JPH0582556B2 - - Google Patents

Description

Detailed Description of the Invention [Technical Field of the Invention] The present invention relates to a content determination device for determining the contents inside a container, and in particular improves an image preprocessing technique for a fluoroscopic image obtained by radiation irradiation. The present invention relates to a content determination device comprising:

[Technical background of the invention and its problems]

This type of determination device converts an X-ray fluoroscopic image of a container obtained by X-ray irradiation into a video signal using an X-ray television camera, and then binarizes this video signal to distinguish between a container area and a background area. This container area image is then binarized by varying the threshold that allows separation into the container frame and the contents inside the container to obtain a binary image. By ORing the binary image and the container area image, the image is divided into a background area, a container frame area, and a contents area, and only the image of the contents is extracted.

However, the image processing means described above is effective when the contents 101 are in the container frame 1 as shown in FIG.
02 and the contents 101 are in contact with the container frame 102 as shown in FIG. 101 cannot be determined.

[Purpose of the invention]

The present invention has been made with attention to the above points, and an object of the present invention is to provide a content determination device that can separate the contents from the container frame and determine the accuracy of the contents even when the contents are in contact with the container frame. be.

[Summary of the invention]

The present invention predicts the thickness of the container frame in advance in terms of the number of pixels, and then calculates the number of pixels corresponding to the thickness of the container frame from a container area image obtained by binarizing an electrically gray perspective image of the container. A content determination device that creates a mask image of a container area that has been reduced by scraping, and extracts only an image of the contents inside the container by performing a logical operation on this mask image and an electrical gradation perspective image of the container. be.

[Embodiments of the invention]

Hereinafter, one embodiment of the present invention will be described with reference to FIGS. 1 to 6. Figure 1 is an overall configuration diagram of the device, Figure 2 is a diagram explaining the operation of the device, and Figure 3 is a diagram explaining the operation of the device.
6 through 6 are diagrams for explaining image processing.
In FIG. 1, numeral 10 is an X-ray fluoroscope, which is an X-ray ray 10b emitted from an X-ray generator 10a.
The container 10c to be inspected is seen through the container 10c, and the X-ray fluoroscopic image is converted into a video signal by the X-ray television camera 10d, and the video signal is supplied to the subsequent A/D conversion circuit 11. 1
0e is a sample stage on which the container to be inspected 10c is placed. This A/D conversion circuit 11 is connected to the X-ray fluoroscope 1
The video signal from 0 is quantized at an appropriate level considering the entire device and converted into gray scale perspective image data discretized with an appropriate number of samplings. 12
is an image memory composed of a semiconductor memory or the like, which stores the gradation perspective image data of the discretized electric signal from the A/D conversion circuit 11, and also stores the subsequent binarization circuit 13 and It is sent to the image logic circuit 14. This binarization circuit 13 is a container area extraction means that has a function of comparing image memory data with a predetermined threshold value and extracting a container area image that is a binary black and white image. It has a circuit that compares a predetermined setting value with each pixel value of the gray perspective image stored in the image memory 12, and if the pixel value is less than the setting value, the value of that pixel is set to "0", It has a function that converts to a binary black and white image that is set to "1" if it exceeds the set value. In this embodiment, for example, "6" is used as the setting value.

Reference numeral 15 denotes an area reduction circuit, which specifically has the following processing functions. That is, this area reduction circuit 15 selects one of the required number of pixels.
If all pixels in the vicinity of this pixel of interest are "1", the pixel value of the output image corresponding to the pixel of interest is set to "1", and one of the remaining pixels in the vicinity If the pixel is always "0", the pixel value of the output image corresponding to the pixel of interest is set to "0". In this process, one pixel located outside the image is removed from the entire outer contour side of the container area image corresponding to the container to be inspected 10c having a pixel value of "1" to create a reduced binary image. . The above process is performed by removing pixels corresponding to the thickness of the predicted container frame (10 pixels) from the outer contour side of the container area image. After obtaining a reduced container area image by scraping it inward, the image is supplied to the image AND circuit 14. Here, the significance of reduction is that even if the object inside the container is in contact with the image area of the container frame,
The idea is to remove the container frame while leaving the image of the object inside the container intact. The image AND circuit 14 sets the area of the reduced container area image from which 10 pixels have been deleted by the area reduction circuit 14 to "1" and the other areas to "0" as a mask image. The mask image is used to mask the gradation fluoroscopic image read out from the image memory 12, and the gradation value of the fluoroscopic image pixel corresponding to the "1" pixel of the mask is left unchanged, and the gradation value is changed to the "0" pixel of the mask. The pixel of the corresponding perspective image is “0”
A logical product operation is performed, and thereby a gradation perspective image of the "1" portion of the mask, that is, an image of the contents inside the container is extracted. Reference numeral 16 denotes a result display device that displays a transparent image of the contents inside the container obtained by the image AND circuit 14. Reference numeral 17 denotes a control device that controls each of the components 12 to 16 of the apparatus at predetermined timing, and for example, a control computer or the like is used.

