Method and apparatus for evaluating the sealing performance of a carton package

Abstract

Disclosed is a method (800) for evaluating seals (200, 202) of a carton package (114) for storing food (FP). The seals (200, 202) are performed using ultrasonic sealing technology. The method includes capturing (802) an image dataset (500) depicting the seals (200, 202), the image dataset (500) comprising a first data subset (600) associated with a first color channel and a second data subset (602) associated with a second color channel, identifying (804) adhering portions (702a, 702b) using the first data subset (600), and identifying (806) non - adhering portions (704) using the second data subset (602).

Description

Technical Field

[0001] The present invention relates to food packaging technology. More specifically, it relates to the evaluation of the seal of a carton package using ultrasonic sealing technology.

Background Art

[0002] Currently, it is well known to use induction heat sealing to form seals (sometimes called welds) of carton packages. In a roll-fed carton filling machine, it is common to seal the opposing ends of a web of packaging material to form a tube. This step is called a longitudinal seal. Once the tube is formed, food such as milk is supplied into the tube from above. At the lower end of the tube, a transverse seal is performed. By simultaneously performing two transverse seals and cutting between the two seals, a pouch filled with food is formed from the lower end of the tube. In subsequent steps, this pouch can be folded, for example, into a brick-shaped package.

[0003] In induction heat sealing, since a layer of packaging material that generates heat in response to a magnetic field generated from an inductor of a transverse heat sealing device is required, other sealing technologies have been developed so that reliable and efficient sealing can be achieved even with packaging materials that do not use an aluminum layer or the like. One of them is ultrasonic sealing. Instead of inducing eddy currents in the packaging material to generate heat, as a result, melting the outer and inner plastic layers of the packaging material, an ultrasonic sealing device uses ultrasonic waves that interact with the plastic material to generate heat so that heat is generated.

[0004] To protect the food held within the package, not only is the food prevented from being released from the package by both a longitudinal seal and two transverse seals, one at each of the upper and lower ends of the package, but also microorganisms such as bacteria and germs are prevented from entering the package. To provide a cost - efficient package, thousands of packages are manufactured per hour. Due to the high speed, there is less margin for performing a reliable seal. Conversely, in order to generate sufficient heat in the transverse seal area of the package and still meet the time limit required for high - speed operation, it is necessary to appropriately set the combination of packaging materials and the settings of the transverse sealing device. Furthermore, in order to confirm that a reliable transverse seal has been applied to the manufactured packages, samples for quality evaluation are taken regularly.

[0005] Quality evaluation can be carried out in various ways. Currently, it is common to manually evaluate the seal using a microscope. By using a quality checklist in combination with a microscope, abnormalities in the package due to inappropriate settings of the transverse sealing device or the packaging material not meeting the set requirements can be detected. Once a quality problem is identified, the transverse sealing device and the packaging material can be analyzed in more detail.

[0006] In the current quality assessment, with very few exceptions, there is a risk that food cannot be protected from the surrounding environment, and packages with seals that are at risk of starting to leak are identified. However, with the increase in the speed of the filling machine, the time for sealing the package laterally becomes shorter, and the risk of insufficient sealing is increasing. Furthermore, the demand for cost efficiency is increasing, and the thinning of the layers of packaging materials, etc., may also increase the risk of insufficient sealing. In addition, the requirements for the environment are increasing, for example, materials such as aluminum foil are being avoided, and the risk of insufficient sealing also increases. As a result, even if the sealing device of the filling machine is reliable, when the requirements increase and a situation occurs where the sealing becomes insufficient in these reliable sealing devices, it is desirable to be combined with a reliable seal quality assessment in order to surely identify this. Once identified, measures can be taken that bring about the positive effect of reducing the risk of shipping insufficient packages and keeping product waste low.

Summary of the Invention

Problems to be Solved by the Invention

[0007] An object of the present invention is to at least partially overcome one or more of the above limitations of the prior art. In particular, it is an object to provide a method for evaluating the sealing performance of a carton package in a highly reliable and cost-effective manner.

