JP7277118B2 - Quality judgment device - Google Patents

Description

The present invention relates to a quality determination device. More particularly, the present invention relates to a quality judgment device for judging the quality of labels affixed to articles.

A label on which information necessary for management in the factory is printed is issued and affixed to an article produced in the factory. The label is usually attached to a predetermined position for each item of goods. In addition, it is common for the label to have predetermined characters and figures printed at a predetermined print density. However, the actual situation is that some of the labels that are issued in large quantities have poor quality in terms of print density. In addition, the actual situation is that some of the labels attached to each of a large number of items are of poor quality with respect to the attachment position.

JP 2008-171582 A

As a technique for detecting defects related to labels, the above-mentioned Patent Document 1 has been proposed. Patent Literature 1 discloses that one end of an exterior label wound around the body of a battery pack is provided with a misalignment detection mark that, when properly wound, is hidden behind the other end of the label and becomes invisible. ing. According to the technique disclosed in Patent Literature 1, it is possible to visually check the displacement when the label is wound.

However, the sticking position where the label is stuck differs depending on the item of the article. Therefore, labels are not always applied in a rolled fashion. When a label is affixed to a flat portion of an article, it is difficult to detect misalignment of the affixing position using the technique disclosed in Japanese Patent Application Laid-Open No. 2002-200012.

In addition, it is preferable to check the quality of the label not only in terms of the sticking position, but also in terms of printing such as ink density, character blurring, and staining. However, with the technique disclosed in Patent Literature 1, it is difficult to detect quality defects other than misalignment of labels attached by the winding method.

SUMMARY OF THE INVENTION Accordingly, it is an object of the present invention to provide a quality determination device capable of easily determining the quality of a label based on image information obtained from a vision sensor.

The present invention is a quality determination device for determining the quality of labels affixed to each of one or more items, comprising: a robot having an articulated arm; a vision sensor that acquires image information about the label by photographing the item; a storage unit that stores label information about the label corresponding to each of one or more items; the label information and the image. a quality determination unit that determines the quality of the label attached to the item based on the information.

According to the present invention, label quality can be easily determined based on image information obtained from a vision sensor.

1 is a perspective view of a quality determination device according to an embodiment; FIG. FIG. 2 is a top view showing a label photographed by the vision sensor of the embodiment and a part of the article to which the label is attached; 1 is a block diagram simply showing the configuration of a quality determination device according to an embodiment; FIG. It is the flowchart which showed simply the flow of the quality determination process performed by the quality determination apparatus of embodiment.

BEST MODE FOR CARRYING OUT THE INVENTION A quality determination device of the present invention will be described in detail below with reference to preferred embodiments and with reference to the accompanying drawings.

[Embodiment]
FIG. 1 is a perspective view of a quality determination device 10 of this embodiment.

As shown in FIG. 1, a quality determination device 10 is attached to a robot 14 having an articulated arm 12 and to the articulated arm 12 of the robot 14, and acquires image information 20 about a label 18 by photographing an item 16. and a vision sensor 22 to. Each joint of the articulated arm 12 is provided with a motor 24 . The articulated arm 12 is driven as each of the motors 24 rotates.

The quality determination device 10 also has a display unit 26 (see FIG. 3) as an interface for issuing a notification indicating that the quality of the label 18 is defective. The display unit 26 is, for example, a liquid crystal panel that can display notifications. The article 16 to be imaged by the vision sensor 22 is, for example, a part of a machine, and is placed on a pedestal 28 in this embodiment. The vision sensor 22 can move around the article 16 placed on the pedestal 28 as the articulated arm 12 is driven.

Pedestal 28 may be a fixed table or a belt conveyor. The pedestal 28 of this embodiment is a fixed table. Further, when the article 16 of the present embodiment is photographed by the vision sensor 22, it is placed at a predetermined position on the fixed pedestal 28 with a predetermined orientation. Also, as shown in FIG. 1, a label 18 is attached to the surface of the item 16 . In this embodiment, it is assumed that there are two or more items 16, and the position where the label 18 is attached on the surface of the item 16 is determined in advance for each item of the item 16. FIG.

FIG. 2 is a top view showing the label 18 imaged by the vision sensor 22 and a portion of the item 16 to which the label 18 is attached.

