JPH07298249A - Automatic monitoring device - Google Patents

Abstract

PURPOSE:To supervise a fault before the occurrence of an accident automatically and unattendendly by picking up continuously an image of a monitoring region, eliminating undersired information from image video information so as to obtain digital image data and comparing sequentially the recorded digital image data with digital image data just before when no failure is detected thereby discriminating the presence of the failure. CONSTITUTION:An image processing section 21 converts a video image picked up by a video camera into digital data and an image recording section 22 records the data. When train proximity information such as a crossing gate operation signal is received, a control section 25 is operated to receive plural images continuously between a recent point of time when a vehicle comes to an emergent stop before the crossing from the point of time of the operation and a sequential image processing section 21 implements the image processing and the image recording section 22 records tentatively digital data as each image. A discrimination section 23 compares digital data of a reference image when no entry vehicle comes in the crossing with 1st image digital data to discriminate the presence of the entry vehicle.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an automatic monitoring device, for example, an automatic monitoring device used to monitor the presence of an approaching vehicle in a railroad crossing when a train is approaching.

[0002]

2. Description of the Related Art A conventional monitoring device, such as a video camera recording device installed in an automatic cash dispenser, records an image of a monitoring place and uses it for processing after an accident occurs, such as infrared rays. There are automatic monitoring devices that use light rays and ultrasonic sensors, but video cameras cannot automatically detect danger before an accident occurs, and automatic monitoring devices that use infrared sensors ensure accurate and reliable monitoring. Has a drawback that the device configuration becomes complicated.

FIG. 3 is a diagram showing an example of an automatic monitoring device for monitoring the presence or absence of an approaching vehicle in a railroad crossing when a train is approaching by using infrared rays. In the figure, 100A and 100B are train passing directions, and 200A and 200B are automobiles. Etc. approaching and passing direction,
A to F are infrared ray emitters, and a to f are light receivers.

Next, the operation of the conventional device shown in FIG. 3 will be described. Activates the device when trains approach,
Infrared beams are emitted from the light emitters A to F, and are received by the corresponding light receivers a to f. When a vehicle such as an automobile invades the railroad crossing, any of the beams emitted from A to F is blocked by this vehicle, and this is detected by the light receiver to confirm the intrusion of the vehicle, and the alarm device is automatically activated. A warning. In this case, in order to make the device capable of surely detecting the position of a vehicle such as an automobile, it is necessary to have at least 6 or more pairs of infrared sensors as shown in FIG. 3 when the road has a double track. Become.

[0005]

As described above, the conventional monitoring apparatus using the video camera recording apparatus cannot perform automatic monitoring before an accident occurs. In addition, in an automatic monitoring device that uses a light or sound wave sensor such as an infrared beam,
As shown in FIG. 3, there is a problem that the structure becomes complicated. That is, as shown in FIG. 3, a combination of a plurality of sensors is used to detect the point where the vehicle cuts off the beam. Therefore, for example, even if the road inside the railroad crossing is a double track, there are six or more sensors. Is required, and as the number of lanes increases to 3 or 4 lanes, more sensors are required, which makes it difficult to install and maintain the sensors.

Further, in order to ensure accuracy, it is necessary to install a sensor in consideration of height. That is, when a vehicle with a low vehicle height is used as a reference, for example, in a vehicle with a high vehicle height such as a truck, the beam passes through the gap between the tires, and there is no vehicle entering despite the vehicle entering the crossing. There is a risk of making an erroneous decision. Conversely, if a vehicle with a high vehicle height is used as a reference, it becomes impossible to detect a vehicle with a low vehicle height or a fallen object.

The present invention has been made to solve the above problems, and it is possible to accurately detect a target that has entered the detection range by using one video camera regardless of the size of the target or the height of the target. It is intended to provide an automatic monitoring device that can perform.

[0008]

SUMMARY OF THE INVENTION An automatic monitoring apparatus according to the present invention includes a video camera for continuously picking up images in a monitoring range, and unnecessary information is removed from sequentially picked-up image video information to leave only necessary information. An image processing unit that sequentially converts each image into digital image data, an image recording unit that temporarily records the sent digital image data, and a digital image data recorded in the image recording unit is not detected as a sequential abnormality. Further, it is characterized in that it is provided with a judging section for judging whether or not there is an abnormality by comparing with the immediately preceding digital image data, and an alarm signal output section for automatically outputting an alarm signal when the judging section judges that there is an abnormality.

Further, the image processing performed by the image processing section to remove unnecessary information from the image video information to obtain only the necessary information is
It is characterized in that it is a process that removes the lightness and saturation from the video image and makes it an image of only hue.

Further, the image processing performed by the image processing unit to remove unnecessary information from the image video information to obtain only the necessary information is as follows.
It is characterized in that it is a process of converting a video image into an image in which only contours are expressed in monotone.

Further, when a device used to monitor the presence of an approaching vehicle inside the railroad crossing when the train is approaching, the presence or absence of an abnormality is determined by the digital image data of the latest point when the train is stopped before the railroad crossing due to an emergency stop. An alarm signal is output, but if the digital image data cannot be used for the determination, a means for outputting the alarm signal based on the determination result of the immediately preceding digital image data is provided.

[0012]

Embodiments of the present invention will be described below with reference to the drawings. FIG. 1 is a diagram showing an embodiment of an automatic monitoring device of the present invention, in which 1 is an image pickup device (video camera), 2
Is a processing device, 3 is an alarm device, 10 is a monitoring range (inside the railroad crossing)
Indicates. 2 is a block diagram showing the configuration of the processing apparatus shown in FIG. 1. In the figure, 21 is an image processing unit, 2
Reference numeral 2 is an image recording unit, 23 is a determination unit, 24 is an alarm signal output unit, and 25 is a control unit.

