JP2000121580A - X-ray measuring apparatus for can seam part and detecting method for measurement starting point - Google Patents

X-ray measuring apparatus for can seam part and detecting method for measurement starting point

Info

Publication number
JP2000121580A
JP2000121580A JP11224079A JP22407999A JP2000121580A JP 2000121580 A JP2000121580 A JP 2000121580A JP 11224079 A JP11224079 A JP 11224079A JP 22407999 A JP22407999 A JP 22407999A JP 2000121580 A JP2000121580 A JP 2000121580A
Authority
JP
Japan
Prior art keywords
image
measured
dictionary
video camera
input
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Pending
Application number
JP11224079A
Other languages
Japanese (ja)
Inventor
Hiroshi Nakayama
Yoshimi Otake
Atsushi Sodeyama
Yuji Takamoto
博司 中山
義美 大武
敦史 袖山
雄治 高本
Original Assignee
Hitachi Eng Co Ltd
Hitachi Ltd
Sapporo Breweries Ltd
サッポロビール株式会社
日立エンジニアリング株式会社
株式会社日立製作所
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Priority to JP10-226879 priority Critical
Priority to JP22687998 priority
Application filed by Hitachi Eng Co Ltd, Hitachi Ltd, Sapporo Breweries Ltd, サッポロビール株式会社, 日立エンジニアリング株式会社, 株式会社日立製作所 filed Critical Hitachi Eng Co Ltd
Priority to JP11224079A priority patent/JP2000121580A/en
Publication of JP2000121580A publication Critical patent/JP2000121580A/en
Application status is Pending legal-status Critical

Links

Abstract

PROBLEM TO BE SOLVED: To provide an X-ray measuring apparatus for a can seam part, in which the measurement starting point of the can seam part is detected by using the feature part of a can to be measured, photographed by a video camera, in which an irregularity in the measurement starting point is eliminated even when a plurality of cans of the same can type are measured and by which the dimension of the can seam part can be measured precisely. SOLUTION: A video camera 6 is arranged so as to face the feature part of a can 1 to be measured. In a comparison and judgment part 822 inside a video image processing part 82, an extraction image 11 which is extracted from the feature part of a can photographed by the video camera 6 and a dictionary image 10 which is prepared by extracting a feature from an image on the can as the feature part of the can 1 to be measured, photographed in advance are compared and judged. On the basis of a judged result by the comparison and judgment part, a driving device 5 which receives a signal from a drive control device 9 turns the can 1, to be measured, up to a measurement starting point which is set by using the dictionary image 10. Thereby, the measurement starting point of a can seam part can be detected without irregularities, and the dimension of the can be measured precisely.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an apparatus for measuring the dimensions of a canned part of a can by X-ray photography, and more particularly to an apparatus for measuring the dimensions of a canned part based on a detected dimension measurement start point. And a measurement start point detection method.

[0002]

2. Description of the Related Art An X-ray is radiated in a tangential direction to a sealed portion of a can in which a can main body and a can lid are tightly wound to seal the can. A can-tightening measuring device for obtaining the dimensions of each predetermined portion of a tightened portion by image processing an X-ray image to be obtained is disclosed as a prior art in Japanese Patent Application Laid-Open No. 1-305346. In detecting the measurement start point from the circumference, the measurement start point is marked with a writing instrument and optically read to realize the measurement start point.

[0003] However, the work of marking the measurement start point must be performed for each can to be measured, which is troublesome and difficult to apply to an in-line apparatus. There is a problem that the reproducibility of the points varies.

[0004]

SUMMARY OF THE INVENTION It is an object of the present invention to detect a measurement starting point by using a characteristic partial image of a can taken by a video camera, and to determine a measurement starting point even when a plurality of cans of the same can type are measured. Provided are an X-ray can-clamped-portion measuring device and a measurement start point detecting method capable of accurately measuring the dimension of a can-clamped portion without variation.

[0005]

Means for Solving the Problems In order to achieve the above object, a feature of the measuring apparatus for measuring the tightening portion of an X-ray can according to the present invention is that a video camera is arranged opposite to the characteristic portion of the can to be measured. The comparison determination unit of the video image processing unit compares and determines the extracted image extracted from the can feature portion photographed by the video camera and the dictionary image created by extracting the feature from the can feature image of the can measured beforehand. And rotating the measured can by a drive device controlled by a drive control device based on the determination result of the comparison determination unit.

Specifically, the present invention provides the following apparatus and method.

The present invention comprises an X-ray irradiating device for irradiating X-rays, an image receiving device comprising an image receiving device for receiving the irradiated X-rays, and an image processing device, wherein the X-ray irradiating device and the image receiving device An X-ray can-clamping-section measuring device having a driving device for rotatably positioning a tightening section of the can to be measured at the measurement position by the X-ray irradiation, wherein the can to be measured arranged at the measurement position A video camera is disposed so as to face the characteristic portion of the above, and the image processing apparatus is configured to process an X-ray image processing section for processing an X-ray image of the tightening section obtained by the irradiation, and the object photographed by the video camera. A video image processing unit for processing a characteristic portion image of the measuring can, wherein the video image processing unit includes an extracted image extracted from the captured characteristic portion image and a characteristic portion of the measured can previously measured. Dictionary created by extracting features from An X-ray can-tightening device having a comparison / determination unit for comparing and determining an image, and a drive control device for controlling the driving device for rotating the can to be measured based on the determination result of the comparison / determination unit. A part measuring device is provided.

