JPS6160094A - Detection device of status change by tv camera - Google Patents

Abstract

PURPOSE:To permit the change of a monitored object changing with time to be easily recognized by performing the maximizing and minimizing processes for every appropriate pixel unit of each image stored in image memory and calculating the difference between the images for which AND operations of the maximizing and minimizing images are carried out to extract components changing with time. CONSTITUTION:At first, gradation processing (multiplying and adding operation processing) is performed for image data stored in image memory processes are carried out using the image P after gradation processing as input to create output Q and Z in image memory 5c and 5d. In image memory 5a and 5b, the image Q after the maximizing processing and the image Z after the minimizing processing are stored, respectively. In the image memory 5b, image F representing changes with time is stored in 4 specified area and differences between images in specified areas in the sequence of pickup input signals from a television camera 3. The differences are made binary processing with the threshold level larger than the density '0' to extract only the components changing time.

Description

DETAILED DESCRIPTION OF THE INVENTION (Technical Field of the Invention) The present invention relates to a television system that detects changes in a monitored object whose shape, position, state, etc. change over time based on image input from a television camera. The present invention relates to a state change detection device using a camera.

[Technical Background of the Invention] For example, in a nuclear power plant, it is necessary to inspect and monitor various devices, equipment, etc. using robots and the like. These inspection and monitoring items include the inspection and monitoring of items whose shape, position, condition, etc. change over time, such as smoke generated from separate equipment and steam leaking from pipes, and are subject to inspection and monitoring by inspectors. In order to reduce exposure, there was a strong demand for the development of an automatic detection device that could be operated remotely.

Generally, dedicated sensors are used to inspect and monitor things that change over time, such as smoke and steam leaking from pipes, but inspection and monitoring items at nuclear power plants include rust on pipes, lighting status of lamps, and meters. The range of indication values is extremely wide, and it is practically difficult to develop a robot that can be equipped with many types of sensors at the same time.

In this respect, the television camera is a versatile sensor and can be applied to various inspection and monitoring items by performing image processing according to the object to be monitored based on the input signal.

T [Object of the Invention] Therefore, the object of the present invention is to provide a television camera that can easily detect changes in the shape, position, state, etc. of a monitored object that changes over time. The object of the present invention is to provide a change detection device.

[Summary of the invention]

To achieve the above object, the detection device of the present invention includes a television camera that captures and inputs images of a monitored object at predetermined time intervals, an image memory that stores a plurality of images captured and input by the television camera, and means for maximizing and minimizing each image stored in the memory for each predetermined pixel unit and performing a logical product between the maximum value image and the minimum value image; The present invention is characterized by comprising means for determining a difference between images and extracting a component exhibiting a temporal change.

Therefore, it is possible to clearly capture the contours of objects to be monitored that change over time, such as smoke or steam, and to accurately detect them.

[Embodiments of the invention]

Hereinafter, one embodiment of the present invention will be described with reference to the drawings.

FIG. 1 shows the relationship between a state change detection device using a television camera and leaked steam, which is an object to be monitored. Reference numeral 1 indicates steam leaking from a pipe 2 (object to be monitored). Reference numeral 3 denotes a television camera, which is mounted on a remotely controllable vehicle, a programmed robot, or the like (not shown). Reference numeral 4 is an interface for digitizing the imaging input signal from the television camera 3 and writing it into a plurality of image memories 5 . The image memory 5... requires a plurality of image memories for input/output and work for processing by the image processing processor 6, but here, four image memories 5a, 5b, 5c, 5 are used.
Suppose that d is in danger.

The image processing processor 6 has functions such as product-sum operation, logical product, addition, difference, and area division for blurring processing. Reference numeral 7 is an interface for displaying the state of the leaked steam 1 on the color monitor 8. Reference numeral 9 denotes a control device for controlling the interface 3, image memory 5, image processor 6, and interface 7.

The operation of the state change detection device configured as described above is as follows.

Assume now that steam 1 is leaking from piping 2 that is being inspected and monitored. This leaked steam 1 from the piping 2 is an object to be monitored that changes significantly in a short period of time. Therefore, as shown in FIG.
The second imaging input signal is stored in the first designated area of the image memory 5a. The waveform in the upper row of FIG. 2 is the frame synchronization signal output from the interface 4, and the waveform in the lower row is the same as the first one. This shows the second imaging input signal gate. The imaging input signal gate interval at this time is determined by the generation status of the leaked steam 1.

In this way, the interface 4 stores a plurality of images in the first to fourth designated areas of the image memory 5a at predetermined time intervals, as shown in FIG. As a result, each designated area of the image memory 5a stores a mixture of image data such as the leaked steam 1 that changes over time and the piping 2 and valves (not shown) that are associated with the leaked steam 1. .

Next, only the leaked steam 1 exhibiting temporal changes is extracted from the image data stored in each designated area of the image memory 5a by the following procedure.

First, the image processing processor 6 performs blurring processing (product-sum calculation processing) on the image data stored in the image memory 5a to smooth the arrangement of the image data. Then, as shown in FIG. 4(a), the images P(i, j) obtained by the blurring process are sequentially generated in four image areas of the image memory 5b, and the following processes are then performed.

As shown in FIGS. 4(a) and (b), the image P(i,j)
The pixel unit is a specific pixel (1, J) and 8 pixels around it, that is, 3 x 3 pixels, and the pixel with the maximum density value is extracted (maximization processing), and this is regarded as the density value (first representative surface element data) of the above-mentioned specific pixel (1, J). Then, such maximization processing is performed for all specified areas.