[0008] According to a first aspect, there is provided a method for evaluating the seal of a carton package arranged to hold food, the seal being generated by using ultrasonic sealing technology, the method comprising: capturing an image dataset depicting the seal, the image dataset including a first data subset associated with a first color channel and a second data subset associated with a second color channel, identifying the bonded portion using the first data subset, identifying the non-bonded portion using the second data subset.

[0009] The sticking part, that is, the part where the first and second surfaces of the packaging material are adhered to each other. For example, as an effect of the melting and bonding of the plastic layers on both sides, the advantage of using different subsets to identify the sticking part is that both the sticking part and the non-sticking part are positively identified. Conversely, the non-sticking part is not regarded as the part that is not the sticking part. In this way, by positively determining both the sticking part and the non-sticking part, a more reliable seal quality evaluation can be realized.

[0010] A further advantage is speed. In contrast to the general approach used for quality evaluation of induction heating seals that identify the print formed from the bulge of the induction coil, the approach proposed herein does not require identifying such a print when evaluating ultrasonic seals.

[0011] The heat generation by ultrasonic sealing technology is different from that by induction heat sealing technology. Since there is less contact between the sealing device and the packaging material, the packaging material is less likely to be affected by the mechanical effects of ultrasonic sealing technology. As an effect of this, by analyzing the first subset and the second subset linked to different color channels, it becomes possible to perform a reliable seal evaluation. This is because these color channels can be respectively linked to how the light is affected when passing through the sticking part and the non-sticking part.

[0012] The method further comprises further identifying the sealed part using the first and second data subsets, where the sealed part corresponds to the top fin, bottom fin, or longitudinal seal part of the carton package, both the sticking part and the non-sticking part are included within the sealed part.

[0013] The advantage of identifying both the adhered and non - adhered parts using different subsets of the image dataset is that it can provide a reliable identification of the sealed part by combining both at a later stage. The adhered part may be a section where two sides of the packaging material are joined. On the other hand, the non - adhered part is a portion where both sides of the packaging material are in contact with each other but not integrated. The non - adhered part can be used to extract characteristic lines, such as product lines or cut lines, by using the average of the lines formed from the edges of the adhered and non - adhered parts. In this way, a more robust line fitting is achieved.

[0014] The image dataset may be captured using a camera placed on a first side of the seal of the package, and the method further comprises illuminating the sealed part using a light source provided on a second side of the sealed part, the second side of the sealed part being opposite to the first side of the sealed part, whereby the light emitted from the light source passes through the sealed part of the package and irradiates the camera.

[0015] The light provided can be natural light, diffused light, polarized light. Since different types of light may be associated with different advantages, the image dataset comprises a series of image datasets captured in natural light, diffracted light, and polarized light. For all light options, first and second data subsets can be captured.

[0016] The light source may be one or more light - emitting diodes (LEDs) that emit light at at least one identification frequency, and the first frequency range of the first color channel and / or the second frequency range of the second color channel are set based on at least one identification frequency of one or more LEDs.

[0017] The advantages of using LEDs are that they can be manufactured cost - effectively and can generate light at the identification frequency.

[0018] The first frequency range of the first color channel and / or the second frequency range of the second color channel are set based on the characteristics of the packaging material.

[0019] By adapting the light source to the details of the identification of the packaging material, a more reliable seal quality assessment can be achieved.

[0020] The image dataset may be captured in multiple versions, and the multiple versions have different light settings, where the light settings include the frequency of the identification of the light emitted from the light source and / or the amplitude of the light emitted from the light source.

[0021] As described above, in addition to multiple versions, it is also possible to use different types of light.

[0022] The packaging material may include a non-metallic light barrier layer, the non-metallic light barrier layer is a cellulose-based light barrier layer, and the cellulose-based light barrier layer is adapted to block light having a frequency that causes the growth of microorganisms and / or the loss of vitamins in the food held in the carton package.