[0008] Preferably, the video image processing section is provided with a dictionary table for defining the dictionary image and the measurement start point.

Further, according to the present invention, a can to be measured is mounted on a can holder rotatable by a driving device, and the measured can is measured from a video camera arranged to face a characteristic portion of the mounted can to be measured. A video image of the characteristic portion of the can is input to the image processing apparatus connected to the output side of the video camera, and an image created by extracting the characteristic from the input image is a dictionary image. The dictionary image and data in which a measurement start point is set from the input image are registered in advance in the dictionary table of the image processing apparatus as a dictionary of the measured can, and the can in which the measured can is placed While the pedestal is rotated by the driving device receiving the signal from the drive control device, the extracted image obtained by extracting the feature from the image of the can feature portion input by the video camera is compared with the dictionary image, and a normalized correlation process is performed. The performed, based on a result of calculating the correlation values, said providing a measurement start point detecting method and detecting the measurement starting point from the circumference of the object to be measured can.

Further, according to the present invention, a can to be measured is mounted on a can holder rotatable by a driving device, and the measured can is measured from a video camera arranged to face a characteristic portion of the mounted can to be measured. A video image of the characteristic portion of the can is input to the image processing apparatus connected to the output side of the video camera, and an image created by extracting features from the input image is used as a reference image. The video is obtained by rotating the can receiver by a predetermined angle by the driving device that has received a signal from a drive control device with reference to the position of the can receiver when the reference image is created. An image created by extracting a feature from each image of a can feature portion input by a camera is acquired as a plurality of dictionary images, and the acquired reference image and the plurality of dictionary images, the reference image and the plurality of features are acquired. The data of the rotation angle from the reference image at the time of acquiring each image is registered in advance in the dictionary table of the image processing apparatus as a type dictionary for measurement for each type of the can to be measured, and the The video image of the mounted can to be measured is input by the video camera, the extracted image obtained by extracting the feature from the image of the characteristic portion of the can to be measured and the reference image and the plurality of reference images and the plurality of target object type dictionaries. A normalized correlation process is performed by comparing the dictionary image with the dictionary image, and the rotation angle data of the reference image or the dictionary image having the maximum correlation value among the reference image and the plurality of dictionary images for each of the measurement target type dictionaries is obtained. A measurement start point detection method for detecting a measurement start point from the circumference of the measured can based on the measurement start point.

Preferably, a signal is output from the drive control device to rotate the can cradle to the measurement start point registered in the dictionary table.

Here, the "characteristic part of the can to be measured" refers to the characteristic of the can, such as the shape of the can lid including the pull tab, the shape of the can body including the side seam and the picture, as described in the embodiment. Means the part to be represented.

[0013]

BEST MODE FOR CARRYING OUT THE INVENTION Hereinafter, an X-ray can tightening portion measuring apparatus and a measuring start point detecting method according to an embodiment of the present invention will be described.
This will be described with reference to the drawings.

FIG. 1 is a block diagram of an embodiment of X according to an embodiment of the present invention.
FIG. 2 shows a plan view of a can lid which is a characteristic portion of the can to be measured in FIG. 1 and which is a portion for detecting a measurement start point. The characteristic portion here is the shape of the pull tab 30 provided on the can lid, the pattern drawn on the can lid, and the shape of the can lid itself.

The measuring device for the X-ray can-clamping section includes a can cradle 2 on which the measured can 1 is placed in contact with the bottom surface of the can 1 to be measured and a can cradle 2 placed on the can cradle 2. And a can positioning device 2a for positioning the can to be measured 1
And a drive device 5 that can be driven to move forward, backward, left, right, up and down, and can be driven to rotate, and X-rays that irradiate X-rays in the tangential direction to the tightening portion of the can 1 to be measured positioned at the center of the can support 2. An irradiation device 3, a receiver 4 opposed to the X-ray irradiation device 3 for receiving an X-ray image of the tightening unit, and a top surface of the can, which is a characteristic part of the can 1 to be measured placed on the can receiver 2, are moved back and forth. A video camera 6 that can be driven to move left, right, up and down, captures an image, an image processing device 8 that processes the received X-ray image and the captured can upper surface image, and a driving device 5 based on data output from the image processing device 8. And a drive control device 9 for controlling.

The image processing device 8 includes an X-ray image processing unit 81
, A video image processing unit 82, a data processing unit 83, and a storage unit 84.

The X-ray image processing unit 81 processes the X-ray image of the tightening unit sent from the receiver 4.

The video image processing section 82 includes a dictionary table 821 storing, as a dictionary image 10, an image obtained by extracting features from a can upper surface image, which is a characteristic portion of the can 1 to be measured, in advance, and a video camera 5. And a comparison / determination unit 822 for comparing and determining the extracted image 11 extracted from the upper surface of the can, which is a characteristic part of the measured can 1, and the dictionary image 10, for example, a normalized correlation process.

The data processing unit 83 calculates the rotation amount of the can 1 to be measured based on the determination result of the comparison determination unit 822, and outputs the rotation amount data. The data processing unit 83 outputs a drive command to the drive control device 9 based on the result of the X-ray image data processing from the X-ray image processing unit 81.

The storage unit 84 stores a processing program, X-ray image data, processing results of the data processing unit 83, and the like.

The drive control unit 9 includes a drive control processing unit 91
, A drive control storage unit 92, a motor controller 94
And an X, Y, O axis driver 95 and a position detection unit 96.

