WO2008104273A1

WO2008104273A1 - Method and inspection device for examining containers

Info

Publication number: WO2008104273A1
Application number: PCT/EP2008/001032
Authority: WO
Inventors: Herbert Menke; Martin Weisgerber; Tammo Zorn; Guido Sommer
Original assignee: Khs Ag
Priority date: 2007-03-01
Filing date: 2008-02-12
Publication date: 2008-09-04
Also published as: DE102007010018A1

Abstract

The object of the invention is a method and an inspection device for examining containers, particularly of bottles (1). At least one container image is generated by means of at least one container imaging device (5) and analyzed in a connected image processing unit (6). According to the invention the container image is evaluated for surveying the related container topology.

Description

Method and inspection device for inspecting containers

The invention relates to a method for the examination of containers of all kinds, such as bottles or cans for receiving goods of all kinds, especially bottles or cans for holding drinks, which at least one container image is generated by means of at least one container receiving device and analyzed in a connected image processing unit ,

Such methods and associated inspection devices for containers, in particular bottles, are known from practice and are otherwise described in the two documents DE 10 2005 017 957 A1 and DE 103 22 459 B3. As a container receiving device usually comes at this point a camera used. In the process, an associated image processing software ultimately makes sure that the inspection device downstream is subject to downstream sorting and / or discharge devices based on predetermined criteria. As a result, bottles with different contours and / or those with labels of label-free bottles can be separated (see DE 103 22 459 B3, section [0031]).

Thus, the prior art uses the container image or image processing only to make a SoI I / I st- comparison and to classify the associated containers or bottles in ultimately acceptable and unacceptable units. The unacceptable units or containers or bottles are logically sorted out. This has proven itself in principle. However, the described procedure for error analysis is unsuitable. That is, the discharged bottles or containers do not give a detailed conclusion about why they were sorted out and whether and where an upstream error occurs in the bottle treatment. Likewise, it is not known from the prior art to use the results of the inspection process to influence the preceding treatment steps. Here, the invention aims to provide a total remedy.

The invention is based on the technical problem of further developing such a method so that errors occurring in the container, or errors which arise as a result of a preceding treatment step, are taken into account in bottle treatment. In addition, an associated inspection device should be specified. To solve this technical problem, a generic method is characterized in that the container image is evaluated for the measurement including, where appropriate, color / symbol analysis of the associated container topology.

According to the invention, therefore, the container image is not used primarily for a target / actual comparison for bottle ejection. Rather, the container image serves to measure the container topology as completely as possible and not only qualitatively in the sense of "good / bad", but in particular to capture quantitatively. The container topology in the context of the invention means all optical and geometric properties of the container to be examined, in particular its length, size and diameter. But the seat, the location and attachment of additional aggregates such as closures, labels, etc. falls under this. Likewise belong to the generic term of the container topology, the level of a filled liquid, the color of the bottle, the color of the filled liquid or the color of labels, any symbols, etc. Finally, in the context of surveying the container topology and the lateral surface of the container and / or Contour or shape of the container with regard to any irregularities measured and examined.

Such irregularities can be attributed, for example, to damage, imperfections in the container material, cracks, dents, etc., which are measured quantitatively in the context of the invention, like the container properties already described. This means that, for example, a crack in the container wall, the position of the label or a closure are geometrically determined exactly, for example, by assigning the crack a certain angular position of the usually rotationally symmetrical container and a height above a base surface. The position and orientation of the label is documented as part of the survey of the container topology, for example, also by the swept angle range between transverse edges of the label and the height profile of the associated longitudinal edges relative to the base. A closure can be characterized in the examination of the container topology characterized in that the height profile of its upper and lower edge of the cover relative to the base surface and also, for example, the distance of the lower edge of the cover of individual threads of a screw thread are measured. As a result, any damage can be clearly assigned to a specific position on the bottle. In addition, from the container data obtained in each case, the exact position of the label can be derived, for example, in terms of its height position relative to the base as well as in terms of any inclinations. Also, the seat of the closure can be checked based on the container data, namely, whether a possible oblique seat is present, a matching retaining ring is present, etc ..

In addition to these container data resulting from the geometric measurement of the container topology, the invention optionally wins further container data by an optical measurement of the container topology resulting from the color / symbol analysis of the container image. As part of the color analysis of the container image not only the color or base color of the container and / or the liquid filled can be detected and optionally checked, but it can of course also deviations from the base color individual container areas geometrically assign. Likewise, the color analysis of the container image allows a virtually faithful reproduction and reproduction of a colored label, a closure, an associated imprint etc .. These individual color information can be assigned within the scope of the survey performed according to the associated container topology clearly individual positions on the container or its lateral surface become.

