JP2008232814A - Device for inspecting x-ray foreign matter - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a device for inspecting X-ray foreign matter capable of detecting foreign matter in an inspection target, without determining appended product as a foreign matter, regardless of the position of the appended product in the inspection target. <P>SOLUTION: In the device for inspecting X-ray foreign matter, body region searching operation is performed so as to determine whether the pixels P of an X-ray perspective image are pixels which constitute the body region of sub-content, on the basis of the relation between the gradation data of all of the pixels in the first searching unit which surrounds a single pixel P of the X-ray perspective image and a content level threshold (th), magnified sub-content region searching operation is successively performed to form the magnified sub-content region extracted image B3(x, y) of a magnified sub-content region and the gradation data of the respective pixels of the X-ray perspective image img (x, y), corresponding to a region, other than the magnified sub-content region of the magnified sub-content region extracted image B3 (x, y), is determined on the basis of a first foreign matter determining level so as to determine the presence of foreign matters. <P>COPYRIGHT: (C)2009,JPO&INPIT

Description

The present invention relates to an X-ray foreign substance inspection apparatus for taking an X-ray fluoroscopic image of an inspection object and inspecting the presence or absence of a foreign substance in the inspection object. The present invention relates to an X-ray foreign matter inspection apparatus that performs foreign matter inspection on an object to be inspected that includes sub-contents (attachments, etc.) that are obstacles to inspection.
Specifically, inspected items that contain accessories (sub-contents) such as oxygen scavengers, powdered soups, extra toys, etc. in contents (main contents) such as confectionery, processed foods, and medicines, Or, inspected objects in which bone (sub-contents) is mixed in part of raw meat or fish fillet (main contents), or buttons or fasteners (sub-contents) on clothes or bags (main contents) The present invention relates to an X-ray foreign matter inspection apparatus for an object to be inspected and the like.

  Conventionally, for example, an X-ray foreign substance inspection apparatus for detecting foreign matter (metal, bone, glass, stone, synthetic resin material, etc.) in an inspected object (including the surface) such as raw meat, fish, processed food, medicine, etc. Is used (see, for example, Patent Document 1). FIG. 16 is a diagram showing a partial configuration of the X-ray foreign substance inspection apparatus. In such an X-ray foreign matter inspection apparatus 110, X-rays are generated by the X-ray tube 1 disposed above the belt conveyor 5, and the X-rays are spread in a fan shape using a collimator (not shown). The X-ray beam is detected by a linear X-ray detector 3 including a large number of X-ray detection elements disposed below the upper conveyor belt 5b.

The belt conveyor 5 has a conveyor belt 5b wound between rollers 5a. Then, the inspection object 7 is placed on the upper surface of the upper conveyance belt 5b. By driving the roller 5a by a driving means (not shown) such as a motor, the conveyor belt 5b circulates and moves, and the inspection object 7 moves.
The collimator forms an X-ray beam that spreads in a fan shape so as to be orthogonal to the moving direction of the inspection object 7. The X-ray detector 3 is arranged so as to be orthogonal to the moving direction of the inspection object 7 so that the X-ray beam spreading in a fan shape can be detected.

  As a result, the X-ray foreign substance inspection apparatus 110 irradiates the moving inspection object 7 with X-rays, creates X-ray transmission image data of the inspection object 7 with the computer 109, and transmits the X-ray transmission image data. Are collected and arranged along the time axis, a two-dimensional X-ray fluoroscopic image can be obtained. At this time, if a foreign object exists in the inspection object 7 and the transmitted X-ray dose of the foreign object is different from the transmitted X-ray amount of the inspection object 7 itself, the foreign object appears as a shadow on the X-ray fluoroscopic image. Therefore, the operator can find a foreign object in the inspection object 7 by observing the X-ray fluoroscopic image displayed on the monitor screen.

  With such an X-ray foreign substance inspection apparatus, from contents such as raw meat, fish, processed food, etc., and clips for holding the contents contained (for example, metal clips that are fastened to seal both ends of the sausage) When determining whether or not there is a foreign object in the object to be inspected, a clip area is determined in advance by setting a mask area at a predetermined position where a clip is assumed to be present, and excluding the mask area from the inspection target. What prevents this is known (see, for example, Patent Document 2).

  However, in the X-ray foreign matter inspection apparatus, the presence or absence of foreign matter in the inspected body 7 consisting of the contents and accessories such as oxygen scavenger, powder soup and toys (bonus) packaged together with the contents is determined. In this case, when it is unclear at which position on the contents the attachment is present, the mask area cannot be set in the X-ray fluoroscopic image, and the attachment may not be determined as a foreign object. There was a problem that it was not possible.

In order to solve this, a division threshold value is set for dividing an attachment region corresponding to an attachment in the object to be inspected and an outside region of the attachment in the X-ray image generated based on the transmitted X-ray dose. Then, the accessory region and the non-attachment region are divided based on the pixel density of the X-ray image and the division threshold, and the pixel density of the X-ray image layer and a predetermined value are divided with respect to the X-ray image of the divided non-attachment region. It is disclosed that the presence or absence of a foreign object is determined based on a threshold value. Furthermore, it is disclosed that the presence / absence of a foreign object is determined based on another threshold value for an accessory region (see Patent Document 3).
JP 2006-275853 A JP 2002-243665 A JP 2006-308315 A

  According to the foreign substance inspection apparatus of Patent Document 3 described above, the inspection can be performed while excluding the position of the accessory as a mask region regardless of the position of the accessory that becomes an obstacle when inspecting the presence or absence of the foreign substance. In addition, it is possible to reduce erroneous detection in which an accessory is detected as a foreign object. Also, with respect to the accessory area, it is possible to detect foreign matter that overlaps with the accessory by adjusting the setting of the division threshold.

  However, in the method described in this patent document, it is an accessory region by comparing with the density (that is, gradation data) of each pixel on the basis of a preset level (division threshold). Or if it is a region outside the attachment, if some noise image of about a pixel unit appears in the X-ray fluoroscopic image, the noise image portion may be erroneously determined as the attachment region. There was a possibility of producing an erroneous detection result. For example, some attachments have protrusions on the periphery of the attachment, and erroneous detection that determines that the protrusion is not a part of the attachment but a foreign object may occur.

  This is not only the relationship between contents and attachments, but the relationship between the area you want to inspect (main contents) and the area that obstructs the inspection (sub-contents), such as the relationship between raw meat and bones. The same applies to the case of inspected objects mixed as contents.

  Therefore, the present invention includes main contents and sub-contents (hereinafter, the main contents and sub-contents include the relationship between the contents and attachments (in this case, the contents are the main contents). In addition to using only the gradation level (division threshold) as a reference, the other criteria are also used as the determination criteria. An object of the present invention is to provide an X-ray foreign substance inspection apparatus which can perform foreign substance inspection while removing the influence of noise.

  Further, in the present invention, when a sub-content area that becomes an obstacle to inspection is excluded as a mask area, a mask area is an area that is enlarged around the area where the sub-content is actually displayed, including the peripheral area. Thus, an object is to eliminate an error that erroneously detects an abnormal portion such as a protrusion irregularly generated around the subsidiary contents as a foreign object.

  In the X-ray foreign matter inspection apparatus of the present invention made to solve the above-mentioned problems, a sub-content region that becomes an obstacle to the foreign matter inspection for the main content is separated from the main content region, and a mask region similar to the conventional one is obtained. In the calculation, when the sub-content area is extracted, the sub-content area (and the same for the foreign substance area) is the same X having a certain area. Taking advantage of the presence of an extension (a set of pixels) of a region having line-level gradation data, the boundary is not determined based on only the threshold value of the gradation data of each pixel, but the level of adjacent pixels is determined. The key data is also used for determining the boundary, so that the boundary is determined while removing the influence of noise. Further, a boundary is defined by adding an operation for making the enlarged sub-content area including the peripheral area of the actual sub-content area a mask area.

