JPS62222391A - 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 abnormality monitoring device that images a monitoring area using an image input means such as a television camera, and recognizes and determines the presence or absence of an abnormality in the image. More specifically, the present invention relates to an abnormality monitoring device that has an improved method of updating a reference image that is compared with a current image of a monitoring area.

(Background Art) Conventionally, this type of abnormality monitoring device has an image input means for capturing an image of a monitoring area and quantizing an image signal, and compares an image obtained by this image input means with a reference image. I was trying to obtain the information necessary for abnormality determination. The reference image used is an input image transferred to the reference image memory when there is no abnormality in the monitoring area. The reference image needs to be updated at regular intervals to offset changes in background brightness in the monitoring area.

Figure 4 shows a conventional reference image updating method. As shown in the figure, the current image is captured at a short cycle of, for example, 30 frames per second, but the reference image is captured at a short cycle of, for example, 30 frames per second. It is updated at a longer cycle. Further, when capturing a reference image, the reference image is not updated when an abnormality detection signal is present, and the reference image is updated only when an abnormality detection signal is not present. That is, it is necessary to update the reference image when there is no target (object to be monitored) such as a person or a car in the input image. This is because if the image where the target exists is used as the reference image, if the target disappears or moves afterwards, there will be a difference between the input image and the target part, and the wrong target will always be extracted. It is. Therefore, the update control unit of the reference Wi image checks the presence or absence of the target and permits the update only when the target does not exist. You need to be careful.

This method can be used when the target appears infrequently, such as when monitoring a store at night, but if the target is almost always present, such as on a public sidewalk, the reference image cannot be updated. It will last a long time. During this time, the background brightness and the position of shadows in the monitoring area may change due to changes in sunlight, etc., and there is always a possibility that a difference will occur between the current image and the reference image.

(Object of the invention) The present invention has been made in view of the above points, and
The object is to provide an abnormality monitoring device that can update a reference image to be compared with a current image of a monitoring area at any time regardless of whether there is a monitoring target.

(Disclosure of the Invention) Hereinafter, the configuration of the present invention will be explained with reference to illustrated embodiments.
As shown in FIG. 1, an image input means 1 captures an image of a monitoring area and quantizes an image signal, and compares the image obtained by the image input means 1 with a reference image to obtain information necessary for abnormality determination. In the abnormality monitoring apparatus, the abnormality monitoring apparatus includes an image processing means 2 for obtaining information, an abnormality determination means 3 for determining the presence or absence of an abnormality based on information obtained by the image processing means 2, and an output means 4 for outputting the determination result. an input image memory 21 that stores an image obtained by the input image memory 21; a reference image memory 22 that stores a reference image to be compared with the image stored in the input image memory 21 in an updatable manner; an image processing unit 23 that extracts a portion in which a substantial change has occurred within the monitoring area by comparing the stored contents with the image processing unit 23;
A labeling image memory 24 stores the changed portion obtained by the image processing unit 23, and the outputs of the labeling image memory 24 and the input image memory 21 are inputted, respectively, and the input image excluding the changed portion stored in the labeling image memory 24 is provided. It also includes an arithmetic circuit 25 that outputs the stored contents of the memory 21 to a reference image memory 22.

In the present invention, as described above, the changed portion within the monitoring area is extracted, and the stored contents of the input image memory 21 are outputted to the reference image memory 22 without the changed portion, so that the monitoring area The reference image in the reference image memory 22 can be updated even if there is a changed part in the reference image memory 22.

Preferred embodiments of the present invention will be described below with reference to the accompanying drawings. FIG. 1 is a block diagram showing the overall configuration of an abnormality monitoring device according to an embodiment of the present invention.

The image input means 1 includes an imaging device consisting of a visible/near-infrared imaging camera such as a vidicon or a 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 taken in by this imaging device is A/D converted and then sent to the image processing means 2.The imaging operation of the monitoring area by the image input means 1 is, for example, an image signal of 30 frames per second. It is done at such speed that it can be achieved.

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 that stores in advance an image when no abnormality exists in the monitoring area. 22. 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 image signals of only pixels with image changes are obtained.

FIG. 2 is a flowchart showing an example of an algorithm for image processing in the image processing means 2. As shown in FIG. First, pixel-to-pixel subtraction is performed between a reference image that does not include abnormal signals that is updated at a constant cycle, and an input image that is taken in from the image input means 1 from time to time, and only pixels that have changed in brightness have values. will be converted to an image like this. After that, filtering processing using a 3×3 mask is performed to remove noise. Next, slicing processing is performed for each pixel using the upper and lower limit values,
Only pixels within a certain range are binarized and emphasized. Furthermore, after noise is removed by filtering processing again, labeling processing begins. Of each labeled object, objects with a certain area or less than a certain number of pixels are removed, and then features such as the center of gravity position and second-order moment are calculated for objects with an area larger than a certain area. The above procedure is executed for each current image frame, the extracted object is tracked between frames, and the moving trajectory of the target is determined.