The drive control processing section 91 includes a data processing section 83
The drive device 5 is controlled via the motor controller 94 and the X, Y, and O-axis drivers 95 based on the rotation amount data output from.

The position detecting section 96 obtains position data of the measured can 1 of the can receiving stand 2 based on a signal from the driving device 5.

The drive control storage unit 92 stores a processing program, position data acquired by the position detection unit 96, and the like.

Next, the operation of the image processing device 8 and the drive control device 9 for detecting the measurement start point will be described.

An image of the upper surface of the can, which is a characteristic part of the can 1 to be measured, positioned at the center of the can pedestal 2 by the positioning device 2a is inputted from the video camera 6, and the input image is inputted to the video image processing section 82 of the image processing device 8. It is taken in. The video camera 6 is configured to be movable back and forth and left and right with respect to the upper surface of the can, which is a characteristic part of the can 1 to be measured.

Comparison / decision unit 82 in video image processing unit 82
Reference numeral 2 denotes a dictionary image corresponding to the can upper surface, which is a characteristic part of the can 1 to be measured, which is called from a dictionary table 821 previously arranged on a memory in the video image processing unit 82, and is extracted from an input image of the video camera 6. A normalized correlation process is performed with the extracted image. If the correlation value is not the predetermined value (99% or more), the driving device 5 is driven by the signal from the driving control device 9, that is, the rotation amount data, into the can cradle 2 on which the measured can 1 is mounted.
, And the rotation operation is repeatedly performed until the correlation value becomes equal to or more than the predetermined value.

FIG. 3 shows a dictionary image 10 extracted as a feature from the can upper surface input image 31 which is a feature portion of the can 1 to be measured, which is previously registered as a dictionary in a dictionary table 821 on a memory in the video image processing unit 82. It is a conceptual diagram.

When the characteristics of each can type are captured, the difference in the shape of the pull tab in the can lid portion can be effectively used. Therefore, the can 1 to be measured is placed so that the X-ray irradiation point 32 and the pull tab 30 intersect vertically. The can receiver 2 is rotated by the driving device 5 that has received a signal from the drive control device 9, and the can upper surface input image 31, which is a characteristic part of the can 1 to be measured, is acquired by the video camera 6, and the can upper surface input image is obtained. A region is defined for the pull-tab 31 so as to surround the pull-tab 30 portion, and a dictionary image 10 of the measured can 1 is extracted in that region. At the same time, a measurement reference point 33 and a measurement start point 34 are determined based on the can upper surface input image 31. Dictionary table 82
1 is registered as a dictionary (data).

FIGS. 4 to 6 show flowcharts for detecting the measurement start point when the pull-tab at the can lid is used as a feature in the X-ray can-clamping section measuring device of FIG.

First, in step 41, the video camera 6 performs an image input process (step 411 in FIG. 5) and a feature extraction process (412) of the can upper surface which is a characteristic portion of the can 1 to be measured to obtain an extracted image 11. I do. Next, a normalized correlation process (413) is performed between the dictionary image 10 and the extracted image 11 corresponding to the measured can 1, and a correlation value calculation process (414) is performed.

Next, whether or not the calculated correlation value is a specified value (99% or more) is compared (401), and the specified value (99% or more) is compared.
If the above is not the case, the rotation operation of the driving device 5 for rotating the can receiving table 2 by the signal from the driving control device 9 is performed, for example, at every rotation angle of 12 ° (402). This rotation angle may be any angle, and is preferably 10 ° to 15 °.

Returning to step 41, the process is repeated until the measured can 1 has been rotated all around (403).

If the correlation value does not reach the specified value (99% or more), the highest correlation value among the obtained correlation values is set as a provisional reference value (404). 2 is rotated from the provisional reference value to, for example, a position of −30 ° from the provisional reference value (405). After rotation,
The image input processing (411) and the characteristic extraction processing (41) of the upper surface of the can that is the characteristic part of the can 1 to be measured by the video camera 6 again
2) is performed to obtain the extracted image 11.

Next, the dictionary image 10 corresponding to the measured can 1
A normalized correlation process (413) is performed between the extracted image 11 and the extracted image 11, and a correlation value calculation process (414) is performed. If the calculated correlation value does not become the specified value (99% or more), the can receiver 2 is rotated by the driving device 5 by the signal from the drive control device 9 (406), and the process returns to step 41 again to return the measured can 1 Performs image input and feature extraction processing on the upper surface of the can. At this point, since the can cradle 2 does not need to be rotated all around, the rotation angle is set to 3 ° or less so that the measurement reference point can be more easily detected (407). For example, when the rotation angle is set to 3 °, the can 1 to be measured is rotated 20 times within the range of 60 °, and the image input / feature extraction processing is repeated each time (408).

The extracted image 11 when the correlation value becomes a specified value (99% or more) becomes the same as the dictionary image 10, so that the reference position is set and the measurement reference point 33 is set (409). Next, the measurement start point setting process of FIG. 6 is performed.

After setting the reference position, in step 421, the data processing unit 83 of the image processing device 8 sets the measurement reference point 3
The angle difference from the measurement start point 34 from the measurement start point 3 is calculated, a signal is output from the drive control device 9 based on the calculated angle difference, and the can receiving stand 2 is rotated to the measurement start point 34 (422). Thereby, the measurement start point 34 of the can upper surface input image 31 can be detected.