  That is, the X-ray foreign substance inspection apparatus according to the present invention acquires a two-dimensional X-ray fluoroscopic image img (x, y) showing an inspection object including main contents and sub-contents, and performs this X-ray fluoroscopy. An X-ray foreign matter inspection apparatus for inspecting the presence or absence of foreign matter other than main contents and sub-contents in an inspection object by performing arithmetic processing on gradation data of each pixel forming an image. Content level for setting a gradation level value for dividing the gradation level corresponding to the transmitted X-ray dose of the object and the gradation level corresponding to the transmitted X-ray dose of the sub-contents as the content level threshold (th) As a first search unit (S1) used for calculation for extracting a main body region from a sub-content region shown in an X-ray fluoroscopic image, a threshold setting unit and a first mass region formed of a set of two-dimensional shapes of pixels A first search unit setting unit to be set, and an X-ray fluoroscopic image The first search unit is set so as to surround the pixel P with respect to the pixel P (x, y), and the relationship between the gradation data of all the pixels in the first search unit including the pixel P and the content level threshold value Based on the sub-contents, a main body region search calculation is performed to determine whether or not the pixel P is a pixel constituting the main body region of the sub-contents, and the main body region search calculation is performed for all the pixels of the fluoroscopic image. A sub-contents body region extraction image B2 (x, y) for extracting a sub-contents body region extraction image B2 (x, y), and a two-dimensional shape set of pixels, which is larger than the first lump region The second search unit used for the calculation for extracting the expanded sub-content area including the peripheral area of the main content area of the sub-contents shown in the sub-content main area extraction image B2 (x, y). Second search unit area set as (S2) A second search unit is set so as to surround the pixel Q with respect to one pixel Q (x, y) of the setting unit and the sub-contents body region extraction image B2 (x, y). Based on the comparison result between the gradation data of all the pixels in the second search unit and the gradation data of the sub-contents main body area shown in the sub-contents main body area extraction image B2 (x, y), the pixel Q is expanded An enlarged sub-content area search operation for determining whether or not the pixel is a content area is performed, and the expanded sub-content area search operation is performed for all the pixels of the sub-content body area extracted image B2 (x, y). Thus, the enlarged sub-content area extraction image B3 (x, y) obtained by extracting the enlarged sub-content area is used for the foreign substance determination of the enlarged sub-content area crest and the area other than the enlarged sub-content area. Set the gradation level value as the first foreign substance determination level value (th1) Each pixel of the X-ray fluoroscopic image img (x, y) corresponding to a region other than the enlarged sub-contents region of the first sub-contents region extraction image B3 (x, y). As the first inspection range, an enlarged sub-contents area foreign matter inspection unit that determines the presence or absence of foreign matter by determining gradation data of each pixel included in the first inspection range based on the first foreign matter determination level value I have to prepare.

  Here, “gradation data” refers to data representing the light and dark density of each pixel in the image. Typical gradation data is generally 256 gradation data in which light and dark densities are divided into 256 levels. The “gradation data” referred to here is data that can express an image density difference. If it is, it will not be limited to this. For example, a color display of pixel density is also handled as gradation data.

  In addition, the content level threshold (th) divides the X-ray dose that passes through the region of the main content only and the X-ray dose that passes through the region where the main content and the other (sub-contents and foreign matter) overlap. Is a threshold value used as a reference for this purpose, and is required to be a gradation level smaller than the gradation level corresponding to the transmitted X-ray dose that passes through the main contents (a gradation that is darker than the main contents region). It is necessary that the gradation level is higher than the gradation level corresponding to the X-ray dose transmitted through the overlapping region between the contents and the other (sub-contents and foreign matter) (brighter gradation than the overlapping region).

  In addition, the first search unit (S1) performs a small noise included in the image (for example, irregularly generated sub-contents) in the operation for extracting the main body area from the sub-content areas shown in the fluoroscopic image. This is a calculation range used for excluding the influence of projections and foreign matter), and consists of a set of two-dimensional shapes of pixels. The two-dimensional shape is preferably a rectangular pixel set or a circular pixel set, but other two-dimensional pixel sets may be used. If the first search unit (S1) has a square shape (square or rectangular), each pixel included in the calculation range can be easily expressed using xy coordinates. For example, in the case where the first search unit (S1) is square, N × N pixel sets (for example, x direction × y direction: 10 × 10 pixels) composed of a set of N pixels in the xy direction are arranged. It can be easily set as the first search unit (S1). In addition, when the first search unit (S1) is formed in a circular shape, the pixels included in the calculation range are complicated in that they are calculated using the xy coordinates, but are equal in all directions on the two-dimensional image. It is possible to set a simple calculation range.

  Further, the second search unit (S2) performs an operation for completely masking the sub-contents so as to include the peripheral region of the main content region of the sub-contents in the operation for extracting the expanded sub-contents region. The calculation range to be used is composed of a set of two-dimensional shapes of pixels, but this two-dimensional shape may also be a square pixel set as in the first search unit, or a pixel set of other shapes. There may be. In addition, in order to expand so that it may become larger than a main body area | region, an area | region larger than a 1st block area | region is set as a 2nd search unit (S2) (for example, a 1st search unit is x direction xy direction: 10 piece x In the case of 10 pixel sets, the second search unit sets x direction × y direction: 16 × 16 pixel sets). Further, the number of pixels arranged in the X direction and the number of pixels arranged in the y direction can be set individually, and the enlargement amount may be different vertically and horizontally.

According to the foreign matter inspection apparatus of the present invention, a two-dimensional X-ray fluoroscopic image img (x, y) showing an inspection object including main contents and sub contents is acquired.
Then, for each pixel P (x, y) of the X-ray fluoroscopic image, the sub-content main body region crest portion determines whether the pixel is a pixel constituting the sub-content main region. The “main body area search calculation” is performed. In short, this “sub-content body area” is an area in which only the portion near the center excluding the outer peripheral portion is extracted from the sub-content area shown in the image. It is an area extracted based on a determination formula used by the object body area extraction unit.
That is, the first search unit (S1) is set as the calculation range so as to surround the pixel P (x, y) for one pixel P (x, y) of the X-ray fluoroscopic image, and the pixel P (x, y) ), Whether or not the pixel P (x, y) is a pixel constituting the main content area of the sub-content based on the relationship between the gradation data of all the pixels in the first search unit (S1) including A body region search calculation is performed to determine
This calculation is applied to the gradation data of each pixel in the first search unit (S1) surrounding the pixel P (x, y) and the pixel P (x, y), and is binarized directly or described later. Whether the pixel P (x, y) is a pixel constituting the main body region of the sub-contents by comparing the gradation data of each pixel and the content level threshold value (th) indirectly using a function Determine whether or not.

  For example, as shown in FIG. 15A, a circular pixel set centered on a pixel P (x, y) on the fluoroscopic image img (x, y) is defined as the first search unit (S1). For each pixel in one search unit (S1), the gradation data is directly compared with the level threshold (th) for each content, and all the pixels in the first search unit (S1) (or more than a certain standard) are compared. The pixel P (x, y) is determined to be a pixel constituting the main body region of the sub-contents when the number of pixels) is gradation data (darker than the main content region) smaller than the content level threshold (th). To do. (As will be described later, a binarized image B1 may be created in advance based on the content level threshold (th), and the determination may be performed using the binarized image B1.)

  Then, by performing the body region search calculation performed on one pixel P (x, y) over all the pixels of the X-ray fluoroscopic image img (x, y), as shown in FIG. When there is a region U (that is, a sub-content region) having gradation data less than the content level threshold (th) on the fluoroscopic image img (x, y), A sub-content body area extracted image B2 (x, y) extracted as the sub-content body area V is created. As a result, a noise image portion (including a foreign matter portion and a protruding portion of the subsidiary content) smaller than the first search unit (S1) is completely erased on the subsidiary content body region extraction image B2 (x, y). As a result, only the central portion of the sub-content region U is extracted. The created sub-content body area extraction image B2 (x, y) is binarized so that the sub-content body area V is “1” and the other areas are “0”.

Subsequently, the expanded sub-content area heading portion is one pixel at a time for each pixel Q (x, y) of the sub-content main body area extraction image B2 (x, y), and the pixel is the expanded sub-content area. The “enlarged sub-content area search operation” is performed to determine whether or not the pixel is included in the image. In short, this “enlarged sub-content area” is obtained by enlarging an outer peripheral portion around this area with respect to the sub-content main area shown in the sub-content main area extraction image B2 (x, y). , An area in which the entire sub-content area is completely covered (the amount of enlargement is set so that irregular projections etc. appear in the sub-content) are included. This is an area that is extracted based on the determination formula used by the area head portion.
That is, for the pixel Q (x, y) of the sub-contents main body region extraction image B2 (x, y), the first region of the area larger than the first search unit so as to surround the pixel Q (x, y). The main body of the sub-contents set in the second search unit (S2) and the gradation data of all the pixels in the second search unit (S2) including the pixel Q and the sub-contents main body region extraction image B2 (x, y) Based on the comparison result with the gradation data of the region V (“1” when binarized), the enlarged sub-subject determines whether or not the pixel Q (x, y) is a pixel constituting the enlarged sub-content region. Perform content area search calculation.

For example, as shown in FIG. 15C, a circular pixel set centering on the pixel Q (x, y) on the sub-contents body region extraction image B2 (x, y) is defined as the second search unit (S2). (However, the region is larger than the first search unit (S1)), and for each pixel in the second search unit (S2), the gradation data and the gradation data of the main body region (when binarized) Are compared with each other, and when any one of the second search units (S2) (or the number of pixels below a certain reference) is the gradation data of the main body region V, the pixel Q (x, y ) Is determined to be a pixel constituting the enlarged sub-content area.
In addition, since the area | region expanded as the area | region of 2nd search unit (S2) becomes large becomes large, in the case of the to-be-inspected object in which protrusion parts, such as a protrusion, generate | occur | produce in a subcontent, the magnitude | size of the protrusion which can generate | occur | produce In consideration of the above, by setting the area of the second search unit (S2), even if a protrusion occurs, it is included in the enlarged sub-content area.