For color analysis, a symbol analysis may be added or substituted. As part of the symbol analysis, the container image or the associated container topology is measured with respect to symbols located on the container. These symbols may be printed graphic symbols as label replacement or addition. Usually, however, the symbols are embodied as letter / number combinations and represent, for example, the expiration date applied to the container. By evaluating the container image in the course of the measurement of the associated container topology, the symbol analysis is now carried out, for example, by converting the letter / number sequence into corresponding container data in the sense of text recognition. Of course, the aforementioned container data can also be flanked with position data which reflect the position, extent and optionally color of the examined symbols.

In any case, according to the invention, comprehensive container data are available which are based on the optical and geometric measurement of the container topology the associated container image result. In this case, the container data not only include geometric information about the size, shape, surface finish, etc. of the container per se, but also in the image recording upstream treatment routines on or on the container additionally mounted units and their characteristics, position and seat are checked. These aggregates include, for example, labels and closures. In this case, the respective label can be measured not only in size and location on the respective container, but also allows the additional color / symbol analysis, the inclusion of container data, which reflect the measured color design and any symbols.

Finally, within the scope of the invention, any additional symbols, character or font codes, durability data, etc. applied to the container can also be measured as part of a symbol analysis and converted into associated container data. This container data thus reflects, for example, the information determined in the sense of text recognition (eg OCR, Optical Character Recognition) and their relative position on the container. That is, the container data draw a comprehensive geometric and possibly color and symbolically accurately measured image of the container topology of the examined container.

According to the invention, it is also provided to check the partial results of the inspection with respect to their common occurrence in a single container. So it becomes e.g. allows to automatically check that the color of the contents matches the one on the bottle and that the best before date on the bottle or label is the current date and the known minimum shelf life of the product being bottled.

These container data resulting from the survey of the container topology, and consequently the container, can now be advantageously related to one or more upstream handlers. That is, the comprehensive measurement of the associated container topology based on the recorded container image usually leads to the fact that the container image in question or the container data and the associated container are each individualized. In other words, the container data can be uniquely assigned to a single container image and thus the associated individual container.

- A - It is also provided according to the invention to use the results of the inspection also to control or regulate one or more of the upstream process steps. So z. B. when it is determined in the image processing that the labels located on a bottle are not in the correct orientation or arrangement to each other, for example, the starting point of one or more labeling operations, such as the time of label dispensing in labeling with self-adhesive labels be moved to an earlier or later date to restore the desired alignment of the labels.

If, in the course of the further analysis, one also knows the speed or motion sequences of the respective container before it is taken, then any exemplary errors which are stored in the container data can be precisely assigned to an associated treatment routine or a corresponding treatment system, more precisely even to a single treatment station within one Assign determined treatment plant. So z. For example, specify which filling valve a tuned container was filled, or on which turntable a particular container was positioned during labeling. In fact, the individual treatment routines or associated treatment systems are generally monitored and controlled by a common control unit, which according to the invention cooperates with the image processing unit in such a way that any anomalies of the container topology resulting from the container data can be clearly assigned to an associated treatment facility. It image processing unit and control unit may also be combined into one unit.

For example, the container data of a particular container resulting from the survey of the container topology allow the conclusion that the filling level is insufficient. On the basis of the data from the control unit about the transport of the container in question, in particular the elapsed time from a filling system to the image recording, it is now possible to deduce which of the individual filling valves of a filling machine is responsible for the underfilling described. For example, in the case of the respective filler or individual filler, the determined filling valve may close too early or have a different malfunction.

In a comparable manner, it is possible to deduce from the container data, for example, an obliquely fitting closure. On the basis of the data in the control unit, it is now possible again, the respective Einzelverschließer a Verschließanlage or even the single screw or the individual To identify closure head, which is responsible for this oblique seat. Because this only the speed of the examined container must be evaluated together with the recording time during image acquisition. For all these procedures, it is expressly not necessary to sort out or otherwise investigate the associated container. Rather, only the container image and the container data resulting therefrom or from the measurement of the associated container topology serve to carry out the described error analysis.

According to a further advantageous embodiment, the invention proposes to combine the container data of respective containers to container data pools. These container data pools can be formed according to predetermined criteria, for example, according to the date of image acquisition, the manufacturer of the container, the container material used, the auxiliary equipment used (closure, label, etc.) etc. These container data pools can now be - as well as the respective single image - save and analyze specially. For this purpose, the respective container image or the container data pool can be transmitted to an optionally removed diagnostic unit.