[0023] The advantage of using the non-metallic light barrier layer is that it can reduce the impact of the packaging material on the environment.

[0024] The method further comprises providing a reference object adjacent to the seal of the carton package, the reference object having a reference length and / or a reference width, identifying the reference object depicted in the image data, determining the absolute length and / or the absolute width of the fixed part and / or the non-fixed part using the reference length and / or the reference width.

[0025] The method further comprises applying a substance such as ink to the seal before capturing the image dataset.

[0026] By applying a substance to the seal, instead of passing through the adhered and non - adhered parts, light passes through the parts with the applied substance. This advantage is that the substance also affects the passage of light and does so in different ways for different subsets. Thus, by wisely selecting the substance and the frequency of light in consideration of the characteristics of the packaging material, a reliable seal quality evaluation becomes possible.

[0027] The image data set may further include a third data subset related to a third color channel, and the method further comprises identifying the parts affected by the substance using the third data subset. The step of identifying the seal part is performed using the first, second, and third data subsets.

[0028] By using the third data subset and having a substance selected based on the third color channel linked to this data subset, it becomes possible to more easily find the parts affected by the substance. This part affected by the substance can be the adhered part.

[0029] The seal part may comprise an overlap subsection where the horizontal seal and the vertical seal overlap.

[0030] The problem with a carton package having a vertical seal and a horizontal seal is that they have an overlapping part. When the packaging material is three - layer, problems occur and the risk of an insufficient seal cannot be ignored. Therefore, performing a reliable seal quality evaluation is particularly important for packages with such overlapping parts.

[0031] The method further comprises, before the step of capturing the image data set, taking out the stored food, unfolding the package so that the seal is exposed, dissolving the packaging material so that the carton layer of the packaging material is removed.

[0032] The advantage of removing the carton layer is that the remaining layers can be made transparent, thereby allowing light to pass through the seal.

[0033] According to a second aspect, there is provided an apparatus for evaluating a seal of a carton package arranged to hold food, the apparatus being generated by the seal using ultrasonic sealing technology, the apparatus comprising: A camera configured to capture an image dataset depicting the seal, the image dataset including a first data subset associated with a first color channel and a second data subset associated with a second color channel, the camera; A control unit comprising a processor and a memory and configured to identify fixed parts in the image dataset using the first data subset and non-fixed parts in the image dataset using the second data subset.

[0034] The same advantages shown above for the first aspect also apply to this second aspect.

[0035] The apparatus may further comprise: A light source arranged to illuminate the seal; The camera is arranged on a first side of the seal portion of the package, the light source is arranged on a second side of the seal portion, and the second side of the seal portion is on the opposite side of the first side of the seal portion; Thereby, the light emitted from the light source passes through the seal of the package and irradiates the camera.

[0036] The light source may be one or more light emitting diodes (LEDs) that emit light at at least one identifiable frequency, and the first frequency range of the first color channel and / or the second frequency range of the second color channel is set based on at least one identifiable frequency of the one or more LEDs.

[0037] Further objects, features, aspects, and advantages of the present invention will become apparent from the following detailed description and drawings.

Means for Solving the Problems

[0038] Hereinafter, exemplary embodiments of the present invention will be described with reference to the accompanying drawings.

Brief Description of the Drawings

[0039]

Figure 1

Figure 2

Figure 3

Figure 4

Figure 5

Figure 6

Figure 7

Figure 8

Modes for Carrying Out the Invention

[0040] Referring to FIG. 1, a filling machine 100, also called a packaging machine, is schematically illustrated. As shown, a reel 102 of packaging material PM may be provided. In many cases, the packaging material PM comprises multiple layers such as a carton layer for providing robustness and environmental advantages, an inner plastic layer for avoiding direct contact between the carton layer and the food FP, and an outer plastic layer for protecting the carton layer from the surrounding environment. When manufacturing a package, the packaging material PM is provided by unwinding the reel 102 such that a web 104 is provided. The web 104 may be sterilized or subjected to other treatments such that microorganisms are removed. In the example shown in FIG. 1, a sterilization device 106 is provided above the filling machine 100. The sterilization device 106 may be an LVEB device configured to kill microorganisms using a low-voltage electron beam. Also, it is conceivable to kill microorganisms using a hydrogen peroxide bath.