  Then, the enlarged sub-content area search operation performed on one pixel Q (x, y) is performed over all the pixels of the sub-content body area extraction image B2 (x, y), thereby obtaining the result shown in FIG. ), The sub-content body area V is enlarged, and an enlarged sub-content area extraction image B3 (x, y) extracted as the enlarged sub-content area W is created. As a result, the sub-content area is completely covered on the enlarged sub-content area extraction image B3 (x, y), including the protrusions. The created enlarged sub-content area extraction image B3 (x, y) is binarized so that the area of the enlarged sub-content area W is “1” and the other areas are “0”.

  Subsequently, the enlarged sub-contents region outside foreign matter inspection unit performs an X-ray fluoroscopic image img (x, y) corresponding to a region other than the enlarged sub-contents region W of the enlarged sub-contents region extraction image B3 (x, y). Each pixel is determined as the first inspection range, and the gradation data of each pixel included in the first inspection range is determined based on the first foreign matter determination level value (th1), thereby determining the presence or absence of foreign matter.

  As described above, according to the X-ray foreign substance inspection apparatus of the present invention, not only the X-ray content level threshold (th) is used as a reference but also the set of pixels included in the first search unit (S1). By including the gradation data together in the determination criterion, it is possible to create a mask area while removing the influence of small noise (protrusions of sub-contents, foreign matter) and perform foreign matter inspection.

Further, according to the foreign matter inspection apparatus of the present invention, when the sub-content area that becomes an obstacle to the inspection is excluded as a mask area, the area around the area where the sub-content is actually displayed is enlarged including the peripheral area. Since the enlarged sub-content area is used as a mask area, it is possible to eliminate an error that erroneously detects an abnormal portion such as a protrusion irregularly generated around the sub-content as a foreign object.
Further, it is possible to completely eliminate an error that erroneously determines that a foreign substance is in the inspection range outside the enlarged sub-contents area as a part of the sub-contents.

(Means and effects for solving other problems)
In the above invention, the binarized image extraction unit for creating the binarized image B1 (x, y) obtained by binarizing the fluoroscopic image img (x, y) based on the content level threshold (th) is further provided. The sub-contents main body region extraction unit includes one pixel P of the binarized image B1 (x, y) instead of one pixel P (x, y) of the X-ray fluoroscopic image img (x, y). Sub-contents obtained by extracting the main body region of the sub-contents by performing the main body region search calculation on (x, y) and performing the main body region search calculation on all the pixels of the binarized image B1 (x, y). The main body region extraction image B2 (x, y) may be created.

For example, the following expression (5) can be used as an expression for binarization.

    According to the present invention, the binarized image B1 (x, y) is created in advance based on the content level threshold (th), and the body region search is performed using the binarized image B1 (x, y). Do the calculation. As a result, the body region search calculation using the binarization calculation can be performed.

  In the above invention, the sub-contents main body region crest portion is such that the gradation data of all the pixels in the first search unit (S1) including the pixel P (x, y) is equal to or less than the content level threshold (th). On the condition that the pixel P (x, y) is a pixel that constitutes the main content region of the sub-contents, a main body region search calculation is performed. The pixel Q (x, y) is determined to be an enlarged sub-content area on condition that the gray-scale data of at least one pixel in the second search unit (S2) including the sub-content main body area You may make it perform the expansion subcontents area search calculation to perform.

  According to the present invention, when one pixel P (x, y) is less than or equal to the content level threshold (th) in the surrounding first search unit, the pixel P (x, y) Since it is determined as the sub-contents main body region, the small noise portion is surely removed, and only the central portion of the sub-contents region is extracted as the main body region. Furthermore, if at least one of the pixels of the second search unit around the one pixel Q (x, y) is a sub-contents main body region, the sub-contents main body region is determined to be an enlarged sub-contents region. Therefore, it is possible to enlarge only the central portion (main body region portion) of the subsidiary content to be an enlarged subsidiary content region, and by appropriately setting the second search unit, the projecting portion of the subsidiary content is also It is possible to extract an enlarged content region that completely covers the entire sub-contents included.

In the above invention, the sub-contents main body region starting portion may perform a main body region search calculation using the following equation (1).

According to the present invention, using the binarized image B1 (x, y), a rectangular first search unit (S1) having x and y directions in the vertical and horizontal directions (x direction × y direction: N × N) ) When a binarization operation is performed in the operation range, and all the pixels in the first search unit are “1”, B2 (x, y) is “1”, that is, a sub-content body area is formed. Pixel.
Thereby, it can be determined whether it is a main body area | region by simple calculation using the square 1st search unit (S1).

In the above invention, the enlarged sub-content area starting portion may perform an enlarged sub-content area search operation using the following equation (2).

According to the present invention, the second search unit (S2) having the x and y directions in the vertical and horizontal directions (x direction x y direction: kx pieces) using the sub-content body region extraction image B2 (x, y). When binarization is performed with the calculation range within (× ky), and any one pixel in the second search unit is “1”, B3 (x, y) is “1”, that is, enlargement This is a pixel constituting the sub-content area.
As a result, it is possible to determine whether or not the region is the enlarged sub-content region by simple calculation using the square second search unit (S2).

In the above invention, the foreign substance inspection unit outside the enlarged sub-content area may determine the presence or absence of a foreign substance outside the enlarged sub-content area using the following equation (3).

According to the present invention, the entire sub-content area is calculated using the enlarged sub-content area extraction image B3 (x, y), and any one pixel outside the enlarged sub-content area is the content. When it is below the object level threshold, it is determined that M1 is other than “0”, that is, there is a foreign object.
As a result, it is possible to determine whether or not the area is the enlarged sub-content area by simple calculation.

    In the above invention, a second foreign substance determination level value setting unit that sets a gradation level value used for foreign substance determination for the entire region of the fluoroscopic image img (x, y) as a second foreign substance determination level value (th2); A third mass region composed of a set of two-dimensional shapes of pixels is set as a third search unit (S3) used for calculation for extracting a foreign substance region from the entire region of the fluoroscopic image img (x, y). In this state, the number of pixels used as a determination criterion when performing foreign object recognition based on the relationship between the gradation data of all the pixels included in the third search unit (S3) and the second foreign object determination level value (th2) The pixel number threshold setting unit for setting the threshold (MS3) and the pixel R (x, y) of the X-ray fluoroscopic image Img (x, y) are arranged so as to surround the pixel R (x, y). Three search units (S3) are set and pixel R Based on the number of pixels extracted by comparing the gradation data of each pixel in the third search unit (S3) including x, y) and the second foreign substance determination level value (th2) and the pixel number threshold (MS3). Then, a foreign substance region search calculation is performed to determine whether or not the pixel R (x, y) is a pixel constituting a foreign substance, and all the pixels of the X-ray fluoroscopic image img (x, y) are set as the second inspection range. A whole area foreign substance inspection unit that determines the presence or absence of foreign substances by performing a foreign substance area search operation for each pixel in the second inspection range may be further provided.

Here, the third search unit (S3) is a calculation range used to exclude the influence of small noise included in the image in the calculation for extracting the foreign substance region, and is composed of a set of two-dimensional shapes of pixels. . The two-dimensional shape is preferably a rectangular pixel set or a circular pixel set, like the first search unit and the second search unit, but the shape is not particularly limited. For example, as the third search unit (S3), a pixel set of x direction × y direction: 7 × 7 (total of 49) is set. The setting of the third search unit (S3) may be changed to an arbitrary value or may be a fixed value.
On the other hand, the pixel number threshold value (MS3) is a threshold value set as the number with respect to the total number of pixels included in the third search unit (S3). For example, if the pixel number threshold is set to 20, and if the third search unit is set to 49, 20/49 is a criterion for determining whether or not it is a foreign object region.

