Method for identifying a section of a material web

Abstract

The invention relates to a method for identifying a section of a material web (10), comprising the steps of: a) recording a digital image (B) of a structural feature (32) in the section of the web to be identified, and b) storing the image together with location information (s, py, x) indicating the location of the structural feature on the material web, characterised in that in step a) a web movement control is carried out on the section of the web to be identified, by means of which web movement control the lateral position (x) of the web is controlled to a target value (s), and in that in step b) the target value (s) is stored as part of the location information.

Description

[0001] METHOD FOR IDENTIFYING A SECTION OF A MATERIAL WEB

[0002] The invention relates to a method for identifying a section of a material web, comprising the steps of: a) recording a digital image of a structural feature in the section of the web to be identified, and b) storing the image together with location information indicating the location of the structural feature on the material web.

[0003] In manufacturing processes, it is often desirable or even required for quality assurance or liability reasons to document and individually trace relevant manufacturing steps performed on individual products. This can be achieved, for example, by applying a marking to each product, such as a label, imprint, or engraving, that identifies the product and—if necessary in conjunction with data stored in a database—provides information about when, where, and by whom the relevant processing step was performed on the product. However, this often has the disadvantage that the marking can impair the appearance or even the functionality of the product.

[0004] An alternative way of identifying products takes advantage of the fact that most products have certain (micro) structural features that only occur in this form in a single product and therefore enable individual identification. Examples of such structural features include fiber structures in textiles or paper goods, micro surface structures in metal foils or, for example, brushed or sanded metal surfaces, grain patterns in wood materials, and the like. Instead of applying a marking to the product, a certain section of the product's surface within which the structural feature in question can be recognized is photographed with a high-resolution camera, and the image is then stored digitally, possibly together with location information that indicates where the structural feature can be found on the product in question.This makes it possible to recognize the structural feature on the product at a later point in time and thus to identify the product.

[0005] For web materials such as paper webs or endless webs of metal sheet, the storage of location information is particularly important because the web often has no or too few features visible to the naked eye that could be used as a guide when searching for microstructural features. The location information to be stored can then, for example, consist of location coordinates that specify the position of the photographed surface section of the web in the longitudinal and transverse directions of the web. The more precise this location information is, the less complex the subsequent search for the relevant structural feature will be. In order to obtain the most precise information possible about the position of the photographed surface section in the transverse direction of the web, it is advisable to select a surface section that borders a longitudinal edge of the web so that the edge is visible in the recorded image and can serve as a reference for the position in the transverse direction of the web.

[0006] However, this limits the freedom in selecting the surface section to be photographed. This is particularly problematic in cases where the surface areas near the edges of the web are, for example, painted or polished and therefore do not exhibit suitable structural features, as well as in cases where the surface areas near the edges are painted, polished, or coated after the image is taken, so that the structural feature is no longer visible. In cases where the edges of the web are trimmed or edged in a subsequent processing step, later retrieval of the structural feature can be significantly more difficult or even impossible.

[0007] The object of the invention is therefore to provide a method which offers greater freedom in the choice of the lateral position of the structural features.

[0008] This object is achieved according to the invention in that in step a) a web guide is carried out on the section of the web to be identified, by which the lateral position of the web is controlled to a desired value, and in that in step b) the desired value is stored as part of the location information.

[0009] With this method, the structural features to be photographed can also be located at a greater distance from the web edge. This allows the camera used to capture the structural features to be positioned stationary over a lane of the material web where the structural features are particularly clearly visible and / or where particularly low-noise images of these structural features can be captured, or where particularly favorable lighting conditions exist. The stored target value indicates the lateral position of the web relative to the position of the camera used to capture the structural feature, thus facilitating retrieval of this feature.

[0010] Web guiding devices are usually already present in systems for processing material webs. According to the invention, the longitudinal position of the camera used to record the structural features is selected relative to the web guiding device such that the lateral position of the web, especially in the area where the camera is located, corresponds to the target value with high accuracy. The target value can then, for example, directly or indirectly indicate the lateral distance between the camera position and an edge of the web (or another suitable reference), so that when later searching for the structural feature, one knows exactly where to position the camera relative to the edge of the web.

[0011] Advantageous embodiments and further developments of the method are specified in the subclaims.

