JPS62222392A - Abnormality monitor - Google Patents

Abstract

(57) [Summary] This bulletin contains application data before electronic filing, so abstract data is not recorded.

Description

DETAILED DESCRIPTION OF THE INVENTION (Technical Field) The present invention relates to an image recognition type abnormality monitoring device that detects the presence or absence of an abnormality within a monitoring area using image input means such as a television camera.

(Background Art) In a conventional abnormality monitoring device of this type, an image processing unit calculates the luminance difference between each pixel of an input image and a reference image, binarizes it at a certain setting level, and then converts it to a set level. The number of pixels in which the above luminance difference occurs is counted, and when the counted value exceeds a certain set value, it is determined that there is an abnormality. therefore,
For example, there was a problem in that the system often malfunctioned due to image changes caused by environmental factors such as trees widely distributed within the screen, shaking of the water surface, rain/snowfall, and sunlight.

Therefore, in order to eliminate malfunctions caused by these environmental factors, an abnormality monitoring device has been proposed that determines the presence or absence of an abnormality based on pre-stored knowledge for abnormality determination. 60-227398), FIG. 2 is a block diagram showing the configuration of this conventional example. Images of the monitoring area obtained by the image input means 1 are sequentially input to an input image memory 21 in the image processing means 2. The image processing unit 23 compares the stored contents of the input image memory 21 and the reference image memory 22 and generates a change image. Take out.

The detection area setting means 36 divides the monitoring area into a plurality of detection areas and sets a guard against intrusion for each detection area. The detection area memory 37 stores this detection area and the warning bottom for each detection area. The abnormality determination means 3 determines in which detection area the moving object obtained by the image processing means 2 is present, and determines whether the moving object is moving from a detection area with a low alert base to a detection area with a high alert base. When it is determined that an abnormality has occurred,
The output means 4 is configured to issue an alarm or the like.

However, with devices that have such advanced judgment functions, as the detection area is subdivided, it takes longer to judge anomalies, and there is a possibility that they cannot be used to monitor areas that require high-speed processing. there were.

(Object of the Invention) The present invention has been made in view of the above-mentioned points, and its purpose is to provide abnormality monitoring that enables rapid abnormality determination and has high reliability of abnormality determination. The purpose is to provide equipment.

(Disclosure of the Invention) As shown in FIG. An image processing means 2 that compares the image with a reference image and obtains the information necessary for abnormality determination, and a main detection area memory 3 that stores a main detection area for making a rough determination of the abnormality occurrence location within the monitoring area.
3, a main abnormality determination means 31 for determining the presence or absence of an abnormality within the main detection area, and sub-detection area memories 341 to 34T1 for storing sub-detection areas for making detailed determinations of abnormalities within the main detection area. It includes a sub-abnormality determining means 32 that determines the presence or absence of an abnormality while referring to a sub-detection area memory based on stored knowledge for abnormality determination, and an output means 4 that outputs the determination result. , the secondary abnormality determining means 32 performs a sophisticated determination only on the location where the occurrence of an abnormality is detected by the main abnormality determining means 31;
It is designed to determine whether an abnormality has occurred.

That is, in the present invention, it is not necessary to carry out sophisticated monitoring all the time, but rather simple and high-speed monitoring is carried out all the time, and only when an abnormality is detected is a more sophisticated judgment method performed.
This method focuses on the fact that if it is determined whether or not there is a real abnormality, a high-speed determination can be made without reducing detection reliability.For this reason, the main abnormality determination means 31 simply detects a change in brightness. It is sufficient to use a simple determination means that determines that an abnormality has occurred even if only the error occurs.

Embodiments of the present invention will be described below with reference to FIG. The image input means 1 is an imaging bag π consisting of visible and near-infrared imaging cameras such as vidicon and CCD, as well as a pyroelectric vidicon-type far-infrared camera that is particularly effective for intruder detection, fire detection, etc.
The image signal of the monitoring area captured by this imaging device, including This is done at such a speed that an image signal can be obtained.

The image processing means 2 includes an input image memory 21 that stores the image of the current monitoring area obtained by the image inputting means 1, and a reference image memory 22 that stores in advance an image when no abnormality exists in the monitoring area. It has The stored contents of the input image memory 21 and the reference image memory 22 are input to the image processing section 23 and subjected to pixel-to-pixel subtraction.