Predetermined information determined corresponding to the item of the article 16 is printed on the label 18 . For example, as shown in FIG. 2, the name A, storage location B, bar code C, and stock information D are printed on the label 18 of the present embodiment as information related to the item 16 . Information printed on the label 18 is not limited to that shown in FIG. The items and number of items of information printed on label 18 may be predetermined or varied by those skilled in the art.

FIG. 3 is a block diagram simply showing the configuration of the quality determination device 10 of this embodiment.

The quality determination device 10 has a control section 30 . This control unit 30 includes a calculation unit 32 and a storage unit 34 . The computing unit 32 is, for example, a processor. The calculation unit 32 has a quality determination unit 36 that determines the quality of the label 18 attached to the item 16 based on the label information 40 and the image information 20 . Further, when the quality determination unit 36 determines that the quality of the label 18 attached to the product 16 is defective, the calculation unit 32 generates a notification indicating the determination result and causes the display unit 26 to display the notification. It has a generator 38 . The storage unit 34 is, for example, a memory. Storage unit 34 stores label information 40 that is information about labels 18 corresponding to each of one or more items 16 . The storage unit 34 may also store a program or the like that can be executed by the calculation unit 32 .

Label information 40 includes, for example, basic image information 42 . The basic image information 42 refers to information obtained by previously photographing the well-printed label 18 from a predetermined photographing position with the vision sensor 22 . A predetermined photographing position for obtaining the basic image information 42 may be determined within a range in which the vision sensor 22 can be moved by the articulated arm 12 for each item to which the label 18 corresponds. In order to store the basic image information 42 captured by the vision sensor 22 in the storage unit 34 , for example, the basic image information 42 may be transmitted from the vision sensor 22 to the control unit 30 . Then, the control unit 30 may receive the basic image information 42 sent from the vision sensor 22 and store it in the storage unit 34 . The form of communication between the vision sensor 22 and the control unit 30 at this time may be wired or wireless.

Label information 40 may also include position information 44 that indicates a predetermined affixing position where label 18 should be positioned for item 16 . The label information 40 of this embodiment includes at least basic image information 42 and position information 44 .

In this embodiment, it is assumed that the control unit 30 of the quality determination device 10 also controls the driving of the robot 14 . In addition to the quality determination unit 36 described above, the calculation unit 32 of the present embodiment further includes a signal generation unit 46 that generates and transmits a signal for controlling the driving of the motor 24 provided in the articulated arm 12 . The robot 14 also has a signal processor 48 that receives and analyzes the signal transmitted from the signal generator 46 and drives the motor 24 .

FIG. 4 is a flow chart simply showing the flow of quality determination processing executed by the quality determination device 10 of the present embodiment.

Hereinafter, the flow of quality determination processing executed by the quality determination device 10 will be described with reference to FIG. 4 as well. As a premise, it is assumed that the work of issuing and attaching the label 18 corresponding to the item of the article 16 has been completed in advance. It is also assumed that the number of items 16 to be determined and the order of determination are determined in advance. The quality determination device 10 executes the quality determination process, for example, by having the calculation unit 32 read out and execute a predetermined program stored in the storage unit 34 .

In step S<b>1 , the vision sensor 22 is moved by the articulated arm 12 . Specifically, the signal generator 46 of the controller 30 sends a signal to the signal processor 48 of the robot 14 to rotate the motor 24 and drive the articulated arm 12 .

In the present embodiment, the articulated arm 12 is driven so that the vision sensor 22 is positioned at a photographing position predetermined for each item. In this embodiment, the photographing position predetermined for each item is the same photographing position as when the basic image information 42 corresponding to the article 16 was obtained.

In step S<b>2 , the quality determination device 10 acquires image information 20 of the item 16 . As already explained, the image information 20 is obtained by photographing the item 16 on the pedestal 28 with the vision sensor 22 . The acquired image information 20 is transmitted to the control section 30 . The form of communication between the vision sensor 22 and the control unit 30 when transmitting the image information 20 to the control unit 30 may be the same as when transmitting the basic image information 42 from the vision sensor 22 to the control unit 30 .

