JP2541531B2 - Parking detection method - Google Patents

Description

TECHNICAL FIELD The present invention relates to a parking detection system for detecting the presence or absence of parking on a roadside strip or the like of a road.

[Conventional technology]

BACKGROUND ART Conventionally, there is a so-called parking meter that detects whether or not a roadside belt is parked. This is installed in a road section where parking is possible, and under the condition that the driver inserts coins, a certain time (for example, 40 minutes)
It is only intended to allow parking.

On the other hand, various systems have been proposed to measure a traffic flow on a road by recognizing a traveling vehicle by photographing a road surface with a CCTV camera or the like and processing the image of the image. And according to this, since the traffic flow can be measured without using human hands by using the image from the existing surveillance camera, vigorous research has been conducted in various fields. Specific examples thereof include “Development of image processing type traffic flow measurement system”, Sumitomo Electric No. 127 (September 1985).

[Problems to be solved by the invention]

However, a technique for detecting the presence / absence of parking on a road by such an image processing method has not been proposed yet because it is difficult to distinguish the surrounding environment change and a parked vehicle that has been stopped for a long time. Therefore, conventionally, the above-mentioned parking meter is used for detection, but this has the following problems. That is, parking on a road section where a parking meter is not installed cannot be detected at all, and even if installed, the presence or absence of parking cannot be detected unless the driver inserts a coin into the parking meter for operation. Also,
If a parking meter is installed for each parked vehicle, the cost will increase, which may impair the aesthetics of the city and hinder the traffic of pedestrians.

Therefore, an object of the present invention is to provide a parking detection system capable of accurately and inexpensively detecting the presence or absence of a parked vehicle in a wide range.

[Means for solving problems]

A parking detection method according to the present invention is a parking detection method for detecting the presence or absence of parking from an image pattern obtained by photographing a road surface and a vehicle parked on the road surface. A first step of setting a plurality of measurement points in a predetermined measurement region in a captured image, (b) a second step of measuring the brightness of the plurality of measurement points, and (c) a plurality of measurement points In each of a plurality of sections that are set to include a predetermined number of measurement points of 3 or more, the total sum of the absolute values of the differences in brightness is calculated for all of the two measurement points adjacent to each other among the measurement points included in the section. A third step of calculating the spatial total change amount of the brightness in the section by calculation, and detecting the presence or absence of the vehicle in each section based on the total spatial change amount of the brightness, and (d) the vehicle in each section Section based on temporal change of presence / absence A fourth step of determining whether the parking, characterized in that it comprises a. Here, the plurality of measurement points may be present at regular intervals on the measurement points set along the roadside belt on the road surface.

Here, the third step is a step of detecting the vehicle from the spatial change of the luminance at the plurality of measurement points obtained in the second step and detecting the presence or absence of parking from the temporal change thereof. desirable.

[Action]

Since the parking detection method according to the present invention is configured as described above, the brightness of the measurement point set on the road surface is measured,
The presence / absence of a vehicle is detected from the spatial total change amount of the brightness calculated for each section, and the presence / absence of parking is detected from the temporal change.

〔Example〕

An embodiment of the present invention will be described below with reference to FIGS. 1 to 3 of the accompanying drawings. In the description of the drawings, the same elements will be denoted by the same reference symbols, without redundant description.

FIG. 1 is an explanatory view of a parking detection system according to the same embodiment, FIG. 2 is a block diagram of a parking detection device to which this embodiment system is applied, and FIG. 3 is a measurement point setting according to this embodiment system. It is an explanatory view of an example.

First, the structure of the parking detection device will be described with reference to FIG. 2. The parking detection device includes an ITV camera 1 and an image processing device 2 connected to the ITV camera 1. Then, the image processing apparatus 2 is ITV
An AD conversion unit 3 connected to the camera 1, an arithmetic processing unit 4 connected to this, a detection processing unit 5 connected to this, and an output unit 6 for outputting the detection result via a communication line 7. have.

Next, the function of each element shown in FIG. 2 will be described with reference to FIGS. 1 and 3.

The ITV camera 1 is installed above the road 10 as shown in FIG. 1 and takes an image of a certain section in front of it. Then, the obtained image is sent to the image processing device 2 as an analog image signal.

The AD converter 3 converts the analog image signal from the ITV camera
For example, it is converted into luminance data of 2 ⁸ = 256 gradations. The curve indicated by reference numeral 11 in FIG. 1 is an example of the brightness level. Then, this digital signal (luminance data) is given to the next arithmetic processing unit 4.

The arithmetic processing unit 4 performs so-called pixel extraction from the given luminance data. This pixel is a road as shown in FIG.
It corresponds to the measurement points 13 at equal intervals on the plurality of measurement lines 12 set along the roadside band of 10.

The detection processing unit 5 detects the presence or absence of a parked vehicle based on this pixel extraction data. That is, the presence or absence of the vehicle is detected by spatially checking the change in the brightness level at the measurement point. Generally, the spatial change of the brightness level of the road surface is small, and the spatial change becomes large when there is a vehicle. Therefore, the presence or absence of the vehicle can be easily detected by the spatial comparison.