[0012] The method is suitable, for example, for applications in which an inspection system is used to detect defects in the material web that would impair the quality of the final product and / or could lead to damage or malfunctions in subsequent processing steps. Using the method according to the invention, the beginning and end of the web section containing the detected defect can then be identified. For this purpose, the camera that records the structural feature can be positioned at such a longitudinal distance from the inspection system and triggered in time that the recorded structural features are located at the beginning and end of the section to be identified.If processing steps are carried out on the web that change the length or width of the web, for example cutting out damaged areas from the web, trimming the web, or treatments that lead to stretching or shrinking of the web, the stored location information can be adjusted accordingly.

[0013] The invention also relates to a system for processing material webs, comprising a web guide device, a camera for recording structural features on the material web, and a control device configured to carry out the method described above.

[0014] If the web guiding system includes a detector that provides a feedback signal for the lateral position of the material web, the camera can be positioned approximately in the same longitudinal position as this detector. In this case, the detector signal can also be used to detect deviations in the web's position from the target value and compensate for them when generating the position information about the structural feature.

[0015] In the following, exemplary embodiments are explained in more detail using the drawings.

[0016] They show:

[0017] Fig. 1 is a floor plan sketch of a device for inspecting a material web and for identifying a defective web section;

[0018] Fig. 2 and 3 show the device according to Fig. 1 in later process stages;

[0019] Fig. 4 is a flowchart for a method according to the invention; Fig. 5 is a plan view of a device for retrieving the identified material from a section;

[0020] Fig. 6 shows the device according to Fig. 5 at a later stage of the process;

[0021] Fig. 7 and 8 show the material web shown in Fig. 6 before and after cutting out a defective web section;

[0022] Fig. 9 is a flowchart for a method for retrieving the defective track section using the device according to Fig. 5; and

[0023] Fig. 10 is a floor plan sketch of a device for retrieving an identified track section according to another embodiment.

[0024] Fig. 1 shows a portion of a material web 10 running over transport rollers 12, 14 and being transported in the direction of arrow A (y-direction). The transport roller 12 is part of a web guide control device 16, with which the lateral position of the web (in the x-direction) is controlled to a setpoint value s in a known manner. The setpoint value s set in the control device is reported to a control device 18. Downstream of the web guide control device 16, a detector 20 is arranged, which detects an edge of the material web 10 and reports its lateral position x back to the control device.

[0025] An inspection camera 22 is arranged between the transport rollers 12 and 14 above the material web 10. It serves to detect defective areas 24 on the material web. The camera is, for example, a line scan camera that extends across the entire width of the material web 10 and records a digital image of the passing material web. A stream of inspection image data IB generated in this way is transmitted to the control device 18, where it is digitally evaluated.

[0026] Another digital camera 26, which has a much smaller field of view 38 than the inspection camera 22 and only captures a small area of ​​the material web, is arranged at a distance Ay downstream of the inspection camera 22 in a longitudinal position which is approximately at the same height as the detector 20.

[0027] In the situation shown in Fig. 1, the front end of the defective area 24 has just reached the position of the inspection camera 22. The control device 18 therefore recognizes, by evaluating the inspection image data IB, that a defective section of the material web has begun and then calculates a time period Δt that the front end of the defective area 24 will require until it approximately reaches the position of the camera 26. This time period Δt depends on the distance Δy and on the transport speed ν of the material web 10. An increment encoder 30 is arranged on the transport roller 14, which counts the angular increments of the rotation of the transport roller 14 and uses this to calculate the transport speed ν of the material web as well as a position parameter νp that indicates the current longitudinal position of the material web relative to the camera 26.In other words, the position parameter py indicates how many meters or centimeters of the material web have moved past the camera 26 since the beginning of the production process. The values ​​v and py are reported to the control device 18 so that it can, among other things, calculate the time interval Δt.

[0028] Fig. 2 shows the state at the moment the time interval Δt has elapsed. At this time, the control device 18 sends a trigger signal T to the camera 26, which then records an image of a structural feature 32, i.e., an image of the surface structures of the material web 10 in a surface section that corresponds to the field of view 28 of the camera and that passes under the camera 26 at this moment. The camera 26 then returns image data B representing this image to the control device 18.

[0029] In the example shown here, the time interval Δt is slightly smaller than the quotient Δy / v. This takes into account, on the one hand, that the image processing system in the control device 18 requires a certain amount of time to detect the edge of the defective area 24 in the digital image data. On the other hand, this ensures that the structural feature 32 is recorded at a location that is slightly in front of the front end of the defective area 24 and therefore definitely within the section marked by the structural feature 32. At the same time, this ensures that the image of the structural feature 32 is not distorted by the defective area 24.