According to the present invention, the entire region foreign matter inspection unit performs a foreign matter region search operation on the pixels included in the entire region of the X-ray fluoroscopic image img (x, y).
That is, the third search unit (S3) is set so as to surround the pixel R (x, y) for one pixel R (x, y) of the X-ray fluoroscopic image img (x, y). Then, the gradation data of each pixel in the third search unit (S3) including the pixel R (x, y) is compared with the second foreign matter determination level value (th2) (the level below the second foreign matter determination level). Pixels of tone data (that is, pixels having dark gradation data equal to or lower than the second foreign substance determination level value) are extracted, and the number of extracted pixels is calculated, and the calculated number of pixels and the pixel number threshold (MS3) ) And a pixel number threshold (MS3) or more is detected, it is determined that the pixel R (x, y) is a pixel that constitutes a foreign substance region, and otherwise no foreign matter area is constituted. By performing such a foreign substance region search operation performed on the pixel R (x, y) for all the pixels of img (x, y) of the number of fluoroscopic images, The presence or absence of foreign matter is determined.
As a result, in the inspection outside the enlarged sub-content area by the foreign substance inspection section outside the enlarged sub-content area, there was an enlarged sub-content area excluded from the inspection as a mask area. The entire inspection object can be inspected. In addition, the inspection is continuously performed in two modes by the enlarged sub-contents region foreign matter inspection unit and the entire region foreign matter inspection unit, and based on thresholds suitable for the outside of the enlarged sub-content region and the enlarged sub-content region, respectively. The inspection can be performed with high accuracy.

Here, the whole area foreign matter inspection unit may perform foreign matter search calculation of the entire image using the following equation (4).

According to the present invention, the foreign object region search calculation is performed on one pixel R (x, y) of the X-ray fluoroscopic image img (x, y). That is, the gradation data of each pixel and the second foreign object determination are set in the rectangular third search unit (S3) (x direction × y direction: M × M) whose vertical and horizontal directions are the x direction and the y direction. An operation of comparing the level value with the level value (th2) is performed, and the number of pixels equal to or smaller than the second foreign matter determination level value (th2) is calculated. When the calculated number of pixels is equal to or less than the pixel number threshold value (MS3), one pixel R (x, y) is the binarized data of the pixels (pixel R (x, y) constituting the foreign substance region. “1”), and other than that, it is determined that the pixel that does not form a foreign object (the binarized data of the pixel R (x, y) is “0”). Such a foreign substance region search operation is performed over the pixels of the entire region of the X-ray fluoroscopic image img (x, y), and when at least one pixel is a pixel constituting the foreign substance region, that is, the value of M2 is “0”. ", It is determined that there is a foreign object.
Thereby, the presence or absence of the foreign substance region can be determined by a simple calculation using the square third search unit (S3).

  Further, in the above invention, a missing determination threshold value setting unit that sets a fourth mass region formed of a set of two-dimensional shapes of pixels as a missing determination threshold value (S4) used in a calculation for determining the loss of sub-contents; The sub-contents for determining the presence / absence of the sub-contents in the inspected object based on the main content area of the sub-contents shown in the sub-contents main body region extraction image B2 (x, y) and the lack determination threshold (S4) You may further provide a missing inspection part.

Here, the missing determination threshold (S4) is a threshold that is compared with the area extracted as the main body area of the sub-content, and specifically, an area that is slightly smaller than the area that can be extracted as the main-body area of the sub-content. (Pixel set) is set.
According to the present invention, the subsidiary content missing inspection unit compares the body content region of the subsidiary content shown in the subsidiary content body region extracted image B2 (x, y) with the region of the missing determination threshold (S4). When the main body area is smaller than the missing determination threshold (S4), or when there is no main body area, it is determined that the sub-content is missing in the inspection object. Further, when there are a plurality of regions larger than the missing determination threshold (S4), it is determined that an extra sub-content is included.
As a result, it is possible to find an inspection object in which the sub-contents that should originally exist are insufficient or too much.

Hereinafter, embodiments of the present invention will be described with reference to the drawings. In addition, this invention is not limited to the following embodiment, It cannot be overemphasized that various aspects are included in the range which does not deviate from the meaning of this invention.
In the following embodiments, the inspected object includes contents (main contents) and attachments (sub contents). There are exceptionally defective products with missing attachments. Further, a product that does not contain foreign matter in the contents is a non-defective product, and a product that contains foreign matter in exceptional cases is a defective product.

(Device configuration)
FIG. 1 is a perspective view showing an appearance of an X-ray foreign substance inspection apparatus according to an embodiment of the present invention, and FIG. 2 is a schematic block diagram showing a partial configuration of the X-ray foreign substance inspection apparatus.
The X-ray foreign substance inspection apparatus 80 includes a main body 70, an entrance belt conveyor 60, an outlet belt conveyor 50, and a control system (computer) 20 that controls the entire X-ray foreign substance inspection apparatus 80.
The inlet belt conveyor 60 is provided at the inlet portion 6 a of the main body 70, and the outlet belt conveyor 50 is provided at the outlet portion 6 b of the main body 70.

  The main body 70 includes a protection box 6 having an inlet 6a and an outlet 6b, an X-ray tube 1 that generates X-rays, an X-ray detector 3 that detects X-rays from the X-ray tube 1, and an X-ray A belt conveyor (conveying mechanism) 5 for conveying the inspection object 7 so that the inspection object 7 passes between the tube 1 and the X-ray detector 3 and a drive mechanism (not shown) for driving the belt conveyor 5. Z)).

  The protective box 6 is a frame that supports the entire structure, and puts the inspection object 7 from the inlet portion 6a and takes out the inspection object 7 from the outlet portion 6b. Further, a radiation shielding material is used so that a harmful amount of X-rays does not leak from the inside of the protection box 6 to the outside of the protection box 6. In addition, protective entrances 8 are lowered at the entrance 6 a and the exit 6 b so that harmful amounts of X-rays do not leak from the inside of the protection box 6 to the outside of the protection box 6.

  The belt conveyor 5 has a conveyor belt 5b that is wound between a pair of rollers 5a. Then, the inspection object 7 is placed on the upper surface of the upper conveyor belt 5b. When one of the rollers 5a is driven by the drive mechanism, the conveyor belt 5b circulates and moves, and the object 7 to be inspected moves from the inlet 6a to the outlet 6b.

  The drive mechanism has, for example, a rotation drive motor. The drive mechanism is controlled by receiving a drive signal output from a transport mechanism control unit 34 described later.

  The X-ray tube 1 is arranged at the upper part of the protection box 6 and emits X-rays toward the X-ray detector 3 arranged at the lower part of the protection box 6. At this time, the generated X-ray is formed as an X-ray beam spreading in a fan shape so as to be orthogonal to the moving direction of the inspection object 7 moved by the belt conveyor 5 using a collimator (not shown).

FIG. 3 is a plan view of the X-ray detector, and FIG. 4 is a cross-sectional view taken along line AA shown in FIG. The X-ray detector 3 includes a phosphor 91 that converts X-rays into light, a photodiode array 92, and a support 93. The X-ray detector 3 is disposed below the upper conveyor belt 5b and is disposed above the lower conveyor belt 5b. The photodiode array 92 is formed by arranging a large number (eg, 256) of photodiodes (92a, 92b,...) Constituting one pixel (1ch) in a straight line. This is converted into a signal (detection signal). This signal is output to the computer 20.
At this time, the X-ray detector 3 is arranged so as to be orthogonal to the moving direction of the inspection object 7 to be moved by the belt conveyor 5 so that the fan-shaped X-ray beam from the X-ray tube can be detected.

  The computer 20 includes, as hardware, a CPU 21 that performs control for realizing various functions, a keyboard 22a and a mouse 22b as input devices, a display device 23 having a monitor screen 23a, and a memory 25 (RAM, ROM, HDD, etc.). Prepare.

  In order to explain the functions processed by the CPU 21, each function will be described as a block. The setting unit 31, the X-ray transmission image data storage control unit 32, the X-ray transmission image data display unit 33, and the transport mechanism control unit 34. A binarized image extraction unit 35, a sub-contents main body region extraction unit 36, an enlarged sub-contents region extraction unit 37, an enlarged sub-contents region outside inspection unit 38, an entire region inspection unit 39, And a content missing inspection unit 40.

The memory 25 also stores a setting data storage area 41 for storing setting data, and an X-ray fluoroscopic image (X-ray transmission image data A) created based on the detection signal acquired by the X-ray detector 3. Various images (binarized image B1, sub-contents main body region extraction image B2, and enlarged sub-contents region extraction image B3 described later) created from the detection signal storage region 42, X-ray fluoroscopic image (X-ray transmission image data A) ) Is stored.
The setting data storage area 41 includes a content level threshold (th), a first search unit (S1), a second search unit (S2), a third search unit (S3), a missing determination threshold (S4), and a first foreign object. The setting values of the determination level value (th1), the second foreign substance determination level value (th2), and the pixel number threshold value (MS3) are stored. By changing the setting value as necessary, the changed setting value is It is supposed to be memorized.

The specific setting contents of these setting data will be described. FIG. 5 is a diagram illustrating an example of an input monitor screen 23a displayed as an image for inputting (changing settings) or confirming setting data. FIG. 6 is a schematic diagram showing an example of an X-ray fluoroscopic image displayed as an image.
In some input monitor screens, in order to make it easier for an operator to input setting data, the name of the setting data is replaced with an expression that is easy for the operator to understand. For this reason, the setting contents stored in the setting data storage area 41 will be described with respect to data in which the setting data name displayed on the input monitor screen is different from the setting data name in the setting data storage area 41.
For convenience of explanation, it is assumed that a normal inspection object 7 is composed of contents and one accessory that is an oxygen scavenger enclosed in a packaging bag. The gradation level of the content displayed on the X-ray fluoroscopic image is set to 150. Further, it is assumed that the gradation level of the oxygen scavenger is 40 and the size is 60 × 60.