As a result, the fixed background of the monitoring area is removed, and the image is converted into an image containing only pixels that have changed in brightness. Then 3x
Noise is removed by filtering using a mask such as 3, and the changed portion of the image is extracted by binarizing at a certain level.

In this embodiment, in order to determine the location of an abnormality based only on brightness changes, the area setting means 35 sets a plurality of main detection areas within the monitoring area.

For example, when detecting intruders into a house, roads,
Areas such as fences, gardens, plants, bolts, entrances, windows, walls, roots, etc. are set as main detection areas. The main abnormality determining means 31 determines whether the image change portion extracted by the image processing means 2 is within any of the main detection regions. When there is an abnormality here, in order to more accurately determine the abnormality, the sub-detection area memory 34. - 3411, the sub-abnormality determining means 32 determines whether or not it is truly abnormal. The sub-detection area memory 34i stores sub-detection areas obtained by further dividing the main detection area i (1≦i≦n), and each sub-detection area has its own warning bottom set. For example, when monitoring a garden, the alarm base is set so that the closer you get to the house, the higher the alarm base is, and when monitoring a road, the alarm base is set so that the closer you get to the fence, the higher the alarm base is. The alert bottom is set so that the The sub-abnormality determining means 32 does not judge whether or not a changed part exists in all the sub-detection areas within the monitoring area, but corresponds to the main detection area i limited by the main abnormality determining means 3. Only for the sub-detection area, it is determined whether a changed part exists or not. For example, when the changing part moves from a sub-detection area with a low alert base to a sub-detecting area with a high alert base, it is determined that the changing part is an f person, and When moving from a sub-detection area to a sub-detection area with a low alert base,
When the alarm base is moving within the sub-detection area at the same level, it is determined that the changing part is not an intruder.

In addition, the sub-abnormality determination means 32 recognizes that moving objects can hide behind plants and houses, that there is a change in brightness in the windows of houses due to turning on and off the lights, and that there are moving objects on the road. However, there are few moving objects in the garden, and there are swaying plants and water surfaces due to the wind.
Based on extensive pre-memorized knowledge, such as that an object moving from the road to the garden by climbing over a fence is abnormal, but an object moving from the road to the garden through a gate is not abnormal.
This is to accurately determine the presence or absence of an abnormality.

The output means 4 issues an alarm with a buzzer, lamp, etc., issues an urgent signal, or prints out and records the details of the abnormality using a printer, depending on the result of the above-mentioned abnormality determination.

With this configuration, the main abnormality determining means 31 only determines where the brightness change occurs in the square, so the processing speed is high. The secondary abnormality determining means 32 needs to make a more accurate determination, but since the area in which the abnormality determination should be made is limited in advance by the main abnormality determining means 31, it is possible to perform faster determination than in the conventional example. This is possible, and the detection reliability can be the same as that of the conventional example.

(Effects of the Invention) As described in Part 2, according to the present invention, the main abnormality determination means determines P1 where the image change occurs,
The secondary abnormality determination means performs a more accurate abnormality determination for the area previously limited by the main abnormality determination means.
This has the effect of making it possible to perform faster determination than before while maintaining high detection reliability.

[Brief explanation of drawings]

FIG. 1 is a block diagram of an abnormality monitoring device according to an embodiment of the present invention, and FIG. 2 is a block diagram of a conventional example. 1 is an image input means, 2 is an image processing means, 3] is a main abnormality determining means, 32 is a sub-abnormality determining means, 33 is a main detection area memory, 34. .about.34n is a sub-detection area memory, and 4 is an output means.

Claims (2)

[Claims] (1) An image input means that images the monitoring area and quantizes the image signal, an image processing means that compares the image obtained by the image input means with a reference image and obtains information necessary for abnormality determination, and the monitoring area a main detection area memory that stores a main detection area for making a rough judgment of the location of an abnormality within the main detection area; a main abnormality determination means that determines the presence or absence of an abnormality within the main detection area; and a detailed judgment of an abnormality within the main detection area. a sub-detection area memory for storing a sub-detection area for carrying out the detection; and a sub-abnormality determining means for determining the presence or absence of an abnormality while referring to the sub-detection area memory based on pre-stored knowledge for abnormality determination. , and output means for outputting the determination result. (2) An abnormality monitoring device according to claim 1, wherein the main abnormality determining means is a determining means that determines the presence or absence of an abnormality only based on a change in brightness within the main detection area.