What is important here is to set a plurality of sections including a predetermined number of measurement points of 3 or more, and detect the presence or absence of a vehicle for each section from the total spatial change in the brightness of the measurement points included in each section. However, the presence or absence of parking for each section is detected from the change over time that is examined at regular time intervals.

The output unit 6 sends the detection result of the detection processing unit 5 and an image showing the parked vehicle to a CRT or the like (not shown) via the communication line 7 or the like. As a result, the presence or absence of parking is output and displayed moment by moment.

 Next, the operation of the above embodiment will be described.

Now, it is assumed that two vehicles 21, 23 are parked on the road 10. Then, the level of the brightness data obtained by the A / D converter 3 via the camera 1 becomes as shown by the solid line 11 in FIG. This brightness data is sampled for each measurement point 13 by the arithmetic processing unit 4, and the detection processing unit 5 checks the spatial change in each section. This section is set so as to include four measurement points, as indicated by symbols a1 to d5 in FIG. The number of measurement points included here is not limited to four, and the number of measurement points can be set arbitrarily.

The spatial change of the brightness level is checked as follows, for example. That is, when the sample point is j and the section length is k, the total brightness change amount B _{i in} the section _i is Ask for. Then, the predetermined value C and the total change amount B _i are compared, and B _i
If ≧ C, the vehicle exists, and if B _i <C, the vehicle does not exist. The brightness level check can be grasped not only in terms of spatial changes but also in terms of temporal changes. That is, the above formula (1) is checked at intervals of t seconds, and when the vehicle is continuously present for T time (t ≦ T), it is determined that the parked vehicle is present.

The method of checking the brightness level is shown in FIG. That is, the section a in which the vehicles 21 and 23 are present
In 1 to a5 and b4 to c2, the amount of change in brightness level is large, so B _i
≧ C and the vehicle is present. In contrast, vehicles 21,22
In the spaces b1 to b3, d4, d5 which do not correspond to the above, since the road 10 is detected and the change in the brightness level is small, B _i <C and it is detected that there is no vehicle.

When it is detected that the vehicle is present in the sections a1 to a5 and b4 to c2 for a certain period of time, it is determined that the vehicle is parked.

In this state, when the vehicle (truck) 22 is newly parked as shown by the broken line in FIGS. 1 and 3, the brightness level becomes as shown by the broken line 41 in FIG. Therefore, section b1 ~
In b3 as well, B _i ≧ C and the vehicle is present. Therefore,
When this state continues for T time, parking is enabled in the sections a1 to d3.

The present invention is not limited to the above embodiments, and various modifications are possible. For example, the camera is not limited to the ITV camera and may be any camera. Further, the parking detection location is not limited to the roadside zone of the road, and may be any location such as a parking lot. For example, if the measurement region is wide in the direction crossing the road, it is possible to detect a vehicle parked away from the road side of the road, double parking, or the like. In addition, the brightness level is converted into, for example, three-valued in the arithmetic processing unit,
The vehicle may be detected based on the amount of change. Also, not only monitoring the detection result with CRT etc., but also alerting the parked driver by interlocking with a voice warning device,
You may decide to carry out enforcement.

〔The invention's effect〕

As described above in detail, according to the present invention, the brightness of the measurement point set on the road surface is measured at constant time intervals, and the presence or absence of the vehicle is detected by the spatial total change amount of the brightness of each section,
Since the presence / absence of parking is detected based on the temporal change, it is possible to accurately and inexpensively detect the presence / absence of a parked vehicle in a high range.

In particular, in the present invention, since it is possible to use the existing monitoring ITV in the traffic control system, it is possible to exert the effect of eliminating traffic congestion without requiring special equipment.

[Brief description of drawings]

FIG. 1 is an explanatory diagram of an embodiment of the system of the present invention, and FIG. 2 is a first diagram.
FIG. 3 is a block diagram of an apparatus to which the embodiment system shown in the figure is applied, and FIG. 3 is an explanatory diagram of measurement point setting on a road. 2 ... Image processing device, 7 ... Communication line, 10 ... Road, 1
1,41 …… Brightness level, 12 …… Measurement line, 13 …… Measurement point, 21
~ 23 …… Vehicle.

Claims (2)

(57) [Claims] 1. A parking detection method for detecting the presence or absence of parking from an image pattern obtained by photographing a road surface and a vehicle parked on the road surface, wherein a predetermined measurement in the image of the road surface is taken. Each includes a first step of setting a plurality of measurement points in the area, a second step of measuring the brightness of the plurality of measurement points, and a predetermined number of three or more measurement points of the plurality of measurement points. In each of the multiple sections set in
The spatial total change amount of the brightness in the section is calculated by calculating the sum of the absolute values of the differences in the brightness for all of the two measurement points adjacent to each other among the measurement points included in the section. A third step of detecting the presence / absence of a vehicle in each section based on the total spatial variation, and a fourth step of detecting the presence / absence of parking in each section based on a temporal change in the presence / absence of a vehicle in each section. And a parking detection method comprising: 2. The parking detection method according to claim 1, wherein the plurality of measurement points are present at regular intervals on a measurement line set along the roadside belt on the road surface.