  The “content level threshold (th)” divides gradation data indicating the content 81 (gradation level: 150) and gradation data indicating the accessory 82 (gradation level: 40) and the foreign object 83. It is a threshold value used for the binarization calculation, and is a threshold value when extracting the binarized image B1 from the X-ray fluoroscopic image. Specifically, a gradation level intermediate between the gradation level “150” of the contents and the gradation level “40” of the accessory part, for example, the gradation level “50” is set.

  “Calculation range (S1)” sets the first search unit (S1). The first search unit (S1) is a range formed by a set of two-dimensional shapes of pixels. This is a calculation range when performing the “main body region search calculation” for extracting the sub-contents main body region using the binarized image B1 described above, and in the calculation, the small foreign matter 83 and the deoxygenation that is the accessory 82 This is setting data for excluding the influence of the protruding portion 82b of the agent. Here, a square pixel set is set. For example, if “10” is input, “10 × 10” is set, and 100 square pixel sets (x direction × y direction: 10 × 10) are set.

  As a result of the binarization operation using the content level threshold (th) and the main body region search operation using the first search unit (S1), the main body portion 82a of the oxygen scavenger (sub-content) which is the attachment 82 The sub-contents main body region extraction image B2 in which (exactly the central portion excluding the periphery of the main body portion 82a) is extracted is created.

  “Enlarged range horizontal (kx), vertical (ky)” sets the second search unit (S2). The second search unit (S2) is a range formed by a set of two-dimensional shapes of pixels larger than the first search unit (S1) described above. This is an enlarged sub-content area extraction image obtained by extracting an enlarged sub-content area (82 ') that covers the entire attachment 82 (main body portion 82a and protrusion 82b) using the sub-content main area extraction image B2. This is setting data for setting a calculation range when performing “enlarged sub-content area search calculation” for obtaining B3. Here, two values are set so that the enlargement range can be individually adjusted vertically and horizontally. For example, when “16” and “16” are set, “horizontal × vertical: 16 × 16” is set, and 256 pixel sets are set.

  “Attachment range (S4)” sets a missing determination threshold (S4). The missing determination threshold value (S4) is a threshold value for determining the presence / absence of the accessory 82 in the inspection object 7, which is a two-dimensional set of pixels. In the “attachment mode” described later, the presence / absence of the accessory 82 is determined. It is a threshold value used for the calculation which determines. A value slightly smaller than the size of the sub-contents (size: 60 × 60), for example, “50 × 50”.

  The “first foreign matter determination level value (th1)” is a foreign matter outside the enlarged subsidiary content area that covers the entire attachment 82 (main body portion 82a and protruding portion 82b) on the enlarged subsidiary content region extraction image B3. This is a threshold of the gradation level when performing inspection (referred to as “inspection mode 1”). For example, gradation level “100” is set. Thereby, the presence / absence of the foreign matter 83 outside the enlarged sub-content area is determined in the inspection mode 1 described later.

  The “second foreign matter determination level value (th2)” is a gradation level threshold value when foreign matter inspection is performed on the entire region of the X-ray fluoroscopic image (referred to as “inspection mode 2”). 30 "is set. Thereby, in the inspection mode 2 described later, the presence / absence of the foreign matter 83 in the entire region is determined, and in particular, the presence / absence of the foreign matter 83 is determined in the overlapping region where the attachment 82 overlaps the contents 81.

The “pixel number range (MS3 / S3)” sets a third search unit (S3) and a pixel number threshold (MS3).
The third search unit S3 is a range formed by a set of two-dimensional shapes of pixels. The third search unit (S3) is a calculation range used for excluding the influence of small noise included in the image in the calculation for extracting a foreign substance region in “examination mode 2” to be described later. Here, it is assumed that a square pixel set of “7 × 7” (total “49”) is set. In addition, this setting is a fixed value (it may be settable, but it is a fixed value because setting of the pixel number threshold (MS3) is sufficient).
The pixel number threshold value (MS3) is a threshold value set as the number with respect to the total number of pixels (that is, “49”) included in the third search unit (S3). For example, when the pixel number threshold is set to “20”, if the third search unit is set to “49”, “20/49” is a criterion for determining whether or not it is a foreign object region. This is used when the presence / absence of the foreign matter 83 is determined in inspection mode 2 to be described later.

Next, each functional block of the CPU will be described.
The setting unit 31 uses the input monitor screen shown in FIG. 5 to display the initial setting values of the various setting data described above, and prompts the operator to input or change the final set data. Control to store in the setting data storage area 41 is performed.

  The X-ray transmission image data storage control unit 32 performs control to acquire a signal (X-ray transmission image data A) from the X-ray detector 1 and store the signal in the detection signal storage area 42 along with the acquired time t. .

Based on the X-ray transmission image data A, the X-ray transmission image data display unit 33 performs control to display the X-ray transmission image img (x, y) on the monitor screen 23a. Since the X-ray transmission image data A is one-dimensional data at a certain moment, the X-ray transmission image data A is arranged along the time axis so that the two-dimensional X-ray transmission of the inspection object 7 is performed. An image img (x, y) is obtained. At this time, if the foreign object 83 exists in the inspection object 7 and the X-ray transmission amount of the foreign object 83 is different from the X-ray transmission amount of the inspection object 7 itself, the foreign object 83 appears in the X-ray fluoroscopic image img (x, y). Will appear as a shade.
In the X-ray fluoroscopic image img (x, y), a two-dimensional xy coordinate system is set, but an arrangement of one pixel (1ch) in order is an x axis (0ch to 255ch), and a time axis is y. It is defined to be axial. The y direction may be changed according to the length of the object to be inspected, but here, for convenience of explanation, it is assumed that the y axis is also (0 ch to 255 ch).

  The transport mechanism control unit 34 performs control to output a drive signal for driving the belt conveyor 5 to the drive mechanism by receiving input signals input by various operations of the keyboard 22a and the mouse 22b.

The binarized image extraction unit 35 creates a binarized image B1 (x, y) that is binarized based on the content level threshold (th) for the fluoroscopic image img (x, y). I do.
Specifically, binarization of each pixel of the X-ray fluoroscopic image img (x, y) is performed using Expression (5), and a binarized image B1 (x, y) is extracted.

  In the binarized image B1 (x, y), a pixel “1” is displayed at a gradation level 0 (dark), and a pixel “0” is displayed at a gradation level 255 (bright).

  The sub-content main body region extraction unit 36 performs the first search so as to surround the pixel P (x, y) with the one pixel P (x, y) as the center in the binarized image B1 (x, y). The unit (S1) is set, and the binarized data of all the pixels in the first search unit (S1) including the pixel P (x, y) is “1” (that is, gradation level 0 (dark)). As a condition, a body region search calculation is performed to determine that one pixel P (x, y) is a pixel constituting the body region of the subsidiary content.

Then, by performing a body region search operation on all pixels on the binarized image B1 (x, y) (or all pixels on the X-ray fluoroscopic image img (x, y)), Control is performed to create a sub-contents body region extraction image B2 (x, y) from which the body region is extracted.
Specifically, the body region search calculation is performed using equation (1).

In the created sub-contents body region extraction image B2 (x, y), the “1” pixel is displayed at the gradation level 0 (dark), and the “0” pixel is displayed at the gradation level 255 (bright). Shall be.
As a result, on the subsidiary content body region extraction image B2 (x, y), the body region of the subsidiary content that is an accessory (the central portion of the accessory) is displayed in black, and the others are displayed in white.
(It should be noted that the operations from the binarization calculation to the main body region search calculation described above are performed with the pixel P (x, x) centered on one pixel P (x, y) in the X-ray fluoroscopic image img (x, y). , Y) is set to surround the first search unit (S1), and the gradation data of all the pixels in the first search unit (S1) including the pixel P (x, y) is the content level threshold (th). (This is substantially the same as performing a body region search operation for determining the pixel P (x, y) as a pixel constituting the body region of the subsidiary contents on condition that

  The enlarged sub-content area crest portion 37 surrounds the pixel Q (x, y) around the one pixel Q (x, y) in the sub-content body area extraction image B2 (x, y). Is set to the second search unit (S2), and any one of the binarized data of all the pixels in the second search unit (S2) including the pixel Q (x, y) is “1” (that is, the floor). On the condition that the tone level is 0 (dark)), an enlarged sub-content area search operation for determining that one pixel Q (x, y) is a pixel constituting the enlarged sub-content area is performed.