It is conceivable at this point to pick up the container images or the container data resulting from the measurement of the associated container topology on site, that is to say in the case of an inspection device in a bottle treatment plant and store them here if necessary. In addition, the container data in question can then be transmitted to the manufacturer of the bottle treatment plant, which may be located remotely from it. Conventional transmission paths, for example wireless or wired (via Internet, telephone line, etc.), are suitable for the transmission. As a result, the manufacturer of the bottle treatment plant with the aid of the diagnostic unit is able to analyze the individual container images to the extent in which the upstream treatment routines or treatment facilities may be faulty and can teach the user accordingly or take targeted countermeasures. Such a remote evaluation of errors was previously not possible.

Due to the fact that the container data and, if appropriate, the respective container data pool are stored, the described analysis can not only be offset in position, but also be performed with a time delay, for example even years later. This allows the manufacturer of the bottle treatment plant a complete documentation of the filled containers and in particular the answer to the question of whether any errors in the container treatment to later occurring Impaired or not. As a result, consumer safety as well as process reliability and system availability are increased enormously and a complete history of the respective treated containers - even after years and their consumption - can be made available.

In this case, it has additionally proven useful if two or more image pickup devices, which are offset by 90 ° from one another, are realized. Because their respective container images can be assembled in the image processing unit to a two- or three-dimensional container full frame or container partial image. Incidentally, the image processing unit can classify the individual containers on the basis of their respective examined container image taking into account one or more specifications and possibly eject them. For example, it is conceivable to dispose of a certain number of containers for laboratory examinations according to specific specifications or at specific time intervals. Or, targeted investigations are made that relate to specific changes in the container material.

The invention also provides an inspection device for containers, which is advantageous but not restrictive for carrying out the method described and is described in claim 8. Advantageous embodiments of this inspection device are the subject matter of claims 9 and 10.

As a result, a method for inspecting containers and an associated inspection device are proposed, with the aid of which a comprehensive treatment management is possible. Because by the fact that the individual container images (container images or container partial images) are used and evaluated for the measurement of the associated container topology and advantageous at the same time data on the passage of the container in question by a treatment plant are known, can combine a variety of information with each other. Because the time of the optical image acquisition is known and also the speed of the container through the bottle treatment plant or their individual treatment facilities (filling, capping, labeling, etc.) conclusions about the person (s) can be drawn for any anomalies. That is, the errors and their causers can be limited and determined in a previously unprecedented way.

Since the container data are stored in total, they are also locally and temporally independent and virtually unlimited available. This is the Possibility to detect errors in the treatment plant for the containers also by remote diagnosis and to initiate countermeasures. In addition, a complete history can be substantiated in order to be able to verify or rule out any errors and subsequent health problems, etc. that occur. Here are the main benefits.

In the following the invention will be explained in more detail with reference to a drawing showing only one exemplary embodiment. The single FIGURE shows a container inspection device according to the invention schematically.

In the figure, one recognizes an inspection device for containers, in this case for bottles 1. The said bottle 1 has been filled with a liquid 2 to a certain level in an upstream treatment routine in a filling plant. In addition, the bottle 1 a in the context of a likewise upstream treatment routine applied by means of a labeling not shown label 3. In addition, the bottle 1 has an indicated closure 4, which was mounted in a capping. The inspection device and the described upstream treatment plants (filling plant, capping plant and labeling plant) together form essential parts of a bottle treatment line or plants.

With the aid of two container receiving devices 5 arranged offset by approximately 90 ° in the form of CCD cameras in the example case, two or more container images are now generated at a specific instant T. These container images are assembled into a container frame in an image processing unit 6 connected to the container receiving devices or CCD cameras 5. This three-dimensional container frame reproduces the detected container or bottle 1 three-dimensionally, over its entire peripheral surface.

The three-dimensional container frame or, in general, the container image is now analyzed in the image processing unit 6. For this purpose, the image processing unit 6 is set up to measure the associated container topology, as has already been described in the introduction.

That is, in the image processing unit 6, the container frame or the individual container images are measured, among other things also in such a way that the required dimensions as real actual dimensions, ie as a numerical value be determined. This means in the context of the embodiment, a measurement of a length L of the bottle 1, the diameter D, and any variations in diameter. For example, in the present exemplary embodiment, the length L is determined to be 160.34 mm and the diameter D to be 44.70 mm. In addition, the size and position of the label 3 is determined by the extension of the label 3 is measured over an indicated angle range of about 110 ° in the example between transverse edges. In addition, the course of the lower and upper longitudinal edges of the label 3 with respect to a base G is detected on the basis of the container data. This is represented by corresponding height arrows Hi and H ₂ in the drawing. It can be seen that due to the misalignment of the label 3 are lower and, if appropriate, the upper longitudinal edge of the label 3 tilted so that therefrom different values for the one hand and on the other hand Hi H ₂ result. - Finally, the closure 4 is measured as described in the introduction.