FIG. 7 is a conceptual diagram for detecting a measurement start point from a measured can. While rotating the can holder 2 with the can 1 to be measured positioned at the center by a signal from the drive control device 9, the can upper surface input image 51, which is a characteristic part, is input, and the first can upper surface input image which is first input as an image 511 is the rotation reference point 53
(Angle: 0 degrees) is set. And can upper surface input image 5
1 to the extracted image 52 in the same area as the dictionary image 10
To get.

Each time the can receiver 2 is rotated, the image input processing of the first can upper surface input image 511, the second can upper surface input image 512, and the third can upper surface input image 513 is performed. Also, each time,
The angle from the rotation reference point 53 is obtained from the drive control device 9 and the drive device 5. Extracted image 5 of can upper surface input image 51
A normalized correlation process is performed between 21 to 523 and the dictionary image 10 registered in the memory of the can 1 to obtain respective correlation values.

Using the angle from the rotation reference point 53 of the extracted image (corresponding to the extracted image 513 in this case) whose correlation value exceeds the specified value, the measurement reference point from the rotation reference point 53 is used. 7, 90 ° is calculated in FIG. 7, and then, based on the result of calculating the angle difference between the measurement reference point 33 and the measurement start point 34 and 180 ° in FIG. Then, the can receiver 2 on which the measured can 1 is placed is rotated by the driving device 5 to detect the measurement start point 34 of the measured can 1.

FIG. 8 shows an overall configuration of an X-ray can-clamp measuring unit according to another embodiment of the present invention, and FIG.
FIG. 2 is a side view of a can body which is a characteristic part of the can to be measured and is a measurement start point detection part.

The measuring device for the X-ray can-clamping section includes a can cradle 2 on which the can 1 to be measured is placed in contact with the bottom surface of the can 1 to be measured and a can cradle 2 placed on the can cradle 2 And a can positioning device 2a for positioning the can to be measured 1
And a drive device 5 that can be driven to move forward, backward, left, right, up and down, and can be driven to rotate, and X-rays that irradiate X-rays in the tangential direction to the tightening portion of the can 1 to be measured positioned at the center of the can support 2. Photographing the irradiation device 3, the image receiving device 4 opposed to the X-ray irradiation device 3 for receiving the X-ray image of the tightening unit, and the can upper surface which is a characteristic part of the can 1 to be measured placed on the can receiving stand 2. Video camera A6
A video camera B7 for photographing the side surface of the can, which is a characteristic part of the can 1 to be measured placed on the can holder 2, and an image processing apparatus for processing the received X-ray image and the photographed can top surface image and can side surface image The driving control device 9 controls the driving device 5 based on data output from the image processing device 80.

The image processing device 80 includes an X-ray image processing unit 81
0, a video image processing unit 820, and a data processing unit 830
And a storage unit 840.

The X-ray image processing section 810 processes the X-ray image of the tightening section sent from the receiver 4.

The video image processing section 820 includes a dictionary table 8230 storing, as a dictionary image 100, an image obtained by extracting a feature from a can top image or a can side image which is a characteristic portion of the measured can 1 previously photographed. Video camera A6
Alternatively, it is composed of a comparison / determination unit 8220 for comparing and judging the dictionary image 100 with the extracted image 110 extracted from the can upper surface or the can side of the can 1 to be measured photographed by the video camera B7, for example, a normalized correlation process.

A plurality of dictionary images 10 are stored in the dictionary table 8230 for each type of object to be measured, for example, for each type due to the difference in the shape of the can lid due to the difference between the pull tab shape and the manufacturing company, and the difference in the pattern on the side of the can. .

The selection of the video camera for acquiring the characteristic portion of the can to be measured is, for example, in the case of a three-piece metal can, a side seam 6 which must be excluded from X-ray image processing.
Since 0 is included, a video camera B7 whose side seam position can be determined from the input image is used as an extraction of the characteristic portion of the can 1 to be measured. In the case of a two-piece metal can, since the side seam 60 does not exist, the video camera A6 is used. Alternatively, one of the video cameras B7 is selected and the one that can acquire an image in which the characteristics of the can 1 to be measured are more easily grasped is used.

The data processing section 830 includes a comparison / determination section 822
Based on the determination result of 0, the rotation amount of the measured can 1 is calculated,
Outputs the rotation amount data. Also, the data processing unit 830
Outputs a drive command to the drive control device 9 based on the X-ray image data processing result from the X-ray image processing unit 810.

The storage unit 840 stores a processing program, X-ray image data, processing results of the data processing unit 830, and the like.

The drive control unit 9 includes a drive control processing unit 91
, A drive control storage unit 92, a motor controller 94
And an X, Y, O axis driver 95 and a position detection unit 96.

The drive control processing section 91 includes a data processing section 83
The drive device 5 is controlled via the motor controller 94 and the X, Y, and O-axis drivers 95 based on the rotation amount data output from.

The position detector 96 acquires the position data of the can 1 to be measured on the can receiver 2 based on the signal from the driving device 5.

The drive control storage section 92 stores a processing program, position data acquired by the position detection section 96, and the like.

Next, the operation of the image processing device 80 and the drive control device 9 for detecting the measurement start point will be described.

For example, when the can 1 to be measured is a three-piece can, an image of the can side, which is a characteristic portion of the can 1 measured at the center of the can holder 2 by the positioning device 2a, is inputted from the video camera B7. The input image is the image processing device 80
The video image processing unit 820 of FIG. The video camera B7 is configured to be able to move back and forth and left and right with respect to the side surface of the can, which is a characteristic part of the can 1 to be measured.