Then, an enlarged sub-content area extraction image obtained by extracting an enlarged sub-content area by performing an enlarged sub-content area search operation on all the pixels on the sub-content main body area extraction image B2 (x, y). B3 (x, y) is created.
Specifically, the body region search calculation is performed using equation (2).

In the created enlarged sub-content region extraction image B3 (x, y), the “1” pixel is displayed at the gradation level 0 (dark), and the “0” pixel is displayed at the gradation level 255 (bright). Shall be.
As a result, on the enlarged sub-content area extraction image B3 (x, y), the main body portion (the center portion of the accessory) of the sub-content that is an accessory, the protruding portion of the sub-content, and the peripheral portion of the sub-content The enlarged sub-content area including the sub-contents is displayed in black, and other areas are displayed in white, and a black portion (mask area) is formed so as to cover the entire sub-content area.

The enlarged sub-content area non-foreign particle inspection unit 38 includes an X-ray fluoroscopic image img (x, x, y) corresponding to an area other than the enlarged sub-content area (ie, mask area) of the enlarged sub-content area extraction image B3 (x, y). Control for determining the presence / absence of a foreign object by determining each pixel of y) as the first inspection range and determining the gradation data of each pixel included in the first inspection range based on the first foreign object determination level value (th1). I do.
Specifically, the presence / absence of a foreign substance is determined using Expression (3).

The whole area foreign matter inspection unit 39 includes a third search unit (S3) so as to surround the pixel R (x, y) with respect to one pixel R (x, y) in the fluoroscopic image img (x, y). And the number of pixels extracted by comparing the gradation data of each pixel in the third search unit (S3) including the pixel R with the second foreign substance determination level value (th2) and the pixel number threshold (NS3), Based on the above, a foreign substance region search calculation is performed to determine whether or not the pixel R (x, y) is a pixel constituting the foreign substance. Then, control is performed to determine the presence / absence of a foreign object by performing a foreign object region search operation for each pixel in the second inspection range, with all pixels of the X-ray fluoroscopic image img (x, y) as the second inspection range.
Specifically, the presence / absence of a foreign substance is determined using Expression (4).

    The subsidiary content missing inspection unit 40 compares the body content of the subsidiary content shown in the subsidiary content body region extracted image B2 (x, y) with the missing determination threshold (S4), thereby determining that the body region is missing. When the area is smaller than the threshold (S4), or when there is no main body area, control is performed to determine that the sub-contents are missing in the inspection object. In addition, when there are a plurality of areas larger than the missing determination threshold (S4), control is performed to determine that the subsidiary contents are included in excess.

(Operation procedure for foreign matter inspection)
Next, an operation procedure of foreign matter inspection by the X-ray foreign matter inspection apparatus 80 will be described. First, a large flow of foreign object inspection will be described. FIG. 7 is a flowchart for explaining the flow of the entire operation procedure. FIG. 8 is a schematic diagram of an image created during the inspection.

First, in the process of s10, an X-ray fluoroscopic image img (x, y) is created (FIG. 8A).
Next, in the process of s11, based on the X-ray fluoroscopic image img (x, y), a binarized image B1 (x, y) is created by Expression (5) (FIG. 8B).
Next, in the process of s12, the sub-contents main body region extraction image B2 (x, y) is created by Expression (1) based on the binarized image B1 (x, y) (FIG. 8C). ).
Next, in the process of s13, based on the content main body region extraction image B2 (x, y), an enlarged sub-content region extraction image B3 (x, y) is created by Expression (2) (FIG. 8D )).
Next, in s14, based on the enlarged sub-content area extraction image B3 (x, y), a foreign object inspection (inspection mode 1) outside the enlarged sub-content area is performed by Expression (3).
Next, in s15, the foreign substance inspection of the entire region is performed by Expression (4) based on the X-ray fluoroscopic image img (x, y). Here, the foreign matter inspection (inspection mode 2) of the sub-content area not inspected in s14 is performed.
Next, in s16, based on the sub-contents main body region extraction image B2 (x, y), an inspection for sub-contents is performed.

  The above is the basic flow. By selecting the inspection mode (operation of a mode setting button (not shown)), only inspection mode 1 can be carried out alone, inspection mode 2 or sub-content missing inspection Or may be performed alone.

  Next, a specific operation procedure will be described. 9 to 14 are diagrams showing a specific operation flow from the creation of the fluoroscopic image to the end of the foreign substance inspection.

First, in the process of step S101, a test start signal is input by various operations of the operator's keyboard 22a and mouse 22b.
Next, in step S102, in order to acquire a signal (X-ray transmission image data A) from the X-ray detector 3 and create an X-ray fluoroscopic image of the inspection object 7 along with the acquired time t. The necessary predetermined time interval (for example, the length of the inspection object 7) is stored in the detection signal storage area 42. Then, the stored signals (X-ray transmission image data A) are arranged in time series along the y direction to create an X-ray fluoroscopic image img (x, y).

Next, in order to create a binarized image B1 (x, y) obtained by binarizing the fluoroscopic image img (x, y), one pixel P (x, y) is inspected in the process of step S103. A point P is set, and x = 0 and y = 0 are set as initial values of the position parameter of the inspection point P.
Next, in the process of step S104, in the X-ray fluoroscopic image img (x, y), the inspection point P (initially, the pixel at the coordinates (0, 0) is the inspection point P) is the content level threshold (th) (floor It is determined whether or not the key level 50 is already set) or higher.
If it is determined that the content level threshold value (th) is less than the threshold value (th), the process proceeds to step S105 to set the value of the binarized image B1 (x, y) at the inspection point P (x, y) to “1”. "
On the other hand, if it is determined that the content level threshold value (th) is equal to or greater than the threshold value (th), the process proceeds to step S106 to obtain the value of the binarized image B1 (x, y) at the inspection point P (x, y). “0”.

Next, in step S107, it is determined whether x = 255 (the coordinate of the last pixel in the x direction). If it is determined that x is not 255, the process proceeds to step S108, where x = x + 1 is stored, and the process returns to step S104. That is, the processing of steps S104 to S107 is repeatedly executed until it is determined that x = 255.
On the other hand, if it is determined that x = 255, it is determined in the process of step S109 whether y = 255 (the coordinates of the last pixel in the y direction). If it is determined that y is not 255, the process proceeds to step S110 to store y = y + 1, and the process returns to step S104. That is, the processes in steps S104 to S109 are repeatedly executed until it is determined that y = 255.
When the processing at all inspection points P is executed, the values (“1” or “0”) of all the pixels of the binarized image B1 (x, y) are determined, and the binarized image B1. (X, y) is completed.

  If it is determined that y = 255, the “main body region search calculation” for searching the main body region of the sub-contents is executed. Therefore, in the process of step S111, as the initial value of the position parameter of the inspection point P Store x = 0, y = 0.

Next, in the processing of step S112, 10 is centered on the inspection point P (x + i, y + j) (i = 0, j = 0) in the binarized image B1 (x, y) according to the equation (1). It is determined whether or not all the values are “1” in the first search unit (S1) of × 10 square, that is, the calculation range (i = −5 to 4, j = −5 to 4).
If it is determined that all the values are “1”, the process proceeds to the process of step S113, and the value is determined at the inspection point P (x, y) in the sub-contents body region extraction image B2 (x, y). “1” (indicating that the pixel in the sub-content body area is an inspection point).
On the other hand, if it is determined that a certain value is “0”, the process proceeds to the process of step S114, whereby a value is obtained at the inspection point P (x, y) in the sub-contents body region extracted image B2 (x, y). Is “0” (indicating that a pixel outside the sub-contents main body area is an inspection point).

Next, in the process of step S115, it is determined whether x = 255. If it is determined that x is not 255, the process proceeds to step S116, where x = x + 1 is stored, and the process returns to step S112. That is, the processes in steps S112 to S115 are repeatedly executed until it is determined that x = 255.
On the other hand, if it is determined that x = 255, it is determined in the process of step S117 whether y = 255. If it is determined that y is not 255, the process proceeds to step S118, whereby y = y + 1 is stored, and the process returns to step S112. That is, the processes in steps S112 to S117 are repeatedly executed until it is determined that y = 255.
When the processing of all the inspection points is executed, the values (“1” or “0”) of all the pixels of the sub-contents main body region extraction image B2 (x, y) are determined. The region extraction image B2 (x, y) is completed.

  On the other hand, if it is determined that y = 255, an “expanded sub-content area search operation” for searching for an expanded sub-content area is executed, so that in the process of step S119, the initial position parameter of the inspection point Q is initialized. As a value, x = 0 and y = 0 are stored.