Since the time T of image acquisition is stored in the image processing unit 6 and the image processing unit 6 cooperates with a control unit 7 for the container treatment plant or the bottle treatment plant, conclusions can be drawn to the already mentioned labeling, with the help of the label 3 on the container or the bottle 1 has been attached. This means that the illustrated inspection device is integrated in total in the described container treatment plant or bottle treatment plant, with several treatment routines and logically also treatment facilities are upstream, such as the already mentioned labeling, one or more fillers filler for filling the liquid, the Verschließeranlage for mounting the Closure 4 etc.

Since the control unit 7 controls the questionable and upstream and optionally downstream additional treatment plants and specifies the speed of associated and not shown conveyor belts, can be deduced from the distance of the respective treatment plant and the speed of each guided through the container treatment plant containers or bottles 1 in each section which labeling unit has not worked properly in the example, so that it has come to the skew of the label 3 shown in the figure. In the simplest case, the speed V of the container or bottle 1 examined from the labeling unit in question up to the inspection device can be multiplied by the time difference of the labeling time t compared to the image acquisition time T, by the distance or the local position or the path S of the Inspection device back to the error producing labeling unit in the sense of

S = V • (t - T)

to determine. This is of course only to be understood as an example and in a simplistic manner.

The image processing unit 6 is still associated with a memory 8. This may be an internal or external memory 8. The container data or the container data pools already described in the introduction are stored in this memory 8. With the aid of a diagnostic unit 9, the (stored) container data can now be analyzed in a time- and place-offset manner, as has already been described. Incidentally, the image processing unit 6 in conjunction with the control unit 7 can classify individual containers or bottles 1 based on the container data or the respectively examined associated container image taking into account one or more specifications and possibly eject them. For this purpose, the individual containers or bottles 1 are identified and are the control unit 7 corresponding Ausschleus- / sorting commands to a inspection device downstream Ausschleus- / sorting.

Claims

claims:

1. A method for examining containers, in particular bottles (1), according to which at least one container image (5) generates at least one container image and analyzed in a connected image processing unit (6), characterized in that the container image is evaluated for measuring the associated container topology ,

A method according to claim 1, characterized in that container data resulting from the survey of the container topology is related to one or more upstream handlers.

3. The method according to claim 1 or 2, characterized in that the container data of respective containers to container data pools are summarized according to predetermined criteria.

4. The method according to any one of claims 1 to 3, characterized in that the respective container image is stored and transmitted to an optionally remote diagnostic unit (9).

5. The method according to any one of claims 1 to 4, characterized in that two or more staggered container receiving means (5) are realized, the respective container images are assembled in the image processing unit (6) to a three-dimensional container full frame or container partial image.

6. The method according to any one of claims 1 to 5, characterized in that the image processing unit (6) classifies the individual containers on the basis of their respective examined container image taking into account one or more specifications and possibly ejects.

7. The method according to any one of claims 1 to 6, characterized in that the image processing unit (6) and a control unit (7) for one or more treatment plants work together in such a way that resulting from the container data any anomalies of the container topology clearly associated Treatment plant and / or an associated treatment station can be assigned.

8. The method according to any one of claims 1 to 7, characterized in that at least two individual results of the analysis of the container topology are checked to see whether they are allowed in this combination.

9. The method according to any one of claims 1 to 8, characterized in that at least one result of the analysis of the container topology for influencing at least one preceding process step is used.

10. The method according to any one of claims 1 to 9, characterized in that at least for a result of the analysis of the container topology is an actual measure.

11. The method according to claim 10, characterized in that the actual measure is a numerical value.

12. Inspection device for containers, in particular for bottles (1), preferably for carrying out the method according to one of claims 1 to 11, with at least one container receiving means (5) and a connected image processing unit (6), wherein the container receiving means (5) receives a container image which is analyzed in the image processing unit (6), characterized in that the image processing unit (6) is adapted to measure the container image with respect to the associated container topology.

13. Inspection device according to claim 12, characterized in that two or more than CCD cameras designed container receiving means (5) are offset from one another to produce a two- or three-dimensional container full frame or container partial image.

14. Inspection device according to claim 12 or 13, characterized in that the image processing unit (6) has an external and / or internal memory (8) for permanent storage of the container images and optionally time and / or location offset analysis.