[0041] After treating the web 106 such that microorganisms are removed, the web 104 can be formed into a tube. Once formed into a tube, the food FP is filled into the tube from above via a filling pipe 108. A longitudinal sealing device 110 can be provided such that the tube is sealed in the longitudinal direction.

[0042] After being sealed in the longitudinal direction, the tube is fed in the feed direction FD to a transverse sealing device 112 where a transverse seal is performed. In many cases, two transverse seals are performed simultaneously, forming the upper seal of one package and the lower seal of another package at the same time. When these two seals are formed, a cut is made between the two seals and the lowermost package is released from the tube. After the transverse seal is applied, the package 114 is folded and formed into a final shape such as a brick shape.

[0043] A roll-fed filling machine is illustrated, but other types of filling machines are also available. For example, as another option, there is a so-called blank-fed filling machine. In this type of filling machine, a blank, i.e., a flat-folded sleeve of packaging material, is supplied and formed into a package filled with food.

[0044] Figure 2 shows the package 114 in more detail. As shown, after being supplied from the horizontal sealing device 112, the package 114 may include a first seal 200 at one end of the package and a second seal 202 at the other end. In a subsequent step, in a so-called folding device (not shown in FIG. 1), the package 114 can be folded into its final shape. For example, a flat top and a flat bottom may be formed, and the horizontal seals may be folded so that they contact the flat top and the flat bottom respectively. The packaging material PM may be provided with a weakened line to facilitate folding.

[0045] In order to ensure that the horizontal seal meets the set quality requirements, an image dataset depicting the horizontal seal can be captured and analyzed. One approach for such analysis is shown in FIG. 3. As shown, in the first step, the image data of the seal may be recorded, that is, the image dataset may be captured. This step can be performed in various ways. One of these methods is to take samples from the manufactured packages and analyze them in detail. The number of samples may vary depending on the filling machine, and the sampling frequency may vary depending on, for example, the packaging material PM used. In the case of a carton package containing several layers (one of these layers is made of carton), it has been proven beneficial to dissolve the package so that the carton layer is removed. When the carton layer is removed, it becomes easier to analyze the plastic layers that are adhered to each other to form the horizontal seal. Furthermore, since the plastic layers are often transparent, dissolving the package to remove the carton layer also has the advantage that light can be irradiated through the remaining layers of the packaging material PM.

[0046] In the image analysis step, the image dataset captured in the seal image recording step can be analyzed to detect various deviations. For example, if it is found that the width of the horizontal seal is below a preset threshold, a notification can be sent to the operator.

[0047] In addition to identifying the deviations indicating quality problems, the features extracted in the image analysis step may be sent to the seal quality characteristic evaluation step. In this step, various features extracted from the image analysis step can be further analyzed.

[0048] Data from the seal quality characteristic evaluation step may be sent to the statistical evaluation step. In the statistical evaluation step, data from multiple packages can be used to find, for example, deviations over time. As shown, the statistical evaluation step may be configured to interact directly with the seal image recording step. This advantage is that the settings used to pinpoint the identifying features of the seals that require additional data can be adapted. For example, an input regarding the frequency used to illuminate the samples can be provided from the statistical evaluation step. The more image data sets captured using this frequency, the more reliable the statistical analysis becomes. The communication between the seal image recording step and the statistical evaluation step may follow a pre-programmed pattern or be based on a machine learning / artificial intelligence approach (ML / AI). The output from the statistical evaluation step may update the machine settings of a filling machine including a lateral sealing device.