This maximization process is performed on the image P(i
, j) as input, and output Q(1°j) is generated in the image memory 5C.

Next, the image IP (i, j) after the blurring process is input and minimized, and the output Z (i, j) is stored in the image memory 5d.
to be generated. Note that the minimization process is similar to the maximization process, but instead of the maximum value, the minimum value (second representative surface element data) is used to calculate the surface roughness of a specific pixel (1, J) and the eight surrounding areas. (3
(x3 pixel units).

As a result of the above, the image memories 5a and 5b respectively store an image Q (i, j) after the maximization process and an image Z (i, j) after the minimization process.

Next, take the AND of the bit pattern of pixel (1, J) for the above two images [IQ (i, j), Z (i, j),
Image F(i,j)-Q(i,J) ANOZ(i,j
> is generated in the image memory 5b.

FIG. 5 shows an image after the logical product processing. This image is effective for detecting boundaries and density contours where density changes rapidly.
Even for objects to be monitored, such as steam, smoke, etc., whose boundaries or outlines with the background are not clear, the outlines of the objects can be clearly captured.

The image memory 5b contains images showing situations that have changed over time.
F(i,j) is stored in four specified areas, and the difference is taken between the images of each specified area in the order of the imaging input signal from the TV camera 3, and the binary value is calculated with a threshold value slightly higher than density 0. , and extract only those components that exhibit temporal changes.

FIG. 6 shows an example of a difference image extracted from the difference between the images of the first designated area and the second designated area. The shaded area is the area where the leaked steam 1 has changed, and the other scattered images are areas where the background has changed slightly. Generally, the input image from the television camera 3 changes little by little each time it is input, and the image is not exactly the same. Therefore, as shown in FIG. 6, even though the background appears to change little over time, slight changes appear in the difference image. Therefore, in order to remove image areas that appear due to such minute changes, for each area where changes appear, count the number of pixels (equivalent to the area), and identify areas that exceed a predetermined number of pixels as leaking steam 1. It is recognized as an image, and the position of a rectangle (numeral 10 in FIG. 6) that is in contact with the area is determined. A processing window (window: the area surrounded by the broken line rectangle 11 in FIG. 7) for appropriate area division is created around the position of this rectangle 10, and the image F(i, j) shown in FIG. Binarization is performed using a threshold value slightly higher than O, and region segmentation is performed on the partial images within the processing window. Then, as shown in FIG. 7, different numbers 101, 102, and
103, 104... are attached. Next, area 1 shown in Figure 7
From among 01, 102, 103, 104..., the area containing the difference image shown in Figure 6 is set as a mask image, and a logical product is performed between this mask image and the image after area division in Figure 7. By selecting the region (number) that appears in the output image, and extracting only the region (number) from FIG. 7, an image of the leaked steam 1, which is the object to be monitored, can be obtained.

Next, the same image extraction process is performed between the images of the second designated area and the third designated area, and then between the images of the third designated area and the fourth designated area, and Only the images of the leaked steam 1 exhibiting significant changes are extracted. The image of the leaked steam 1 extracted in this way is displayed on the color monitor 7.

〔Effect of the invention〕

As described above, according to the present invention, it is possible to capture an object to be monitored that exhibits temporal changes such as steam, smoke, etc. as an image with a clear outline, and the observer can understand the shape of the object to be monitored, Changes in position, amount of generation, direction of movement, etc. can be easily detected.

[Brief explanation of the drawing]

The drawings show one embodiment of the present invention. Fig. 1 is a schematic configuration diagram showing the relationship between the detection device and the leaked steam that is the object to be monitored, and Fig. 2 shows the input of a 1ffl image input signal from a television camera. FIG. 3 is a timing diagram showing the timing, and FIG. 4(a) is a diagram showing an example of images stored in four designated areas of the image memory. (b) is an explanatory diagram of the maximization process, Figure 5 is a diagram showing the image after the logical product between the maximum value image and the minimum value image, and Figure 6 is the first image in the image shown in Figure 5. Figure 7 shows an example of the difference image extracted by the difference between the images of the designated area corresponding to the second designated area and the designated area corresponding to the second designated area.
It is a figure which shows the result of dividing the image shown in the figure into regions. 1... Leakage steam (object to be monitored), 3... Television camera, 4.7... Interface, 5... Image memory, 6... Image processing processor, 8... Color monitor, 9 ···Control device. Applicant's agent Patent attorney Takehiko Suzue Figure 1 Figure 4 (a) (b) Figure 5 Figure 6 Figure 7

Claims (3)

[Claims] (1) A television camera that captures and inputs images of objects to be monitored at predetermined time intervals, an image memory that stores a plurality of images captured and input by this television camera, and each image stored in this image memory is divided into predetermined pixels. Means for performing maximization and minimization processing for each unit and calculating the logical product between the maximum value image and the minimum value image, and calculating the difference between these logically processed images to detect temporal changes. What is claimed is: 1. A state change detection device using a television camera, comprising: means for extracting a component exhibiting a change in state. (2) means for maximizing and minimizing each image stored in the image memory for each predetermined pixel unit and performing a logical product between the maximum value image and the minimum value image; The television camera according to claim 1, wherein the means for determining the difference between the AND-processed images and extracting a component exhibiting a temporal change includes an image processing processor. State change detection device. (3) The maximization process and the minimization process are performed using 3×3 pixels centered on a specific image among each image stored in a plurality of designated areas of the image memory, and The device for detecting state changes using a television camera according to claim 1, wherein a maximum value image with the highest density value and a minimum value image with the lowest density value are extracted as representative pixel data, respectively, in a unit. .