[0030] The image data B are stored in a database by the control device 18 together with the position parameter py and the target value s. In the example shown, the lateral position x of the edge of the material web, which is measured by the detector 20, is also stored. If the web guiding device 16 does not operate with absolute precision, the measured lateral position x may deviate from the target value s. If the structural feature 32 is to be found later on the material webs 10, this target / actual deviation can be taken into account when searching for the structural feature.

[0031] Fig. 3 shows the situation at a slightly later point in time. The defective area 24 has passed the inspection camera 22. At the moment the rear end of the defective area 24 has passed the line scan camera, image analysis determined that the defective web section has come to an end. This restarted a timer that counts a delay time. At the moment shown in Fig. 3, this delay time has elapsed, and a trigger signal is again transmitted to the camera 26, with the result that the camera photographs another structural feature 34 and transmits digital image data B to the control device 18. In this case, however, the delay time is somewhat longer than in the situation shown in Fig. 2, with the result that the recorded structural feature 34 lies slightly behind the rear end of the defective area 24.The current transport speed v and the position parameter py are measured again by the increment encoder 30 and transmitted to the control unit 18. These data are stored in a new data record in the database together with the lateral position x, the image data B, and the target value s.

[0032] The essential steps of the method described above are shown in a flowchart in Fig. 4. In step S1, the inspection camera 22 and the image evaluation system are used to check whether a new defective area has been detected on the web or whether the end of a previously detected defective area has been reached.

[0033] As long as this is not the case, step S1 is repeated at regular (short) intervals. If the beginning or end of a defective area is detected (J), a new data record is created in the database in step S2. Optionally, data that further specifies the type and lateral position of the detected defect is stored in this data record. In step S3, the timer that measures the time interval Δt is then started. In step S4, it is checked whether the time interval Δt has elapsed. As soon as this time interval has elapsed, the camera 26 is triggered in step S4 so that an image of the structural feature 32 or 34 is recorded. The image data are saved in the newly created data record together with the target value s, the transport speed v, the position parameter py and the lateral position x. The process then returns to step S1.

[0034] In this way, the material web 10 is continuously provided with virtual markings throughout the entire production or processing process, identifying defective sections of the material web. At the end of the production or processing process, the material web 10 can, for example, be wound into a coil and stored or transported to a facility where further processing steps are carried out.

[0035] Fig. 5 shows a schematic floor plan of such a system, with which further processing steps are carried out on the material web 10. The further processing here specifically consists in locating the defective areas 24 and separating them from the material web. For this purpose, the material web 10 is pulled off the coil with the aid of transport rollers 36, 38 in the direction indicated by an arrow B. An increment encoder 40 is arranged on the transport roller 38, which then counts down the position parameter py. In addition, the transport speed v is measured again and transmitted to the control device 18 together with the position parameter py.

[0036] The transport roller 36 is again part of a web guide control device, with which the lateral position x of the web is controlled to the target value s provided by the control device 18. The web guide control device 42 works together with a detector 44, which detects the current lateral position x of an edge of the material web 10 and reports it to the control device 18 and to the web guide control device 42.

[0037] A camera 46 is arranged above the web near the detector 44, with which a surface section of the web located in a window 48 can be continuously observed. The resulting image data B' is then transmitted to the control device 18 and compared there with the stored image data B, which represent the previously recorded structural features 32, 34. The lateral position of the camera 46 relative to the detector 44 corresponds to the lateral position of the camera 26 relative to the detector 20. Due to the web guide, it is therefore ensured that the structural features 32, 34 will pass through the window 28. The width of this window is slightly larger than the width of the structural features themselves, so that the structural features can still be detected even with slight positional deviations.

[0038] In the situation illustrated in Fig. 5, the most recently recorded structural feature 34 is currently approaching the position of camera 46. The control device 18 compares the position parameters py continuously supplied by the increment encoder 40 with the position parameter stored for the structural feature 34. In the example shown, this comparison is simplified by the fact that the distance measured in the y-direction between the increment encoder 40 and the detector 44 is the same as the distance between the increment encoder 30 and the detector 20 in Fig. 1.

