating the number of edges that are vehicle contour lines intersecting the vertical white line, an edge storing means for storing the edge, and an edge storing means for storing the edge and the far white line storing means. Intersection detecting means for obtaining an intersection of the vertical white line with the pixel and processing the pixel value as a pixel value; and determining whether there is a vehicle in each parking area based on a processing signal from the intersection detecting means. Vehicle monitoring device for a control means for, further comprising a said.