The comparison / determination unit 8 in the video image processing unit 820
Reference numeral 220 denotes a plurality of dictionary images for each model to be measured corresponding to the can side, which is a characteristic portion of the can 1 to be measured, which is called from a dictionary table 8230 previously arranged on a memory in the video image processing unit 820; A normalized correlation process is performed between the extracted image extracted from the input image of the camera B7. If the maximum correlation value with the dictionary image for each type of measurement object used for comparison is not equal to or greater than the predetermined value (99%), the driving device 5 sets the can 1 to be measured based on the signal from the driving control device 9, that is, the rotation amount data. The mounted can receiver 2 is driven to rotate, and the rotation operation is repeated until the maximum correlation value becomes equal to or greater than a predetermined value.

FIG. 10 is a conceptual diagram of dictionary images 74 to 76 extracted as features from can side images of the measured can 1 previously registered as a dictionary in a dictionary table 8230 on a memory in the video image processing unit 820. .

When the characteristics of each type of can are captured, the difference in the pattern on the side of the can can be effectively utilized as in the case of the pull tab shape of the can lid.
The input image of the video camera B7 placed in parallel with the line irradiation device 3 is rotated by the driving device 5 receiving a signal from the driving control device 9 so that the side seam 60 of the can 1 to be measured becomes the center. Then, a can side image 61 which is a characteristic portion of the can 1 to be measured is obtained, an area 73 is taken around the can side image 61 so as to surround a pattern on the can side, and a reference image 74 of the can 1 to be measured is taken in the area 73. The reference image 74 is extracted and registered as a dictionary (data) in the dictionary table 8230 as a measurement reference point.

Since the X-ray tightening measurement of the side seam portion is usually excluded from the measurement target portion, the measurement start point 33 is inevitably determined when it is set at the center portion of the input image.

Next, the can cradle 2 on which the measured can 1 is placed
Is rotated by a specified angle (set to 6 °), a can side image 62 is obtained, and a dictionary image 7 which is a characteristic portion is obtained from the can side image 62.
5 is extracted, and this dictionary image 75 is used as the image of the measurement reference point + (the specified rotation angle × the number of movements) as the dictionary table 823.
Register to 0. The same procedure is repeated for the entire circumference, ie, 60 times, and registered as a dictionary of the can 1 to be measured.

FIG. 11 is a conceptual diagram of a dictionary image for each type of measurement object registered in the dictionary table 8230. As the dictionary image 100, for example, a type A and a type B are registered for each type.

FIGS. 12 to 14 show flowcharts of detecting the measurement start point using the image of the side surface of the can in the X-ray can-clamping section measuring device of FIG.

First, the video camera B7 performs an image input process (step 1211 in FIG. 13) and a feature extraction process (1212) of the can side, which is a characteristic portion of the can 1, to obtain an extracted image 110. Next, a normalized correlation process (1213 to 1215) is performed between the extracted image 110 and the measurement target kind dictionary 72 corresponding to the measured can 1, that is, between all the registered dictionaries in the measurement target kind dictionary 72. , The maximum correlation value calculation process (122).

Next, the calculated maximum correlation value is the specified value (99
% Or more (123), and determines whether it is the specified value (9
If not, the highest value among the acquired correlation values is set as a provisional reference value (124), and the can cradle 2 is moved from the provisional reference value by the drive device 5 using a signal from the drive control device 9. For example, it is rotated to a position of -6 ° (12
5). After the rotation, the video camera B7 again performs the image input process (1261) and the feature extraction process (1262) on the side surface of the can, which is the characteristic portion of the can 1 to be measured, to obtain the extracted image 110.

Next, a normalized correlation process (1263) is performed between the reference image 74 (of an angle of 0 °) of the target product type dictionary 72 corresponding to the measured can 1 and the extracted image 110, and the correlation value is obtained. A calculation process (1264) is performed. If the calculated correlation value does not reach the specified value (99% or more), the can receiving table 2 is rotated by the driving device 5 by a signal from the driving control device 9 (128), and the process returns to step 126 again to return the measured can 1 Performs image input and feature extraction processing on the side of the can.

At this time, since the can holder 2 does not need to be rotated all around, the rotation angle is set to 3 ° or less so that the measurement reference point can be more easily detected (128). For example, if the rotation angle is set to 1 °, 1
The can 1 to be measured is rotated twice, and the image input / feature extraction processing is repeated each time (129).

After the rotation by 12 ° is completed, the one having the maximum correlation value is calculated from the correlation values at each angle. In this case, the calculated extracted image 110 is the same as the reference image 74 of the measurement target type dictionary 72, so that the reference position is set and the measurement start point 33 is set. The obtained maximum correlation value is set as a reference value (130), and the drive control device 9 is set.
The can receiver 2 is rotated to the reference value by the driving device 5 according to the signal from (131).

The extracted image 110 when the correlation value has become the specified value (99% or more) becomes the same as the reference image 74 of the target type dictionary 72, so that the reference position is set and the measurement start point 33 is set. .

FIG. 15 is a conceptual diagram for detecting a measurement start point from an image of the side surface of the can to be measured. A can side input image 610, which is a characteristic part, is input while rotating the can receiving stand 2 in which the can to be measured 1 is positioned at the center by a signal from the drive control device 9, and the target type dictionary 7 is input from the can side input image 610.
An extracted image 110 is obtained in the same part of the area as the area 2.

Normalization is performed between the extracted image of the can side input image 610 and the target type dictionary 72 registered in the memory of the target can 1, that is, all the registered dictionaries in the target type dictionary 72. A correlation process is performed to obtain respective correlation values.