[0049] The concept shown in FIG. 3 can be implemented in different ways. As an example, FIG. 4 schematically shows an apparatus 400 for evaluating the lateral seals 200, 202 of a package 114 manufactured by the filling machine 100 shown in FIG. 1.

[0050] Package 114 may be placed on table 402 so as to obtain an image dataset depicting seals 200, 202. In FIG. 4, as an example, the second seal 202 is being analyzed. Table 402 may include an opening such that light L emitted from a light source 404, such as an LED, strikes seals 202, 202 without being blocked by table 402. Table 402 may include a movable member 406 so as to be able to move seals 202, 202 relative to light source 404 and camera 408. As shown, light source 404 may be disposed below table 402 such that light L strikes seal portions 200, 202 from below or strikes towards a second side S-2 of the seal portion. Camera 408 may be disposed above seal portions 200, 202, i.e., facing a first side S-1 of the seal portion, such that light L emitted from light source 404 passes through seal portions 200, 202 and is then captured by camera 408 and converted into an image dataset.

[0051] The image dataset may be transmitted from camera 408 to control unit 410. This may include a processor 412 and a memory 414 configured to process the image dataset such that adhered portions within the image dataset are identified using a first data subset associated with a first color channel and non-adhered portions within the image dataset are identified using a second data subset associated with a second color channel, as will be described in more detail below.

[0052] Communication between the camera 408 and the control unit 410 may be wireless as shown in the figure. Although not shown, in order to perform such communication, the camera 408 may be equipped with a data communication module, and similarly, the control unit 410 may also be equipped with such a module. Alternatively, although not shown, the camera 408 and the control unit 410 may be integral parts of one and the same unit. Further, although not shown, the control unit 410 may communicate with a backend server, execute a part of data processing on the control unit 410 installed on-site, and execute a part of data processing on the backend server, that is, a remote data processing device. Also, it is possible to make the control device 410 function as a gateway for transferring the image data set to the backend server and perform all data processing on the backend server.

[0053] Figure 5 shows an example of an image data set 500 depicting the seal 202. As shown in the figure, a reference object 502 such as a ruler may be provided so that the length identified in the image data set 500 can be converted into an absolute scale, for example, a distance in mm units. The reference object 502 may form a part of the movable member 406 shown in FIG. 4.

[0054] The image data set 500 illustrated in FIG. 5 may include a number of data subsets. For example, as shown in FIG. 6, the image data set 500 may include first, second, and third subsets 600, 602, 604. Each of these subsets may correspond to a color channel. For example, the three subsets may correspond to the color channels red, green, and blue. In other words, the image data set 500 may include subsets each related to an identification optical frequency range.

[0055] The advantage of analyzing different subsets individually is that different features can be reflected in different ways in different frequency ranges. Thus, as a result, by analyzing different subsets individually and then combining them, it becomes possible to identify more reliable features compared to analyzing the image dataset 500 alone.

[0056] The different frequency ranges of the first to third data subsets may be adjusted with respect to the frequency range of the light L emitted from the light source. By adapting the light L in this way, and also by utilizing the fact that the wavelength of the light is known when analyzing the first to third data subsets 600, 602, 604, features such as the adhered and non - adhered parts of the seal can be identified in a more reliable way. The frequency of the light L emitted from the light source may be set by changing the settings of the light source. Also, although not shown, it is possible to install a filter between the light source and the seal and allow only the light of the identification frequency to pass through.

[0057] As described above, it has been found that it is particularly advantageous to evaluate ultrasonic seals by analyzing subsets. One of the reasons is that the seal pattern formed by an ultrasonic sealing device is different from the seal pattern formed by an induction heating sealing device, and the contact between the ultrasonic sealing device and the packaging material may be less than the contact between the induction heating sealing device and the packaging material, that is, the pressure applied from the ultrasonic sealing device to the packaging material is small.