Next, in step S3, the quality determination device 10 determines whether or not the label 18 attached to the article 16 has a quality defect related to the attachment position. This determination is performed by the quality determining section 36 determining whether or not the label 18 is positioned at a predetermined pasting position by image recognition based on the position information 44 and the image information 20 . That is, the image information 20 is obtained in step S2 to such an extent that at least it is possible to determine where the label 18 is attached to the article 16 . The quality determination unit 36 calculates and finds the position where the label 18 is attached by image recognition based on the acquired image information 20 . Then, the quality determination unit 36 compares the determined sticking position with the pre-stored position information 44 to determine whether the label 18 is correctly stuck at a predetermined position. Here, when it is determined that "there is a quality defect related to the pasting position", the quality determination device 10 generates a notification (message) to the effect that "there is a quality defect related to the pasting position" in the notification generation unit 38. At the same time, the notification is displayed on the display unit 26 (step S7).

If no defect related to the pasting position is found in step S3, the quality determination unit 36 executes step S4 and subsequent steps. Steps S4 to S6 are mainly steps for quality judgment regarding printing. The quality determination unit 36 determines whether the label 18 attached to the item 16 is well printed based on the comparison between the basic image information 42 and the image information 20. Determine the quality of the label 18 .

In step S<b>4 , the quality determination device 10 determines whether the label 18 attached to the item 16 has quality defects related to defects. For this determination, for example, the quality determination unit 36 calculates a matching score by pattern matching the image information 20 and the basic image information 42 on a pixel-by-pixel basis, and determines whether or not the matching score exceeds the threshold value T1. This is done by making a decision based on

The matching score is a numerical representation of how similar the image patterns included in the image information 20 and the basic image information 42 are. A specific method of calculating the matching score and the threshold value T1 used as the criterion may be appropriately determined by those skilled in the art.

If it is determined that the image information 20 and the basic image information 42 have, for example, 90% or more of the same pattern based on the matching score, the quality determination unit 36 determines that the label 18 to be determined has no defects. do.

If it is determined that there is a defect in the label 18 to be determined, the quality determination device 10 shifts the process to step S7 and causes the display unit 26 to display a notification that "there is a defect related to the defect". .

In this embodiment, the image information 20 is acquired by photographing from the same photographing position as the corresponding basic image information 42 . Therefore, in the present embodiment, pattern matching may be performed on a pixel-by-pixel basis not only for the label 18 but also for the periphery where the label 18 is affixed, thereby determining the affixing position in step S4.

Next, in step S<b>5 , the quality determination device 10 determines whether or not the label 18 attached to the article 16 has quality defects related to print density. This determination is made by the quality determination unit 36 calculating a histogram of the image information 20 and comparing the histogram with a basic histogram obtained from the basic image information 42 . Here, for example, the quality determination unit 36 determines whether the histogram calculated from the image information 20 has a distribution matching the basic histogram within an allowable error range. The basic histogram is preferably calculated in advance, but may be calculated each time step S5 is executed.

If it is determined that there is an unacceptable error between the histogram of the label 18 to be determined and the basic histogram, the quality determination device 10 shifts the process to step S7 and states that "there is a print density defect". A notification to that effect is displayed on the display unit 26 .

Next, in step S<b>6 , the quality determination device 10 determines whether the label 18 attached to the item 16 has quality defects related to contamination. For this determination, for example, the quality determination unit 36 extracts the difference between the image information 20 and the basic image information 42, calculates the area of the difference, and determines whether the size of the area exceeds the threshold value T2. This is done by making a decision based on

If it is determined that the label 18 to be determined has dirt of a size equal to or greater than the threshold value T2, the quality determination device 10 shifts the process to step S7 and notifies that "there is a defect related to dirt". is displayed on the display unit 26 .

The above is the flow of the quality determination processing executed by the quality determination device 10 of the present embodiment. According to the quality determination device 10 of the present embodiment, it is possible to easily determine the quality of the label 18 based on the image information 20 obtained from the vision sensor 22 .

[Modification]
Although the embodiment has been described above as an example of the present invention, it is of course possible to add various modifications and improvements to the above embodiment. It is clear from the description of the scope of claims that forms with such modifications or improvements can also be included in the technical scope of the present invention.

(Modification 1)
Any of steps S3 to S6 described in the embodiment may be omitted, and the order of execution may be changed from the order shown in FIG. For example, if a person skilled in the art determines that it is sufficient to check the quality of only the label 18 sticking position, steps S4 to S6 may be omitted as necessary.