Next, the operation of the apparatus configured as described above will be explained with reference to FIGS. 2 to 6. X
The X-ray fluoroscopic image output from the inspection container 10c after being irradiated with the ray 10b is detected by the X-ray television camera 10d as shown in step 21 in FIG. After being converted into an electrical continuous wave gradation image signal, the A/
The signal is sent to the D conversion circuit 11. This A/D conversion circuit 11 converts the continuous signal of the gradation fluoroscopic image sent from the X-ray fluoroscope 10 into step 2 in FIG.
As shown in FIG. 2, the image is converted into a discrete gradation perspective image quantized with, for example, 256 pixels x 256 pixels x 8-bit gradation (that is, 256 gradations) (step 22), and stored in the image memory 12 (step 23). . The gradation fluoroscopic images stored in the image memory 12 in this way are sequentially read out in a predetermined order based on a read command from the control device 17 and sent to the binarization circuit 13.
Here, each pixel value of the input gradation perspective image is compared with a set value, for example, "6 gradations", and when it is below the set value, it is set to "0", and when it is above the set value, it is binarized to "1". Then, noise components are removed (step 24). As a result, as shown in Fig. 3, a binary image divided into a container area (container area image) of "1" and a background area other than the container area is obtained, and among these, the container area image corresponding to "1" is The signal is supplied to the reduction circuit 15. In this area reduction circuit 15, as shown in step 25, in the container area that has become "1", a process is performed to remove a number of pixels, for example, 10 pixels, corresponding to the thickness of the container frame predicted in advance, from the container contour, and then the fourth area is removed. The reduced container area image shown by the solid line in the figure is extracted. Specifically, as shown in FIG. 5, this area reduction circuit 15 sets the pixel located at the center of, for example, nine pixels in each pixel as the pixel of interest A, and all of the eight pixels in its vicinity are set to "1". When,
The pixel value of the output pixel corresponding to the pixel of interest A in the input image is set as "1", and if even one of the eight neighboring pixels is "0", the pixel value of the output pixel corresponding to the pixel of interest A in the input image is set as "1". Set to “0”. If the above image processing is performed for one screen, one pixel can be removed from the outer contour side of the container region image, and if this is repeated 10 times, 10 pixels can be removed from the outer contour side of the container region image. FIG. 6 shows an example in which the center pixel of five pixels is set as the pixel of interest A. Then, the container area image shown by the solid line in FIG. 4, which has been reduced by the area reduction circuit 15, is sent to the subsequent image AND circuit 14 as a mask image. This image AND circuit 14 still has two images from the image memory 12.
A gradation perspective image that has not been digitized is input. Therefore, in this circuit 14, as shown in step 26, the mask image, which has been reduced by removing 10 pixels from the outer contour side of the container region image, and the gradation perspective image stored in the image memory 12 are placed at the same position. By facing each other and performing a logical product operation,
If the pixel of the gradation perspective image corresponding to the "1" pixel of the mask image is output as it is, and the pixel of the gradation perspective image corresponding to the "0" pixel of the mask image is output as "0", Container to be inspected 10c
Images of internal contents can be extracted. Then, the image of the contents extracted by the image AND circuit 14 is displayed on the result display device 16 in step 27, and it is determined here what contents are contained in the container to be inspected 10c. Alternatively, it can be determined automatically using a computer or the like.

Therefore, according to the configuration of the embodiment as described above,
Since the gradation perspective image stored in the image memory 12 is compared and binarized using a predetermined gradation in the binarization circuit 13, it is possible to accurately separate the container area and other background areas. . Also, 2
If the binarized image of the container area that has been subjected to the digitization process is removed by the area reduction circuit 15 by the number of pixels corresponding to the thickness of the container frame to create a reduced container area image, the container frame will essentially be in a state where it does not exist. In other words, the contents inside the container and the background portion remain. Therefore, if the image AND circuit 14 outputs the portion corresponding to the content as it is with its gradation value, and the portion corresponding to the background portion is output as "0", the content can be extracted accurately and reliably. Moreover, since the container is ignored and only the contents are displayed on the display device 16, there is one less recognition target, and the recognition time required can be shortened.