[0058] After extracting features from subsets 600, 602, and 604, these can be overlaid on the image dataset 500 to form an image dataset 700 having the features. As shown, the edges between the adhered portions 702a, 702b and the non - adhered portion 704 may be emphasized by the overlaid lines. In the illustrated example, the non - adhered portion 704 is a portion where a cut can be provided to separate two packages from each other. To provide relevant information to the operator, the overlaid features are not limited to emphasizing particularly relevant lines or regions, and may include information added in the form of text or numbers. For example, the first width of the non - adhered portion may be marked and an indicator associated therewith, and here, the character string "Z1" may be provided. Similarly, the second width of the non - adhered portion may be marked and the character string "Z2" may be provided.

[0059] Figure 8 is a flowchart showing a method 800 for evaluating seals 200, 202. In a first step 802, an image dataset 500 can be captured. In a second step 804, the adhered portions 702a, 702b can be identified. In a third step 806, the non - adhered portion 704 can be identified.

[0060] Optionally, in a fourth step 808, a seal portion that may correspond to a top fin, a bottom fin, and / or a vertical seal can be identified using the first and second subsets.

[0061] Optionally, in a fifth step 810, the seal portion can be illuminated using a light source 404 such that light L emitted from the light source passes through the seal portions 200, 202 of the package 114 and enters the camera 408.

[0062] As an option, in the sixth step 812, a reference object 502 can be provided next to the seals 200, 202, and the reference object has a reference length and / or a reference width. In the seventh step 814, the reference object depicted in the image data can be identified. In the eighth step 816, the absolute length and / or the absolute width of the adhering portions 702a, 702b and / or the non-adhering portion 704 can be determined using the reference length and / or the reference width.

[0063] As an option, in the ninth step 818, before capturing the image data set, a substance such as ink may be applied onto the seal. By applying the substance, the light L emitted from the light source is affected in different ways in different color channels, and as a result, for example, the adhering portion and / or the non-adhering portion may become faster and / or more reliable.

[0064] As an option, in the tenth step 820, the affected portion of the substance can be identified using the third data subset 604. By linking the affected portion of the substance to one of the data subsets, it can be achieved by selecting the substance related to the third color channel related to the third data subset, and the identification of the seal portion can be made more reliable.

[0065] As an option, before the first step 802 of capturing the image data set, in the eleventh step 822, the food held in the package can be taken out, and then, in the twelfth step 824, the package can be unfolded so that the seal is exposed, and then, in the thirteenth step 826, the package can be dissolved so that the carton layer of the package material is removed.

[0066] Even if the different steps are mentioned in a different order, the steps may be executed not only in the order described above but also in other orders. Further, for reasons of time efficiency, some steps may be executed in parallel.

[0067] From the above description, various embodiments of the present invention have been described and illustrated. However, the present invention is not limited thereto, and it may be embodied in other ways within the scope of the subject matter defined in the following claims.

Claims

1. A method (800) for evaluating a seal portion (200, 202) of a carton package (114) that holds food (FP), wherein the seal portion (200, 202) is generated by using ultrasonic sealing technology, and the method comprises: Capturing (802) an image dataset (500) depicting the seal portion (200, 202), wherein the image dataset (500) comprises a first data subset (600) associated with a first color channel and a second data subset (602) associated with a second color channel; Identifying (804) a fixed portion (702a, 702b) by using the first data subset (600); Identifying a non-fixed portion (704) by using the second data subset (602) (806); The method (800).

2. Further comprising identifying the seal portion by using the first data subset (600) and the second data subset (602) (808), The seal portion corresponds to a top fin, a bottom fin or a longitudinal seal portion of the carton package (114), Both the fixed portion (702a, 702b) and the non-fixed portion (704) are included within the seal portion, The method according to claim 1.