The results after labeling are temporarily held in the labeling image memory 24. The labeling image at this time corresponds to the content of the input image stored in the input image memory 21. FIG. 3 shows an example of an input image and a labeled image.

The input image in Fig. 3 is a case where a passerby exists in a background image taken of a road, and in the labeling image corresponding to this input image, the passerby part is labeled ``1''. and other parts are labeled “o
” will be added.

Next, the input image memory 21 and the labeling image memory 24 are simultaneously read out and input to the arithmetic circuit 25.

The function of this arithmetic circuit 25 is that the labeling image memory 24
When the value of is "O", the stored contents of the input image memory 21 are transferred as they are to the reference image memory 22, and when the value of the labeling image memory 24 is other than "0", the contents are transferred from the input image memory 21 to the reference image memory 22. The data transfer route is masked. As a result, the reference image is updated while the target part in the current image (area P surrounded by the dotted line in Figure 3) is masked, and only the parts other than the target are updated9. Since the target is moving, the masked part mentioned above will be updated if the target moves from that location, so it will not affect the abnormality determination in the later stage6.
In the embodiment, the reference image is also updated every time the current image is captured, but the reference image may be updated by thinning out, as in the conventional example shown in FIG.

The abnormality determining means 3 determines the presence or absence of an abnormality based on the result of the image processing performed by the image processing means 2. For example, when a changing object as shown in FIG. 3 appears on the screen, when this changing object appears from the side of the screen and disappears at the bottom of the screen, it is determined that the changing object has approached.

When the changing object appears from the bottom of the screen and disappears at the top of the screen, it is determined that the changing object has moved away. Furthermore, when the changing object moves horizontally within the screen, it is determined that the changing object has crossed the road. The abnormality determining means 3 further uses a pattern recognition method to compare the shape of the changing object with various pre-stored shapes, and calculates the size and moving speed of the changing object to determine if the changing object is It determines whether it is a human, a car, a two-wheeled vehicle, etc.

In addition to such an advanced artificial intelligence type determination means, the abnormality determination means 3 may be a simple abnormality that simply counts the number of changed pixels and determines that an abnormality has occurred if the counted value exceeds a set value. A determining means may also be used.

The output means 4 issues an alarm, issues a caution signal, or records the details of the abnormality depending on the determination result by the abnormality determination means 3. In the example shown in FIG. 3, an operation is performed to record the number of people and cars passing on the road by direction of movement.

In addition, in the description of the embodiment, the case where the abnormality monitoring device of the present invention is used as a road traffic monitoring device has been described, but the application of the present invention is not limited to this.
It can also be widely used in intrusion detection devices, fire detection devices, factory safety monitoring devices, etc.

(Effects of the Invention) As described above, in the present invention, a substantial changed part in the monitoring area is extracted by comparing the stored contents of the input image memory and the reference image memory, and this changed part is removed. Since the storage contents of the reference image memory are updated with the storage contents of the input image memory, there is an effect that the reference image memory can be updated even when a changed portion exists within the monitoring area.

In addition, in the present invention, since it is possible to update the reference image regardless of the presence or absence of a changed part, the reference image can be updated at regular intervals. Therefore, the abnormality monitoring device of the present invention , it is particularly suitable for outdoor monitoring applications where there are many changes in background brightness.

[Brief explanation of drawings]

FIG. 1 is a block diagram showing the overall configuration of an abnormality monitoring device according to an embodiment of the present invention, FIG. 2 is a flowchart showing an example of an image processing algorithm in the same, and FIG. 3 is an update operation of a reference image in the same. FIG. 4 is an explanatory diagram of reference image update behavior (%) in a conventional example. 1 is an image input means, 2 is an image processing means, 3 is an abnormality determination means, 4 is an output means, and 21 is an input An image memory, 22 is a reference image memory, 23 is an image processing section, 24 is a labeling image memory, and 25 is an arithmetic circuit.

Claims (1)

[Claims] (1) Image input means that images the monitoring area and quantizes the image signal, 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 image processing. In an abnormality monitoring device including an abnormality determining means for determining the presence or absence of an abnormality from information obtained by the means, and an output means for outputting the determination result, an input image memory for storing an image obtained by the image input means; A reference image memory stores a reference image to be compared with the image stored in the input image memory in an updatable manner, and a comparison of the stored contents of the input image memory and the reference image memory indicates that a substantial change has occurred within the monitoring area. a changing part extraction means for extracting a part;
a changed part storage means for storing the changed part obtained by the changed part extraction means; and an input image memory into which the outputs of the changed part storage means and the input image memory are respectively input, excluding the changed part stored in the changed part storage means; an arithmetic circuit that outputs the stored contents of the image memory to a reference image memory.