[0039] If the control device 18 determines that the current position parameter py is approaching the value stored for the structural feature 34, the transport speed v is reduced, and the camera 46 is activated. The stored image of the structural feature 34 is "slided" pixel by pixel through the wider window 48 and compared with the currently recorded camera image. Methods for such image comparisons are known in the art. For example, the image data can be reduced to a color in which the structural features are particularly clearly visible using color filters, and the image can then be binarized so that each pixel has either the value 0 or 1, and thus the pixels of the stored image can be directly compared with the pixels of the camera image.Known statistical methods can be used for the comparison, in which a specific similarity measure is calculated that indicates the extent of the deviations between the images. The comparison can be carried out line by line as the trajectory, and thus the structural feature, moves beneath the window 48. The similarity measures found for the successive lines are accumulated, and if the accumulated similarity measure exceeds a specific threshold, the structural feature is deemed to be recognized. Fig. 6 shows the situation in which the structural feature 34 has reached the window 48 and is recognized. The control device 18 then causes the trajectory to be stopped and a cut mark 50 to be applied. In the example shown, the cut mark 50 is located in the same y-position as the trailing edge of the structural feature 34 in the direction B.

[0040] Once the cut mark 50 is set, the web is again transported in direction B, possibly temporarily at a slightly increased speed, until the structural feature 32 approaches the camera 46. Then, the web is slowed down again, and the process described above is repeated to detect the structural feature 32.

[0041] Once the structural feature 32 has been detected, another cutting mark 52 is set, as shown in Fig. 7. Using a cutting device (not shown), the web is cut along the cutting marks 50, 52, and the web section in between, which contains the damaged area 24, is severed. The two remaining web sections are then connected to one another by a splice seam 54, as shown in Fig. 8. Due to the selected position of the cutting marks, the structural features 32, 34 remain undamaged and are still present on the web, so that the splice seam 54 can be relocated at a later time if necessary. The stored location data for the cutting marks, in particular the position parameters py, are corrected according to the length of the severed web section.

[0042] The following describes a further development of the method, which can also be used to handle applications in which the material web 10 can be stretched or shrunk during the processing, so that the length and / or width of the web and thus also the position data of the structural features change. In Fig. 1, in addition to the detector 20, a further detector 56 is shown, which detects the opposite edge of the material web 10 and determines its lateral position, so that by comparing it with the lateral position measured by the detector 20, the width Ax of the web can be measured and transmitted to the control device 18.

[0043] The processing station shown in Fig. 5 also has, in addition to detector 44, another detector 58 for the other edge of the web, so that the current width Ax of the web can be measured here as well. If the web width measured with detectors 44, 58 deviates from the width previously measured with detectors 20, 56, the stored lateral positions of structural features 32, 34 are scaled proportionally to the width change, so that they again indicate the true positions of the structural features after the web has been stretched or shrunk.

[0044] Assuming that the shrinkage or expansion of the web is isotropic, the change in length of the web can also be estimated, so that, if necessary, the point in time at which the transport speed v of the web is reduced in Fig. 5 can be adjusted to enable real-time detection of the structural features. Once the first structural feature 34 has been detected, the change in length can be directly measured by comparing the position parameter py supplied at this point in time by the increment encoder 40 with the stored value, so that the structural features that mark any further defective locations on the web can be approached more precisely.

[0045] The essential steps of the method for retrieving structural features, for example structural feature 34 in Fig. 5, are shown as a flow chart in Fig. 9. In step S10, the web stretching (or shrinkage) is compensated for in the manner described above. Subsequently, in step S11, the target value s for the edge of the web measured with detector 44 in Fig. 5 is calculated. In the example shown, this edge forms the reference for the lateral position of the web. Alternatively, any other track on the material web that extends in the longitudinal direction of the web can be used as a reference. If, as in the example shown, the lateral distances between detector and camera in Figures 1 and 5 match, the stored value for the target value s can simply be adopted.

[0046] The web is then rewound, with the position parameter py being counted back from the maximum value reached at the end of the web (when wound onto the coil) until it has decreased to the value py(34) - A. py(34) is the value stored for the structural feature 34 in the database, and A corresponds to the distance between the structural feature 34 and the camera 46 in Fig. 5, i.e., at the time at which the transport speed is reduced.

[0047] Subsequently, in step S12, the camera 46 is activated, the image comparison is started, and the web is rewound at a reduced transport speed until it is determined in step S13 that the structural feature 34 has been found. Subsequently, in step S14, the cutting mark 50 is set.

[0048] The same procedure is then carried out accordingly for the structural feature 32 and, if applicable, for further structural features recorded with the system shown in Fig. 2.

[0049] In certain applications, it may be necessary to locate structural features in real time along the same web path in which they were recorded. One example would be an application where the defects detected by the inspection system are foreign bodies lying or adhering to the web, which could cause damage to downstream components of the system, such as the calender rolls used to calender the web. For design reasons, it is often not possible to position the inspection system directly upstream of the calender rolls. This is because certain processing steps and equipment are required immediately upstream of the calender, leaving no space for the inspection system. On the other hand, in certain cases it may be impossible to remove the detected foreign bodies directly downstream of the inspection system, for example because the web must first dry or cool down.