3. The image dataset (500) is captured by using a camera (408) disposed on a first side (S-1) of the seal portion (200, 202) of the package (114), Illuminating (810) the seal portion (200, 202) by using a light source (404) provided on a second side (S-2) of the seal portion (200, 202), The second side of the seal portion (200, 202) is opposite to the first side (S-1) of the seal portion, and light (L) emitted from the light source (404) irradiates the camera (408) through the seal portion (200, 202) of the package (114), The method according to claim 1.

4. The light source (404) is one or more light emitting diodes (LEDs) that emit light having at least one identifiable frequency, and the first frequency range of the first color channel and / or the second frequency range of the second color channel are set based on at least one identifiable frequency of the one or more LEDs, The method according to claim 3.

5. The first frequency range of the first color channel and / or the second frequency range of the second color channel are set based on the characteristics of the packaging material (PM). The method according to claim 3.

6. The image dataset (500) is acquired in multiple versions, each of which has different light settings, the light settings include the frequency of the light emitted from the light source and / or the amplitude of the light emitted from the light source. The method according to claim 3.

7. The packaging material (PM) includes a non-metallic light barrier layer, the non-metallic light barrier layer being a cellulose-based light barrier layer, and the cellulose-based light barrier layer is adapted to block light having frequencies that cause the growth of microorganisms and / or the loss of vitamins in the food (FP) held in the carton package (114). The method according to claim 1.

8. A reference object (502) is provided (812) next to the sealing portion (200, 202) of the carton package (114), the reference object (502) having a reference length and / or reference width. The reference object (502) depicted in the image data (500) is identified (814), Using the aforementioned reference length and / or reference width, the absolute length and / or absolute width of the fixed portion (702a, 702b) and / or the non-fixed portion (704) is determined (816). The method according to claim 1.

9. Before capturing the image dataset (500), an ink or other substance is applied to the sealing portion (200, 202) (818). The method according to claim 1.

10. The aforementioned image dataset (500) further includes a third data subset (604) related to a third color channel, The aforementioned method, The method includes using the third data subset (604) to identify the affected area of ​​the substance (820), The step of identifying the seal portion (808) is performed using the first, second, and third data subsets (600, 602, 604). The method according to claim 9.

11. The sealing portion includes an overlapping subsection where the lateral sealing portion and the vertical sealing portion overlap. The method according to claim 1.

12. Before the step (802) of capturing the aforementioned image dataset (500), Take out the food (FP) held in the package (114) (822), Unfold the package (114) so that the seal parts (200, 202) are exposed (824), Dissolve the package (114) so that the carton layer of the packaging material is removed (826), The method according to claim 1, comprising the above.

13. An apparatus (400) for evaluating the seal parts (200, 202) of a carton package (114) arranged to hold food (FP), wherein the seal parts (200, 202) are generated by using ultrasonic sealing technology, and the apparatus (400) comprises: A camera (408) for capturing an image dataset (500) depicting the seal (200, 202), and the image dataset (500) comprises a first data subset (600) associated with a first color channel and a second data subset (602) associated with a second color channel, A control unit (410) comprising a processor (412) and a memory (414), which uses the first data subset (600) to identify the adhered parts (702a, 702b) in the image dataset (500) and uses the second data subset (602) to identify the non - adhered part (704) in the image dataset (500), Comprising. Apparatus (400).

14. Further comprising a light source (404) for illuminating the seal (200, 202), The camera (408) is arranged on a first side (S - 1) of the seal part (200, 202) of the package (114), the light source (404) is provided on a second side (S - 2) of the seal part (200, 202), and the second side of the seal part (200, 202) is opposite to the first side of the seal part (200, 202), Light (L) emitted from the light source (404) passes through the seal part (200, 202) of the package (114) and irradiates the camera (408). The apparatus according to claim 13.

15. The light source (404) is one or more light - emitting diodes (LEDs) that emit light having at least one identifiable frequency, and the first frequency range of the first color channel and / or the second frequency range of the second color channel are set based on at least one identifiable frequency of the one or more LEDs. The apparatus according to claim 14.

Images