The angle at the time of dictionary registration from the reference dictionary 74 of the dictionary image (corresponding to the dictionary image 75 in this case) whose correlation value exceeds the specified value is used (in this case, 6 °). In response to a signal from the drive control device 9, the can-time receiving table 2 on which the measured can 1 is placed is rotated by the driving device 5 to set the measurement start point 33 of the measured can 1.

FIG. 16 is a conceptual diagram for detecting a measurement start point from an image on the upper surface of the can to be measured. While rotating the can holder 2 with the can 1 to be measured positioned at the center by the signal from the drive control device 9, the can upper surface input image 90, which is a characteristic part, is input. An extracted image 91 in the same region is obtained.

Normalization is performed between the extracted image of the can upper surface input image 90 and the target type dictionary 92 registered in the memory of the target can 1, that is, all the registered dictionaries in the target type dictionary 92. A correlation process is performed to obtain respective correlation values.

Using the angle at the time of dictionary registration from the reference dictionary 93 of the dictionary image (corresponding to the dictionary image 95 in this case) in which the correlation value exceeds the specified value (in this case, 90 °) In response to a signal from the drive control device 9, the can-time receiving table 2 on which the measured can 1 is placed is rotated by the driving device 5 to set the measurement start point 33 of the measured can 1.

In the case of the present embodiment, since the characteristic portions for the entire circumference are extracted in advance and registered as a plurality of dictionaries for each product type, the can receiver 2 on which the can 1 to be measured is placed Since there is no need to rotate by the circumference, the measurement start point can be detected at a very high speed.

[0076]

According to the present invention, exactly the same measurement start point can be detected for the same can type. Therefore, even when a plurality of cans are measured, there is no variation in the measurement start points, and the dimensions of the can-clamping portion can be reduced. It can be measured accurately.

In addition, it is not necessary to mark the measurement start point of the can to be measured, and the invention can be easily applied to an in-line apparatus.

[Brief description of the drawings]

FIG. 1 is an overall configuration diagram of an apparatus for measuring the dimensions of a can-clamping portion according to an embodiment of the present invention.

FIG. 2 is a plan view of a top surface of the can, which is a characteristic part of the can to be measured in FIG.

FIG. 3 is a conceptual diagram of a dictionary image extracted as a feature from a can upper surface input image, which is a feature of the measured can of FIG. 2;

FIG. 4 is a flowchart of a measurement start point detection process in the can winding part dimension measuring device of FIG. 1;

FIG. 5 is a flowchart of the video input / feature extraction processing of FIG. 4;

FIG. 6 is a flowchart of a measurement start point setting process of FIG. 4;

FIG. 7 is a conceptual diagram for detecting a measurement start point from the measured can of FIG. 1;

FIG. 8 is an overall configuration diagram of an apparatus for measuring the dimensions of a can-clamping portion according to another embodiment of the present invention.

FIG. 9 is a side view of a can, which is a characteristic portion of the can to be measured in FIG.

FIG. 10 is a conceptual diagram of a dictionary image extracted as a feature from a can side input image which is a feature portion of the measured can of FIG.

11 is a conceptual diagram of a dictionary table configuration of a dictionary image extracted as a feature from a can side input image which is a feature portion of the measured can of FIG. 8;

FIG. 12 is a flowchart of a measurement start point detecting process in the measuring apparatus for measuring the size of a can-clamping portion of FIG. 8;

FIG. 13 is a flowchart of a video input / feature extraction process with all the corresponding-type dictionaries of FIG. 12;

14 is a flowchart of a video input / feature extraction process with the corresponding type dictionary 0 in FIG. 12;

FIG. 15 is a conceptual diagram in the case of using a can side image for the measurement start point detection processing in the can winding part dimension measuring device of FIG. 8;

FIG. 16 is a conceptual diagram in the case of using a can upper surface image for a measurement start point detection process in the can winding part measuring device of FIG. 8;

[Explanation of symbols]

DESCRIPTION OF SYMBOLS 1 ... Can to be measured, 2 ... Can receiving stand, 2a ... Can positioning device, 3
... X-ray irradiation device, 4 ... Receiver, 5 ... Driver, 6 ... Video camera A, 7 ... Video camera B, 8,80 ... Image processor, 81,810 ... X-ray image processor, 82,820 ... Video image processing unit 821 dictionary table 822 82
20: comparison / determination unit, 8230: dictionary table, 83, 8
Reference numeral 30: Data processing unit, 84,840: Storage unit, 9: Drive processing unit, 91: Drive control processing unit, 92: Drive control storage unit, 94: Motor controller, 95: X, Y, O axis driver, 96 ... Position detector, 10,75,76,94,9
5,100: Dictionary image, 11, 52, 91, 110: Extracted image, 30: Pull tab, 31, 51, 90: Can upper surface input image, 32: X-ray irradiation location, 33: Measurement reference point, 34: Start measurement Point, 53: rotation reference point, 60: side seam, 6
1, 62, 63: can side image, 610: can side input image,
72, 92: Dictionary of target varieties, 74, 93: Reference image

Continued on the front page (72) Inventor Atsushi Sodeyama 3-2-1 Sachimachi, Hitachi-shi, Ibaraki Inside Hitachi Engineering Co., Ltd. (72) Inventor Yoshimi Otake 3-2-1 Sachimachi, Hitachi-shi, Hitachi, Hitachi Inside Engineering Co., Ltd. (72) Inventor Yuji Takamoto 1-1, Namiki Motomachi, Kawaguchi City, Saitama Prefecture Inside (72) Inventor Hiroshi Nakayama 1-1, Namiki Motomachi, Kawaguchi City, Saitama Prefecture Sapporo Bill Co., Ltd. Inside