(Modification 2)
In step S1, it is not necessary to move the vision sensor 22 to the photographing position predetermined for each item. In step S1, it is sufficient if the vision sensor 22 can be moved to include at least the label 18 and its surroundings to the extent that it is possible to determine where on the item 16 the label 18 is attached.

Note that if step S2 is performed without determining the predetermined photographing position for each item, the image information 20 and the basic image information 42 are likely to differ in photographing range. In such a case, after the labels 18 shown in the image information 20 and the basic image information 42 are extracted by image recognition, pattern matching may be performed between the extracted images. At this time, pattern matching may be performed while interpolating the scales of the extracted images. As a result, even if there is a difference in the imaging range between the image information 20 and the basic image information 42, it is possible to perform the defect determination with good accuracy.

(Modification 3)
In the embodiment, the display unit 26 is provided as an interface for notifying the operator that a quality defect has been found in the label 18 . The notification may be given to the operator by voice. Therefore, the interface is not limited to the display unit 26, and may be a speaker, for example.

(Modification 4)
In the embodiment, the article 16 photographed by the vision sensor 22 is placed at a predetermined position on the fixed pedestal 28 in a predetermined orientation. For example, the control unit 30 may sequentially calculate the relative positional relationship between the vision sensor 22 and the item 16 . A predetermined positional relationship between the vision sensor 22 and the article 16 may be stored in advance in the storage unit 34 . Furthermore, the robot 14 may be controlled such that the relative positional relationship between the vision sensor 22 and the article 16 matches the predetermined positional relationship stored in the storage section 34 .

As a result, even if the product 16 is not placed on the pedestal 28 at a predetermined position or is not placed on the pedestal 28 in a predetermined direction, the quality determination device 10 can achieve the desired accuracy. , the image information 20 can be obtained.

(Modification 5)
The robot 14 may comprise an articulated arm 12' having a mechanism for grasping and transporting the item 16 with the label 18 to a predetermined location when the quality of the label 18 is found to be defective. The predetermined place is, for example, a so-called "defective product storage area", which is a place for separating defective products from non-defective products. The articulated arm 12' may be an arm different from the articulated arm 12 to which the vision sensor 22 is attached, or may be the same arm. Alternatively, the articulated arm 12 ′ may be provided in a robot 14 ′ different from the robot 14 provided with the articulated arm 12 .

(Modification 6)
The above-described embodiments and modifications may be combined as appropriate within a consistent range.

[Invention obtained from the embodiment]
Inventions that can be understood from the above embodiments and modifications will be described below.

A quality determination device (10) for determining the quality of a label (18) affixed to each of one or more items (16) includes a robot (14) having an articulated arm (12), and the robot (14). a vision sensor (22) attached to the articulated arm (12) of (14) and acquiring image information (20) about the label (18) by photographing the item (16); Based on a storage unit (34) storing label information (40) relating to the label (18) corresponding to each of the above items (16), and the label information (40) and the image information (20) a quality determination unit (36) for determining the quality of the label (18) attached to the item (16);

This allows the quality of the label (18) to be determined based on the image information (20) obtained from the vision sensor (22).

The label information (40) includes basic image information (42), and the quality determination unit (36) determines whether the label (18) attached to the article (16) is printed satisfactorily. The quality of the label (18) affixed to the item (16) may be determined by determining based on a comparison of the basic image information (42) and the image information (20) whether the good. This makes it possible to judge the quality of the label (18), for example, regarding defects, print density, or adherence of dirt.

The quality determination unit (36) performs pattern matching on the image information (20) and the basic image information (42) on a pixel-by-pixel basis, calculates a matching score based on the result of the pattern matching, and calculates the matching score. may determine whether the label (18) affixed to the item (16) is of poor quality with respect to affixed position or defects. This makes it possible to easily detect quality defects related to the label (18) sticking position or missing.

The quality judging section (36) calculates a histogram of the image information (20) and compares the histogram with a basic histogram obtained from the basic image information (42) to detect quality defects related to print density. A determination may be made as to whether the label (18) affixed to the item (16) is present. This makes it possible to easily detect quality defects related to print density of the label (18).

The quality determination unit (36) extracts the difference between the image information (20) and the basic image information (42), calculates the area of the difference, and determines whether the size of the area exceeds a threshold. It may be determined whether the label (18) affixed to the item (16) has a quality defect related to label contamination. This makes it possible to easily detect quality defects related to contamination of the label (18).

