JPH036984A - Fault detector using video image - Google Patents

Abstract

PURPOSE:To discriminate only the occurrence of a fault suitably by obtaining respective absolute deviations of a reference picture and succeeding plural pictures and comparing a change width and a specified value obtained from the absolute deviation for each picture. CONSTITUTION:A reference picture is fetched in a picture buffer 6. Then a picture after DELTAt is fetched in a picture memory 7 to obtain an absolute deviation Hi of both the picture. Moreover, plural pictures obtained at succeeding times are fetched in the picture memory 7 to be revised and the deviation Hi with respect to the reference picture to form discrimination hold signals H1-Hn. Then based on the hold signals H1-Hn, a minimum value is subtracted from the maximum value to obtain a change width DELTAH. Then the change width DELTAH is compared with a specified value obtained by experiments in advance. As a result, when the change width DELTAH is smaller than the specified value, it is discriminated there is no fault and the said processing is repeated. On the other hand, when the change width DELTAH is the specified value or over, it is discriminated that there is a fault.

Description

DETAILED DESCRIPTION OF THE INVENTION [Field of Industrial Application] The present invention relates to an anomaly detection device using an Eirin, which is suitable for use as a device for automatically monitoring the occurrence of an anomaly in, for example, crime prevention monitoring, fire monitoring, etc.

[Conventional technology]

FIG. 2 is a diagram showing the configuration of a conventional monitoring device. The television camera 1 converts images of a surveillance area into electrical signals, and is connected to a receiver 3 and a control circuit 4 via a signal processing section 2.

In the signal processing section 2, the output signal of the television camera 1 is
The signal is input to either the image buffer 6 or the image memory 7 via the signal switching controller 5. Image buffer 6
stores a reference image showing a normal state. The image memory 7 stores images after the reference image. Both outputs are input to the calculation control section 8. This arithmetic control section 8 is
The value obtained by subtracting the image from the image buffer 6 and the image memory 7 is compared with a preset standard value, and as a result, if the above value is more than the specified value, it is determined that there is an abnormality, and if it is less than the specified value, it is determined that there is no abnormality. . Thereafter, the arithmetic control unit 8 updates the image memory 7 and continues the same determination as described above.

[Problem to be solved by the invention]

As mentioned above, in the conventional method, whether or not there is an abnormality is determined based on the result of subtraction between the standard image in the normal state and the subsequent single image. When a disturbance such as a change in occurs, the probability of erroneously detecting the occurrence of an abnormality increases. in this way,
Conventionally, there was no function to distinguish between changes in the environment and the occurrence of an abnormality, and there were problems such as a lack of detection reliability.

The present invention has been made in view of the above points, and an object of the present invention is to provide an abnormality detection device using images that can prevent false detection due to disturbance and accurately determine only the occurrence of an abnormality.

[Means to solve the problem]

As a method for identifying disturbances, we focus on the fact that the image constantly changes when an abnormality occurs, and solve the problem using the following method.

That is, the reference image and subsequent images (multiple planes) exist in an extremely short time domain. Therefore, as a comparison method,
The absolute deviation amount of each image between the reference image and the subsequent plural images is determined, and the variation width obtained from the absolute deviation amount for each image is compared with a specified value to determine whether an abnormality has occurred.

[Effect]

According to the above configuration, a plurality of images within a short period of time are compared and monitored. Therefore, erroneous detection due to disturbance can be prevented, and only the occurrence of an abnormality can be accurately determined.

〔Example〕

Hereinafter, one embodiment of the present invention will be described with reference to the drawings. Note that the detection configuration diagram is the same as that in FIG. 2, and since the present invention is an improvement of the signal processing section 2 in FIG. 2, the configuration will be briefly explained based on FIG. 2.

That is, the television camera 1 shown in FIG. 2 converts the image gI of the surveillance area into an electrical signal. The signal is input to either the image buffer 6 or the image memory 7 via the signal switching controller 5. Each output of the image buffer 6 and the image memory 7 is input to the calculation control section 8, respectively. Further, the output of the arithmetic control section 8 is input to the receiver 3 and the control circuit 4. The arithmetic control unit 8 also controls the signal switching controller 5 regarding the image pack allo or image capture into the image memory 7.

Next, the technical means of the arithmetic control unit 8 related to the present invention will be described in the first section.
This will be explained based on the diagram.

FIG. 1(a) is a diagram showing the timing of image capture in the present invention, in which images are captured at intervals of Δ. Δt shall be set to an extremely short time; for example, due to the nature of television cameras, the shortest time is 1/1
It takes about 60 seconds.

First, an image as a reference image is taken into the image buffer 6. Next, an image after Δ is taken into the image memory 7,
The absolute deviation amount H1 between the two paintings is determined by the following formula.

Aj: Brightness level of an arbitrary pixel in the reference image Bj: Brightness level of an arbitrary pixel in a subsequent image l: Number of comparison images (1 to !1) Hl obtained here is used as a determination pending signal. Also, the image memory 7 is updated for the plurality of planes obtained after that,
The above equation H1 with respect to the reference image is obtained, and determination pending signals for H1 to Hfl are generated.

Next, as shown in FIG. 1(b) K, based on the holding signals H1 to Hn, the minimum value is subtracted from the maximum value to obtain the change width ΔH. Then, the variation width Δ obtained in this way
(2) is compared with a specified value determined in advance through experiments, and if the variation width ΔH is less than the specified value, it is determined that there is no abnormality, and the above process is repeated.

On the other hand, if the variation range ΔH is equal to or greater than the specified value, it is determined that there is an abnormality. If there is an abnormality, only the video of the television camera where the abnormality is detected or the change area is outputted to the receiver 3 shown in FIG. 2, and at the same time, the fact that there is an abnormality is outputted to the control circuit 4.

By comparing and monitoring a plurality of images within a short period of time in this way, it is possible to prevent false detections due to disturbances and accurately determine only the occurrence of an abnormality.

〔Effect of the invention〕

As described above, according to the present invention, by detecting and determining the degree of change in the amount of deviation between a reference image and a plurality of subsequent images in an extremely short time range, (1) accurate early detection of an abnormality location; (2) Preventing false detection due to external disturbances; (3) Improving the reliability of the device.

[Brief explanation of drawings]

FIG. 1 is a diagram for explaining an abnormality detection method according to an embodiment of the present invention, and FIG. 2 is a block diagram showing the configuration of a conventional monitoring device. Reference plane

Claims (1)

[Claims] A device for monitoring the occurrence of an abnormality with a television camera by processing a video signal obtained from the television camera, comprising: a first storage means for storing a reference image; and a first storage means for storing a reference image; The second step is to sequentially store a plurality of images obtained in
storage means, and deviation amount calculation for calculating the absolute deviation amount of both images for each of the images based on the reference image stored in the first storage means and the image stored in the second storage means. means, a change width calculation means for calculating the change width of the absolute deviation amount for each image obtained by the deviation amount calculation means, and a change width obtained by the change width calculation means and a preset specified value. What is claimed is: 1. An abnormality detection device using an image, comprising a change range comparison means for comparing, and determining occurrence of an abnormality according to a comparison result of the change range comparison means.