Claims (5)

    [Claims]
  1. An X-ray irradiator for irradiating X-rays, and an image receiver comprising an image receiver for receiving the irradiated X-rays and an image processor, wherein the X-ray irradiator and the image receiver are connected to each other. An X-ray can-clamping-portion measuring device having a drive unit that rotatably positions a tightening portion of a can to be measured at a measurement position by the X-ray irradiation, wherein the feature of the can to be measured disposed at the measurement position is A video camera is arranged to face the portion, and the image processing device is configured to process an X-ray image of the tightening unit obtained by the irradiation.
    A line image processing section and a video image processing section for processing a characteristic partial image of the measured can taken by the video camera, wherein the video image processing section extracts an image extracted from the captured characteristic partial image And a comparison / determination unit for comparing and comparing a dictionary image created by extracting a feature from the feature portion of the measured can which has been photographed in advance, and rotating the measured can based on the determination result of the comparison / determination unit. An X-ray can-winding-portion measuring device, comprising: a drive control device for controlling the drive device to be driven.
  2. 2. An apparatus according to claim 1, wherein said video image processing section is provided with a dictionary table for determining said dictionary image and said measurement start point.
  3. 3. A characteristic portion of the can to be measured is mounted on a can holder rotatable by a driving device, and a video camera disposed opposite to the characteristic portion of the can to be measured. Is input to the image processing apparatus connected to the output side of the video camera, and an image created by extracting features from the input image is used as a dictionary image. An image and data in which a measurement start point is set from the input image are registered in advance in the dictionary table of the image processing device as a dictionary of the can to be measured, and the can cradle on which the can to be measured is driven is driven. A normalization correlation process is performed by comparing the dictionary image with an extracted image obtained by extracting a feature from an image of a can feature portion input by the video camera while being rotated by the driving device that has received a signal from a control device. Based on the result of calculating the function value, the measurement start point detecting method and detecting the measurement starting point from the circumference of the object to be measured can.
  4. 4. A characteristic portion of the measured can is placed on a rotatable can holder by a driving device, and a video camera arranged opposite the characteristic portion of the placed can. A video input, captures the video input input image to an image processing device connected to the output side of the video camera, obtains an image created by extracting features from the captured input image as a reference image, And while rotating the can cradle by a specified angle by the driving device that has received a signal from a drive control device based on the position of the can cradle when the reference image is created, video input is performed by the video camera. An image created by extracting features from each image of the can feature portion obtained is obtained as a plurality of dictionary images, and the obtained reference image and the plurality of dictionary images, and the reference image and the plurality of feature images are respectively obtained. The rotation angle data from the reference image at the time of acquisition is registered in advance in the dictionary table of the image processing apparatus as a measurement target type dictionary for each type of the can to be measured, and is placed on the can receiver. The video image of the measured can is input by the video camera, the extracted image obtained by extracting the characteristic from the image of the characteristic portion of the measured can input image, the reference image and the plurality of dictionary images for each of the target type dictionary, and And performs a normalized correlation process based on the rotation angle data of the reference image or the dictionary image having the maximum correlation value among the reference image and the plurality of dictionary images for each of the measurement target type dictionaries. A measurement start point detection method, comprising: detecting a measurement start point from a circumference of a can to be measured.
  5. 5. The measurement according to claim 3, wherein a signal from the drive control device is output and the can cradle is rotated to the measurement start point registered in the dictionary table. Start point detection method.
JP11224079A 1998-08-11 1999-08-06 X-ray measuring apparatus for can seam part and detecting method for measurement starting point Pending JP2000121580A (en)

Priority Applications (3)

Application Number Priority Date Filing Date Title
JP10-226879 1998-08-11
JP22687998 1998-08-11
JP11224079A JP2000121580A (en) 1998-08-11 1999-08-06 X-ray measuring apparatus for can seam part and detecting method for measurement starting point

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
JP11224079A JP2000121580A (en) 1998-08-11 1999-08-06 X-ray measuring apparatus for can seam part and detecting method for measurement starting point

Publications (1)

Publication Number Publication Date
JP2000121580A true JP2000121580A (en) 2000-04-28

Family

ID=26525836

Family Applications (1)

Application Number Title Priority Date Filing Date
JP11224079A Pending JP2000121580A (en) 1998-08-11 1999-08-06 X-ray measuring apparatus for can seam part and detecting method for measurement starting point

Country Status (1)

Country Link
JP (1) JP2000121580A (en)