Next, in the process of step S120, the center of the inspection point Q (x + i, t + j) (i = 0, j = 0) in the enlarged sub-content area extraction image B3 (x, y) is obtained by the equation (2). It is determined whether any one value is “1” in the second search unit (S2), that is, the expansion ranges kx and kt (i = −8 to 7, j = −8 to 7). .
If it is determined that any one of the values is “1”, the process proceeds to step S121 so that the inspection point Q (x, y) in the enlarged sub-content area extracted image B3 (x, y) is obtained. The value is “1” (indicating that the pixel in the enlarged sub-content region 82 ′ is an inspection point).
On the other hand, if it is determined that all the values are “0”, the process proceeds to step S122, so that the inspection point Q (x, y) in the enlarged sub-content area extraction image B3 (x, y) is obtained. The value is “0” (indicating that pixels other than the enlarged sub-content region 82 ′ are inspection points).

Next, in the process of step S123, it is determined whether x = 255. If it is determined that x is not 255, the process proceeds to step S124 so that x = x + 1 is stored, and the process returns to step S120. That is, the processes of steps S120 to S123 are repeatedly executed until it is determined that x = 255.
Next, when it is determined that x = 255, it is determined whether or not y = 255 in the process of step S125. If it is determined that y is not 255, the process proceeds to step S126 so that y = y + 1 is stored, and the process returns to step S120. That is, the processes of steps S120 to S125 are repeatedly executed until it is determined that y = 255.
When the processing of all inspection points is executed, the values (“1” or “0”) of all the pixels of the enlarged sub-content area extraction image B3 (x, y) are determined, and the enlarged sub-contents are determined. The region extraction image B3 (x, y) is completed.
The

  If it is determined that y = 255, the foreign object determination is started. That is, in the process of step S127, “inspection mode 1” is executed. Although details of the inspection mode 1 will be described later, the presence / absence of the foreign matter 83 in the inspection object 7 is determined for pixels other than the enlarged sub-content area.

  Next, in the process of step S128, “inspection mode 2” is executed. Although details of the inspection mode 2 will be described later, the presence / absence of the foreign matter 83 in the inspection object 7 is determined for all areas including the enlarged sub-content area.

  Next, in the process of step S129, the “attachment mode” is executed. Although details of the accessory mode will be described later, the presence / absence of the accessory 82 in the inspection object 7 is determined.

Next, in the process of step S130, it is determined whether or not an examination end signal has been input by various operations of the operator's keyboard 22a and mouse 22b.
If the inspection end signal has not been input, the process returns to step S102.
On the other hand, when the inspection end signal is input, this flowchart is ended.

Next, the “inspection mode 1” process (s127) will be described. FIG. 12 is a flowchart for explaining the inspection mode 1.
Here, based on the enlarged sub-content region extraction image B3 (x, y), the fluoroscopic image img (x, y), and the first foreign object determination level value (th1) (tone level 100), the expression (3 ) To determine the presence or absence of the foreign matter 83 in the inspection object 7.

    First, in the process of step S201, x = 0 and y = 0 are stored as initial values of the position parameter of the inspection point R (x, y).

Next, in the process of step S202, in order to check whether the inspection point R (x, y) is inside the enlarged sub-contents region or outside the enlarged sub-contents region, the enlarged sub-contents region at the inspection point R (x, y) The value of the extracted image B3 (x, y) is determined.
If it is determined that the value of the inspection point R (x, y) is “1”, the process proceeds to step S205 to add “+0” to the value of M1 (see Expression (3)).
On the other hand, if it is determined that the value of the inspection point R (x, y) is “0”, the process proceeds to step S203, and the fluoroscopic image img (x, y) at the inspection point R (x, y). ) Gradation data is less than or equal to the first foreign matter determination level value (th1).
If it is determined that the first foreign matter determination level value (th1) has been exceeded, the process proceeds to step S205 to add “+0” to the value of M1.
On the other hand, if it is determined that the value is equal to or less than the first foreign substance determination level value (th1), the process proceeds to step S204 to add “+1” to the value of M1.

Next, in step S206, it is determined whether x = 255. If it is determined that x is not 255, the process proceeds to step S207, where x = x + 1 is stored, and the process returns to step S202. That is, the processes in steps S202 to S206 are repeatedly executed until it is determined that x = 255.
On the other hand, if it is determined that x = 255, it is determined whether y = 255 in the process of step S208. If it is determined that y is not 255, the process proceeds to step S209 to store y = y + 1, and the process returns to step S202. That is, the processes in steps S202 to S208 are repeatedly executed until it is determined that y = 255.
If it is determined that y = 255, this flowchart is terminated.

Next, the “inspection mode 2” process (s128) will be described. FIG. 13 is a flowchart for explaining the inspection mode 2.
Here, the fluoroscopic image img (x, y), the second foreign substance determination level value (th1) (gradation level 30), the pixel number range (pixel number threshold (MS3) / third search unit (S3) = 20 / 49), the calculation of the equation (4) is performed to determine the presence or absence of the foreign matter 83 in the inspection object 7.

First, in the process of step S301, x = 0 and y = 0 are stored as initial values of the position parameter of the inspection point R (x, y).
Next, in the process of step S302, 49 (7 × 7) centered on the inspection point R (x + i, y + j) (i = 0, j = 0) in the fluoroscopic image img (x, y). In the third search unit (S3) composed of pixels, that is, among the pixels in the calculation range (i = -3 to 3, j = -3 to 3), the gradation data is equal to or less than the second foreign substance determination level value (th2). It is determined whether or not the number of pixels is equal to or greater than “20”, which is a pixel number threshold (MS3).
If it is determined that the number equal to or less than the second determination level value (th2) is equal to or greater than the pixel number threshold (MS3), the process proceeds to step S303, and the value of M2 (see Expression (4)) is obtained. Add "+1".
On the other hand, when it is determined that the number equal to or smaller than the second determination level value (th2) is less than the pixel number threshold (MS3), the process proceeds to step S304, and “+0” is added to the value of M2. .

Next, in the process of step S305, it is determined whether x = 255. If it is determined that x is not 255, the process proceeds to step S306, where x = x + 1 is stored, and the process returns to step S302. That is, the processes in steps S302 to S305 are repeatedly executed until it is determined that x = 255.
On the other hand, if it is determined that x = 255, it is determined whether y = 255 in the process of step S307. If it is determined that y is not 255, the process proceeds to step S308 to store y = y + 1, and the process returns to step S302. That is, the processes in steps S302 to S307 are repeatedly executed until it is determined that y = 255.
When it is determined that y = 255, the foreign object inspection in the inspection mode 2 is completed, and then comprehensive determination including the inspection mode 1 and the inspection mode 2 is performed.

That is, in the process of step S309, it is determined whether M1 + M2 = 0.
If it is determined that M1 + M2 = 0, the process proceeds to step S310 to determine that there is no foreign object in the inspection object 7.
On the other hand, if it is determined that M1 + M2 = 0 is not satisfied, the process proceeds to step S311 to determine that there is a foreign object in the inspection object 7.
Next, when the process of step S310 or S311 is completed, this flowchart is ended.

Next, the “attachment mode” process (s129) will be described. FIG. 14 is a flowchart for explaining the accessory mode.
Here, calculation is performed based on the sub-contents main body region extraction image B2 (x, y) and the accessory range (S4) (50 × 50), and the accessory (sub-content) (60 It is determined whether or not x60) is missing.

First, in the process of step S401, the attachment determination unit 39 sets the area (B2 (x, y)) determined as the sub-content body area on the sub-content body area extraction image B2 (x, y) to “1”. It is determined whether the range of “)” is equal to or greater than the accessory range (S4), that is, the missing determination threshold (S4).
If it is determined that the size is greater than or equal to the attachment range (S4), the process proceeds to step S402 to determine that the attachment 82 exists in the inspection object 7.
On the other hand, when it is determined that it is less than the accessory range (S3), it is determined that the accessory 82 does not exist in the inspection object 7 by proceeding to the process of step S403.
Next, when the process of step S402 or S403 is completed, this flowchart is ended.

As described above, according to the X-ray foreign substance inspection apparatus 80, the enlarged sub-area in which the region corresponding to the attachment 82 (sub-contents) in the inspection object 7 is enlarged in the X-ray fluoroscopic image img (x, y). The content area 82 ′ (mask area) and the outside of the enlarged sub-content area are divided, and the first determination level value (th1) (ie, bright gradation) having a large gradation level outside the enlarged sub-content area. Level) as a reference, the presence or absence of the foreign object 83 is determined, so that an error for erroneously determining the accessory 82 as the foreign object 83 does not occur regardless of the position of the accessory 82 in the inspection object 7, and the object is accurately covered. The foreign object 83 in the inspection object 7 can be detected.
At this time, since the enlarged sub-content region 82 ′ enlarged from the actual sub-content is used as a mask region, for example, when the oxygen scavenger is an accessory, part of the iron oxide powder in the oxygen scavenger Even in the case where is distributed in the shape of protrusions protruding from other iron oxide powders, the entire attachment 82 in the inspection object 7 can be surely excluded from the inspection object.
Further, since the accessory 82 present at the position overlapping the foreign object 83 and the other accessory 82 are different in gradation data (that is, X-ray transmission amount), based on the second determination level value (th2), The presence or absence of the foreign matter 83 that overlaps the attachment 82 in the inspection object 7 can be determined.
Further, it is possible to determine the lack of the attachment 82 regardless of the position of the attachment 82 in the inspection object 7.