Next, the operation will be described. Image processing unit 2
Reference numeral 1 converts a video image captured by the video camera 1 into digital data and records it in the image recording unit 22. In this case, in order to reliably determine the presence or absence of an approaching vehicle in a short time, for example, a video image is subjected to a hue-only image excluding lightness and saturation, and is converted into digital data. For example, a video image is converted into an image in which only contours are expressed in monotone, and this is converted into digital data. Etc., and is recorded in the image recording unit 22 as digital data for each image.

Here, when train proximity information such as a breaker operation signal is input, the control unit 25 operates, and a plurality of images are continuously displayed from this time to the latest time when the vehicle can be emergency stopped before the railroad crossing. Are sequentially captured, image processing is sequentially performed by the image processing unit 21, and one image is temporarily recorded in the image recording unit 22 as digital data. The next determination unit 23 compares the digital data of the reference image when there is no approaching vehicle at the level crossing with the digital data of the first image to determine whether or not there is an approaching vehicle. This judgment is made by comparing each font data that constitutes one pixel, and whether or not there is an approaching vehicle is judged by whether or not the number of different font data exceeds the judgment reference value. Is determined by the number of different font data. That is, if the number of different font data exceeds the judgment standard value,
It is determined that there is an approaching vehicle, and if the number is large, it is determined that the approaching vehicle is large.

The digital data of the second and subsequent images is determined by comparing with the image data of the first time when the image of the first time is determined to be "no approaching vehicle".
When it is determined that the image of the first time is “entry vehicle is present”, it is determined by comparing with the reference image data. That is, the image data that is sequentially captured is configured to reduce the determination error as much as possible by comparing it with the immediately preceding image data of “no approaching vehicle (no abnormality)”. Therefore, for example, even when the amount of sunshine changes, such as monitoring inside a railroad crossing from daytime to nighttime, the determination error can be reduced. In order to avoid the risk when the final data cannot be obtained due to equipment failure, etc., it is necessary to take images in succession from the beginning of the circuit breaker until it completely descends and to judge the data of multiple images. Therefore, self-diagnosis when the device fails is also possible.

When it is determined that the final data is “entry vehicle is present”, the determination unit 23 outputs the alarm signal output unit 24.
The information is sent to the navel, and the warning signal is sent out at the warning signal output section. This allows unattended automatic monitoring before an accident occurs. When the final data is determined to be “no approaching vehicle”, the data is recorded in the image recording unit 22 so as to be digital data of the reference image when the train approaches the next train.

The above embodiment has been described as an apparatus for monitoring an approaching vehicle inside a railroad crossing, but since it is based on image information from a video camera, the monitoring range and the size of the monitoring target can be easily changed. It goes without saying that it can be used, for example, for monitoring safes and automatic teller machines when the store is closed, for monitoring entrances and exits of buildings at night, for monitoring dangerous areas, and restricted areas.

[0018]

As described above, the automatic monitoring apparatus of the present invention is capable of automatically monitoring an abnormality before an accident with one video camera unattended, and is capable of centrally monitoring the entire monitoring range. When there are few abnormalities, it is possible to predict to some extent what the abnormality is. Also, because it is based on the image information from the video camera,
Fewer restrictions on what to monitor. Furthermore, since the determination is performed using a plurality of images, there are advantages that there are few determination errors and stable detection is possible, and self-diagnosis is possible when the device is out of order.

[Brief description of drawings]

FIG. 1 is a diagram showing an embodiment of an automatic monitoring device of the present invention.

FIG. 2 is a block diagram showing a configuration of a processing device shown in FIG.

FIG. 3 is a diagram showing an example of a conventional automatic monitoring device.

[Explanation of symbols]

 1 Imaging Device (Video Camera) 2 Processing Device 3 Alarm Device 10 Monitoring Range (Inside Railroad Crossing) 21 Image Processing Unit 22 Image Recording Unit 23 Judgment Unit 24 Warning Signal Output Unit 25 Control Unit

─────────────────────────────────────────────────── ─── Continuation of the front page (51) Int.Cl. ⁶ Identification code Internal reference number FI Technical display location G08B 25/00 510 M 9377-5G H04N 5/225 C

Claims (4)

[Claims] 1. A video camera for continuously picking up images in a surveillance range, and an image processing unit for sequentially converting image data of sequentially picked up image data into unnecessary image information by removing unnecessary information. Then, the digital image data recorded in the digital image data recorded in the digital image data recording unit is temporarily compared with the digital image data immediately before no abnormal image was detected. An automatic monitoring device comprising: a determination unit for determining and an alarm signal output unit for automatically outputting an alarm signal when the determination unit determines that there is an abnormality. 2. The image processing for removing unnecessary information from the image video information performed by the image processing unit to obtain only the necessary information,
2. The automatic monitoring device according to claim 1, wherein the processing is processing of removing only lightness and saturation from a video image to obtain an image of only hue. 3. The image processing for removing only unnecessary information from the image video information performed by the image processing section,
2. The automatic monitoring device according to claim 1, wherein the automatic monitoring device is a process for converting a video image into an image in which only contours are expressed in monotone. 4. When constructing a device used to monitor the presence of an approaching vehicle inside a railroad crossing when a train is approaching, the presence / absence of an abnormality is determined by digital image data of the latest point when the train is stopped before the railroad crossing due to an emergency stop. The alarm signal is output, but when the digital image data cannot be used for the determination, a means for outputting the alarm signal according to the determination result of the immediately preceding digital image data is provided. Automatic monitoring device.