The label information (40) includes position information (44) indicating a predetermined sticking position where the label (18) corresponding to the item (16) should be positioned, and the quality determination unit (36) determines whether or not the label (18) is positioned at the predetermined pasting position by image recognition based on the position information (44) and the image information (20). defects in the label (18) affixed to the item (16). This makes it possible to easily detect quality defects related to the label (18) sticking position.

The robot (14) drives the articulated arm (12) to move the vision sensor (22) to a predetermined photographing position corresponding to each item of the article (16), and The vision sensor (22) may acquire the image information (20) by photographing the article (16) from a predetermined photographing position corresponding to each item of the article (16). As a result, for example, the accuracy of quality judgment performed by comparing the image information (20) and the basic image information (42) is improved.

an interface (26) for issuing a notification indicating that the quality of the label (18) is defective if the quality of the label (18) applied to the item (16) is determined to be defective in any way; You may have more. As a result, for example, the operator can easily know that the quality judging section (36) has detected a quality defect in the label (18).

DESCRIPTION OF SYMBOLS 10... Quality determination apparatus 12, 12'... Articulated arms 14, 14'... Robot 16... Goods 18... Label 20... Image information 22... Vision sensor 24... Motor 26... Display part 28... Pedestal 30... Control part 32... Calculation Section 34 Storage section 36 Quality determination section 38 Notification generation section 40 Label information 42 Basic image information 44 Position information 46 Signal generation section 48 Signal processing section A Name B Storage location C Barcode D ... inventory information T1 ... (pattern matching) threshold T2 ... (area difference) threshold

Claims (5)

A quality determination device for determining the quality of labels affixed to each of one or more items,
a robot having an articulated arm;
a vision sensor attached to the multi-joint arm of the robot for acquiring image information about the label by photographing the item;
a storage unit storing label information about the label corresponding to each of the one or more items;
a quality determination unit that determines the quality of the label attached to the item based on the label information and the image information;
with
The label information includes basic image information,
The quality determination unit determines whether or not the label attached to the item has a quality defect related to an attachment position or defect based on the result of pattern matching using the image information and the basic image information. death,
The quality determination unit calculates a histogram of the image information and compares the histogram with a basic histogram obtained from the basic image information to determine whether the label attached to the product has a quality defect related to print density. further determine whether there is
When the imaging range of the image information and the imaging range of the basic image information are different, the quality determination unit determines a first label image which is an image of the label extracted from the image information by image recognition, an image A quality determination device that performs the pattern matching using a second label image that is an image of the label extracted from the basic image information by recognition. A quality determination device for determining the quality of labels affixed to each of one or more items,
a robot having an articulated arm;
a vision sensor attached to the multi-joint arm of the robot for acquiring image information about the label by photographing the item;
a storage unit storing label information about the label corresponding to each of the one or more items;
a quality determination unit that determines the quality of the label attached to the item based on the label information and the image information;
with
The label information includes basic image information,
The quality determination unit determines whether or not the label attached to the item has a quality defect related to an attachment position or defect based on the result of pattern matching using the image information and the basic image information. death,
The quality determination unit extracts the difference between the image information and the basic image information, calculates the area of the difference, and determines the quality of label contamination based on whether the size of the area exceeds a threshold. further determining whether the label affixed to the item is defective in
When the imaging range of the image information and the imaging range of the basic image information are different, the quality determination unit determines a first label image which is an image of the label extracted from the image information by image recognition, an image A quality determination device that performs the pattern matching using a second label image that is an image of the label extracted from the basic image information by recognition. The quality determination device according to claim 1 or 2,
The quality determination unit performs the pattern matching on a pixel-by-pixel basis and calculates a matching score based on the result of the pattern matching. A quality determination device that determines whether or not the label that has been printed. The quality determination device according to any one of claims 1 to 3,
The label information includes position information indicating a predetermined pasting position where the label corresponding to the item should be positioned,
The quality determination unit determines whether or not the label is positioned at the predetermined attachment position by image recognition based on the position information and the image information. A quality determination device that determines whether or not the label attached to the The quality determination device according to any one of claims 1 to 4 ,
The quality determination device, further comprising an interface that issues a notification indicating that the quality of the label is defective when the quality of the label attached to the item is determined to be defective.

Images