[0050] Fig. 10 shows a schematic floor plan of a system arranged in the transport direction of the material web 10 between the inspection system shown in Fig. 1 and the processing station in which the foreign bodies on the web would cause damage.

[0051] A web guide control device 56 regulates the lateral position of the web to the target value s supplied by the control device 18. The actual position of the web edge is measured by a detector 58 and fed back to the control device. A camera 60 is arranged near the detector 58 and serves to continuously record the surface structure of the web 10 in a strip 62 extending in the longitudinal direction of the web. The lateral position of the camera 60 relative to the web edge is the same as that of the camera 26 in Fig. 1.

[0052] An increment encoder 66 is arranged on a transport roller 64, which measures the position parameter py and is calibrated such that it provides the same position parameter for each point along the path as the increment encoder 30 in Fig. 1.

[0053] In the situation shown in Fig. 10, the defective area 24 is just approaching the camera 60. This camera is not yet active. The control device 18 compares the position parameter supplied by the increment encoder 66 with the position parameter stored for the structural feature 32. When the difference between these parameters has decreased to a certain value (preferably dependent on the transport speed), the camera 60 is activated by a trigger signal T, and the camera begins recording and supplies image data B' to the control device. This image data is continuously compared with the stored image of the structural feature 32. When this structural feature reaches the position of the camera 60, it is detected by image comparison.However, the web transport and image comparison are continued and, if necessary, interrupted when the control device 18 determines that the structure feature 32 and the defective area 24 have passed through the downstream (not shown) process stations and are approaching the calender rolls. Only when the structure feature 34 is also detected during the image comparison is the web transport stopped and the defective area 24 eliminated, either by removing the contaminants from the web or by separating the affected section and subsequently splicing the web. Since separating a web section changes the position parameters py for the subsequent web points, the stored position parameters for subsequent structure features are adjusted accordingly.When the last of the stored structural features has been found, the camera 60 can be deactivated again until a new defect is detected and approaches the position of the camera 60.

[0054] The method described here is not only suitable for applications in conjunction with an inspection system, but can generally be used to enable the tracing of a material web. For example, the camera that records the structural features (like camera 26 in Fig. 1) can be triggered at regular time intervals, independently of the detection results of an inspection system, in order to record a structural material on the material web 10 at regular longitudinal intervals. By accessing the data stored in the database, each section of the material web can then be identified later, if necessary, whereby the spatial resolution with which the identification is possible is determined by the grid of the recorded structural features.

Claims

PATENT CLAIMS 1. Method for identifying a section of a material web (10), comprising the steps of: a) recording a digital image (B) of a structural feature (32, 34) in the section of the web to be identified, and b) storing the image together with location information (s, py, x) which indicates the location of the structural feature on the material web, characterized in that in step a) a web guide is carried out on the section of the web to be identified, by which the lateral position (x) of the web is regulated to a desired value (s), and that in step b) the desired value (s) is stored as part of the location information.

2. Method according to claim 1, in which defective areas (24) of the material web (10) are detected by an inspection system (22), and in which structural features (32, 34) of locations on the web are recorded, which, measured in the longitudinal direction of the web, mark the beginning and the end of the defective area.

3. Method according to claim 1 or 2, combined with a method for retrieving the recorded structural features (32, 34), in which at a position in the longitudinal direction of the material web (10) in which a camera (46, 60) for recording an image of a part of the surface of the material web is located, the lateral position of the material web (10) relative to the position of the camera (46, 60) is controlled to the same setpoint value (s) as in step a).

4. The method according to claim 3, wherein the stored location information comprises a position parameter (py) indicating the longitudinal position of the structural feature (32, 34) on the path, and wherein the camera (46, 60) used to retrieve of the structural features, is triggered depending on a relationship between the stored position parameters (py) and the current position parameter for the location of the camera.

5. The method according to claim 4, wherein the transport speed (v) of the material web (10) is reduced when a structural feature (32, 34) approaches the camera (46).

6. Method according to one of the preceding claims, in which a stretching or compression of the material web (10) is measured in the longitudinal and / or transverse direction of the web and the stored location information is corrected according to the measured stretching or compression.

7. Plant for producing or processing a material web (10), with a camera (26) arranged on the web for recording an image of a structural feature (32, 34) on the web, a web travel control device (16), and a control device (18) which is configured to carry out the method according to one of claims 1 to 6.

Images