Note that the present invention is not limited to the above embodiments. In the above embodiment, a fluoroscopic image using X-rays was used, but other radiation such as α-rays, β-rays, γ-rays, etc.
It may also be possible to use radiation, infrared rays, or the like. In addition, the number of quantization levels and the number of samplings are 8-bit gradation (256 gradations) and 256 x 256 pixels as an example, but other quantization levels and sampling numbers may be used, or equivalent analog circuits may be used. good. Also,
Although the threshold value for the binarization process in the binarization circuit 13 is set to six gradations, it may be changed to other threshold values depending on changes in the density dynamic range of the X-ray fluoroscopic image. Further, in the above embodiment, the reduced pixel width of the container was described as 10 pixels, but other pixel widths may be used depending on the type and thickness of the container 10c to be inspected. In addition, the X-ray generator 10a
and the container to be inspected 10 between the X-ray television camera 10d
c is placed on a fixed or movable sample stage 10e, especially the latter movable one is placed on a sample stage 10e to be inspected.
The X-ray generator 10a detects the passage of 0c at a predetermined position.
In some cases, it is driven and controlled. In addition, the present invention can be implemented with various modifications without departing from the gist thereof.

〔Effect of the invention〕

As detailed above, according to the present invention, after removing the image of the container frame from the image of the container area, the process of extracting the contents inside the container is performed, so that the contents are not in contact with the container frame. Provided is a content determination device that can reliably separate and extract the contents from the container frame even when the container is in the container frame, and that only the contents are displayed so that the content can be determined reliably and contributes to shortening the determination time. can.

[Brief explanation of the drawing]

Figures 1 to 6 are shown to explain an embodiment of the content determination device according to the present invention.
Fig. 1 is a block diagram showing the configuration of the device of the present invention, Fig. 2 is a flowchart explaining the operation of the device, and Fig. 3 is a two-dimensional image obtained by binarizing an electrical gradation fluoroscopic image.
Figure 4 shows a binary image of a container reduced by removing the number of pixels by the thickness corresponding to the container frame from the container image, Figures 5 and 6 show container reduction, respectively. 7 and 8 are diagrams for explaining the problems of the conventional apparatus. 10...X-ray fluoroscope, 10a...X-ray generator, 10c...device to be inspected, 10d...X-ray television camera, 11...A/D conversion circuit, 12...image memory, 13...2 Value conversion circuit, 14... Image AND circuit, 15... Area reduction circuit, 16... Result display device, 17... Control device.

Claims (1)

[Scope of Claims] 1. In an apparatus for determining the contents inside the container by detecting a radiographic image transmitted and output from a container to be inspected by radiation irradiation with a radiation detector, data converting means for converting the analog signal into quantized gradation perspective image data and storing it in a memory, and comparing the gradation perspective image data stored by this means with a predetermined threshold value to obtain binary values. a container region extracting means for extracting an image of the reduced container region; and a container region image reduced by removing a desired number of pixels from the outline side of the container region image extracted by the container region extracting means. A container area reduction means for obtaining a container area, and a reduced container area image obtained by this means as a mask, perform a logical operation on this mask image and the gradation perspective image from the memory to create an image of the contents inside the container. A content determination device comprising a content extraction means for extracting the content. 2. The content determination device according to claim 1, wherein the radiation detector uses an X-ray television camera. 3. The content determination device according to claim 1, wherein the threshold value predetermined by the container area extraction means is a gradation value that makes the outer contour of the container clear. 4. The content determination device according to claim 1, wherein the image extracted by the container region extracting means is a binary image in which the container region image is set to "1" and the other region images are set to "0". 5. The container area reduction means sets the central pixel among the plurality of pixels as a pixel of interest, and changes the pixel of interest depending on whether all of the plurality of pixels in the vicinity of the pixel of interest include "1" or even one "0". Set the pixel value of the output image corresponding to "1" or "0", repeat this for one screen, and remove one pixel from the container contour,
The content determination device according to claim 1, wherein this process is repeated for a desired number of pixels.