Cited By (11)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP2006023123A (en) * 2004-07-06 2006-01-26 Hokkai Can Co Ltd Defect inspection method and its device
CN1323736C (en) * 2004-03-10 2007-07-04 复盛股份有限公司 Golf rod head having joint portion and material filling space
JP2008215955A (en) * 2007-03-01 2008-09-18 Asahi Breweries Ltd Inverted can detector
JP2012233919A (en) * 2004-05-26 2012-11-29 Werth Messtechnik Gmbh Coordinate measuring apparatus and method for measuring an object
WO2013103027A1 (en) * 2012-01-05 2013-07-11 富士ゼロックス株式会社 Image processor, image processing program, and object matching device
WO2014164796A3 (en) * 2013-03-11 2014-11-27 Rexam Beverage Can Company Method and apparatus for necking and flanging a metallic bottle
CN105817784A (en) * 2016-05-27 2016-08-03 苏州华源包装股份有限公司 Positioning structure for searching for welding seam through photographing detection
CN106024341A (en) * 2016-05-11 2016-10-12 国网福建省电力有限公司 X-ray alarm device for loose displacement of internal components of main transformer and operation method of X-ray alarm device
EP3208009A1 (en) * 2016-02-16 2017-08-23 Zhaoqing Canneed Instrument Limited Machine for automatically detecting the quality of the seam of a can body
US10195842B2 (en) 2013-06-11 2019-02-05 Ball Corporation Apparatus for forming high definition lithographic images on containers
US10549921B2 (en) 2016-05-19 2020-02-04 Rexam Beverage Can Company Beverage container body decorator inspection apparatus

Cited By (16)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN1323736C (en) * 2004-03-10 2007-07-04 复盛股份有限公司 Golf rod head having joint portion and material filling space
JP2012233919A (en) * 2004-05-26 2012-11-29 Werth Messtechnik Gmbh Coordinate measuring apparatus and method for measuring an object
JP2006023123A (en) * 2004-07-06 2006-01-26 Hokkai Can Co Ltd Defect inspection method and its device
JP4516788B2 (en) * 2004-07-06 2010-08-04 北海製罐株式会社 Defect inspection method and apparatus
JP2008215955A (en) * 2007-03-01 2008-09-18 Asahi Breweries Ltd Inverted can detector
US9858499B2 (en) 2012-01-05 2018-01-02 Fuji Xerox Co., Ltd. Image processor, non-transitory computer readable medium and object matching device
WO2013103027A1 (en) * 2012-01-05 2013-07-11 富士ゼロックス株式会社 Image processor, image processing program, and object matching device
JP2013140503A (en) * 2012-01-05 2013-07-18 Fuji Xerox Co Ltd Image processing device and image processing program
CN103988225A (en) * 2012-01-05 2014-08-13 富士施乐株式会社 IMAGE PROCESSOR, image processing program AND OBJECT MATCHING DEVICE
CN103988225B (en) * 2012-01-05 2018-01-30 富士施乐株式会社 Image processor, image processing method and object matching device
WO2014164796A3 (en) * 2013-03-11 2014-11-27 Rexam Beverage Can Company Method and apparatus for necking and flanging a metallic bottle
US10195842B2 (en) 2013-06-11 2019-02-05 Ball Corporation Apparatus for forming high definition lithographic images on containers
EP3208009A1 (en) * 2016-02-16 2017-08-23 Zhaoqing Canneed Instrument Limited Machine for automatically detecting the quality of the seam of a can body
CN106024341A (en) * 2016-05-11 2016-10-12 国网福建省电力有限公司 X-ray alarm device for loose displacement of internal components of main transformer and operation method of X-ray alarm device
US10549921B2 (en) 2016-05-19 2020-02-04 Rexam Beverage Can Company Beverage container body decorator inspection apparatus
CN105817784A (en) * 2016-05-27 2016-08-03 苏州华源包装股份有限公司 Positioning structure for searching for welding seam through photographing detection

Similar Documents

Publication Publication Date Title
US8977056B2 (en) Face detection using division-generated Haar-like features for illumination invariance
US8331619B2 (en) Image processing apparatus and image processing method
US8773514B2 (en) Accurate 3D object reconstruction using a handheld device with a projected light pattern
US8300139B2 (en) Image pickup apparatus and image pickup method
US4490617A (en) Optical width measuring system using two cameras
KR100617390B1 (en) Face identification apparatus, face identification method, and medium for recording face identification program
US4849914A (en) Method and apparatus for registering color separation film
JP4320272B2 (en) Specific area detection method, specific area detection apparatus, and program
KR101267205B1 (en) Face feature point detection device and program
US5682242A (en) Method and apparatus for determining a location on a surface of an object
US8102465B2 (en) Photographing apparatus and photographing method for photographing an image by controlling light irradiation on a subject
US8385645B2 (en) Object detecting device, imaging apparatus, object detecting method, and program
US7660442B2 (en) Method and system for capturing fingerprints, palm prints and hand geometry
JP5004379B2 (en) Movable X-ray apparatus and projection position determination method
US8045763B2 (en) Method and apparatus for recognizing and determining the position of at least one teat cup
US7769233B2 (en) Particular-region detection method and apparatus
US7133572B2 (en) Fast two dimensional object localization based on oriented edges
JP4130641B2 (en) Digital still camera and control method thereof
CN100502471C (en) Image processing device, image processing method and imaging device
US6289113B1 (en) Handheld iris imaging apparatus and method
US6822690B2 (en) Image processing apparatus, image pickup apparatus, and image processing method capable of eliminating effect of outside light
US7620242B2 (en) Particular-region detection method and apparatus, and program therefor
US6968094B1 (en) Method of estimating and correcting camera rotation with vanishing point location
DE102012021375A1 (en) Apparatus and method for detecting a three-dimensional position and orientation of an article
US20190052801A1 (en) Methods and apparatus to capture photographs using mobile devices

Legal Events

Date Code Title Description
A131 Notification of reasons for refusal

Free format text: JAPANESE INTERMEDIATE CODE: A131

Effective date: 20050419

A02 Decision of refusal

Free format text: JAPANESE INTERMEDIATE CODE: A02

Effective date: 20050817