  INDUSTRIAL APPLICATION This invention can be utilized for the X-ray foreign material inspection apparatus which detects the foreign material in a to-be-inspected object from an X-ray fluoroscopic image when X-rays are irradiated to to-be-inspected objects, such as raw meat, fish, processed food, and medicine. .

The perspective view which shows the external appearance of the X-ray foreign material inspection apparatus which is one Embodiment of this invention. The block diagram which shows the structure of a part of X-ray foreign material inspection apparatus. The top view of a X-ray detector. Sectional drawing of the AA shown in FIG. The figure which shows an example of the monitor screen displayed as an image in order to set setting data. FIG. 4 is a schematic diagram of a monitor screen on which an X-ray fluoroscopic image is displayed. The flowchart which shows the big flow of the operation | movement procedure of a foreign material inspection. Schematic diagram of each image created during foreign object inspection. The flowchart which shows an example of the flow of the operation | movement procedure of a foreign material inspection. The flowchart which shows an example of the flow of the operation | movement procedure of a foreign material inspection. The flowchart which shows an example of the flow of the operation | movement procedure of a foreign material inspection. 5 is a flowchart showing an operation procedure in inspection mode 1; 7 is a flowchart showing an operation procedure in inspection mode 2; The flowchart which shows the operation | movement procedure in attachment mode. The figure explaining the state which produces a subcontents main body area | region extraction screen and an expansion subcontents area | region extraction screen from a X-ray fluoroscopic image. The figure which shows the structure of the conventional foreign material inspection apparatus.

Explanation of symbols

1: X-ray source 3: X-ray detector 5: Belt conveyor (conveyance mechanism)
7: Inspected object 20: Control system (computer)
22: Input device 23: Display device 25: Memory (setting data storage unit)
31: Setting unit 35: Binarized image extracting unit 36: Sub-contents body region extracting unit 37: Enlarged sub-contents region extracting unit 38: Enlarged sub-contents region foreign matter inspection unit 39: Whole region foreign matter inspection unit 40: Sub-content missing inspection unit 41: setting data storage area 42: detection signal storage area 43: image storage area

Claims (9)

A two-dimensional X-ray fluoroscopic image img (x, y) reflecting the inspection object including the main contents and the sub-contents is acquired, and the gradation data of each pixel forming the X-ray fluoroscopic image is obtained. An X-ray foreign matter inspection apparatus for inspecting the presence or absence of foreign matters other than the main contents and sub-contents in the inspected object by performing arithmetic processing,
Contents for setting the gradation level value for dividing the gradation level corresponding to the transmitted X-ray dose of the main content and the gradation level corresponding to the transmitted X-ray dose of the sub-content as the content level threshold (th) An object level threshold setting unit;
A first search that sets a first block region composed of a set of two-dimensional shapes of pixels as a first search unit (S1) used in an operation for extracting a main body region from sub-content regions shown in an X-ray fluoroscopic image A unit setting section;
For a pixel P (x, y) of an X-ray fluoroscopic image, a first search unit is set so as to surround the pixel P, and gradation data and contents of all pixels in the first search unit including the pixel P Based on the relationship with the object level threshold value, a body area search calculation is performed to determine whether or not the pixel P is a pixel constituting the body area of the subsidiary content, and the body area search calculation is performed for all the pixels of the fluoroscopic image. A sub-contents main body region cuffing section for creating a sub-contents main body region extraction image B2 (x, y) obtained by extracting the sub-contents main body region;
A second lump area that is a set of two-dimensional shapes of pixels and that is larger than the first lump area includes the peripheral area of the sub-contents main body area shown in the sub-contents main body area extraction image B2 (x, y) A second search unit region setting unit that is set as a second search unit (S2) used in the calculation for extracting the expanded sub-content region;
A second search unit is set so as to surround the pixel Q with respect to one pixel Q (x, y) of the sub-contents main body region extraction image B2 (x, y), and within the second search unit including the pixel Q Based on the comparison result between the gradation data of all the pixels and the gradation data of the main body region of the subsidiary content shown in the subsidiary content body region extraction image B2 (x, y), the pixel Q defines the enlarged subsidiary content region. An enlarged sub-content area search operation for determining whether or not the pixel is a constituent pixel is performed, and the enlarged sub-content area search operation is performed on all the pixels of the sub-content body area extracted image B2 (x, y) to expand An enlarged sub-content area extracting section for creating an enlarged sub-content area extraction image B3 (x, y) obtained by extracting the sub-content area;
A first foreign matter determination level value setting unit that sets a gradation level value used for foreign matter determination in a region other than the enlarged sub-content region as a first foreign matter determination level value (th1);
Each pixel of the X-ray fluoroscopic image img (x, y) corresponding to an area other than the enlarged sub-content area of the enlarged sub-content area extraction image B3 (x, y) is set as the first inspection area. An X-ray foreign substance inspection comprising: an enlarged sub-contents out-of-area foreign substance inspection unit that determines the presence or absence of a foreign substance by determining gradation data of each pixel included based on a first foreign substance determination level value apparatus. A binarized image extraction unit that creates a binarized image B1 (x, y) obtained by binarizing the fluoroscopic image img (x, y) based on the content level threshold (th);
The sub-contents body region extraction unit replaces one pixel P (x, y) of the X-ray fluoroscopic image img (x, y) with one pixel P (x) of the binarized image B1 (x, y). , Y), the main body area search calculation is performed, and the main body area search calculation is performed for all the pixels of the binarized image B1 (x, y), thereby extracting the sub-contents of the sub-contents. 2. The X-ray foreign substance inspection apparatus according to claim 1, wherein the main body region extraction image B2 (x, y) is created. The sub-contents main body region cuffing portion sets the pixel P to the sub-contents on condition that gradation data of all pixels in the first search unit including the pixel P is equal to or less than a content level threshold (th). Perform body area search calculation to determine the pixels constituting the body area,
The enlarged sub-content area cuff expands the pixel Q on condition that the gradation data of at least one pixel in the second search unit including the pixel Q is the gradation data of the sub-content body area. The X-ray foreign substance inspection apparatus according to claim 1, wherein an enlarged sub-content area search operation for determining a sub-content area is performed.
The X-ray foreign substance inspection apparatus according to claim 3, wherein the sub-contents main body region protuberance performs a main body region search operation using the following equation (1).
5. The X-ray foreign substance inspection apparatus according to claim 1, wherein the expanded sub-content area crest portion performs an enlarged sub-content area search operation using the following equation (2): .
6. The foreign matter inspection unit outside the enlarged sub-content area determines whether or not there is a foreign substance outside the enlarged sub-content area using the following equation (3). X-ray foreign substance inspection device.
A second foreign matter determination level value setting unit that sets a gradation level value used for foreign matter determination for the entire region of the X-ray fluoroscopic image img (x, y) as a second foreign matter determination level value (th2);
A third lump area composed of a set of two-dimensional shapes of pixels is set as a third search unit (S3) used for calculation to extract a foreign substance area from the entire area of the fluoroscopic image img (x, y). In this state, a pixel number threshold value (MS3) is set as a criterion for foreign object recognition based on the relationship between the gradation data of all pixels included in the third search unit and the second foreign substance determination level value. A pixel number threshold setting unit;
A third search unit (S3) is set so as to surround the pixel R for one pixel R (x, y) of the X-ray fluoroscopic image img (x, y). Whether the pixel R is a pixel constituting a foreign object based on the number of pixels extracted by comparing the gradation data of each pixel and the second foreign substance determination level value (th2) and the pixel number threshold (MS3) The entire foreign matter region search calculation is performed to determine whether or not there is a foreign matter by performing the foreign matter region search calculation for each pixel in the second inspection range with all pixels of the fluoroscopic image as the second inspection range. The X-ray foreign substance inspection apparatus according to claim 1, further comprising an area foreign substance inspection unit. The X-ray foreign substance inspection apparatus according to claim 7, wherein the whole area foreign substance inspection unit performs a foreign substance area search calculation using the following equation (4).
A missing determination threshold value setting unit that sets a fourth mass region formed of a set of two-dimensional shapes of pixels as a missing determination threshold value (S4) used for calculation for determining the loss of sub-contents;
Sub-contents missing for determining presence / absence of sub-contents in inspected object based on sub-contents main body region shown in sub-contents main body region extraction image B2 (x, y) and lack determination threshold (S4) The X-ray foreign matter inspection apparatus according to claim 1, further comprising an inspection unit.