Food packaging quality evaluation method and packaging system

Abstract

A method (200) for assessing the quality of a food package (102) in a packaging system (100), the method (200) comprising: capturing (202) a first image data set (128) depicting the food package (102) at a first station (112) in the packaging system, the first station (112) comprising a first image capture device (114); capturing (204) a second image data set (130) depicting the food package (102) at a second station (118) in the packaging system (100), the second station (118) comprising a second image capture device (120); the second image capture device (120) is positioned at a different location from the first station (112) of the package transport path (P), the second image capture device (120) is positioned differently compared to the first image capture device (114), and the food package (102) is represented differently in the second image data (130) compared to the first image data (128), determining (206) characteristics (136) by processing the first image data (128) and the second image data (130), comparing (208) the characteristics with reference data (140), and sending (210) a positive quality indication (146) in case of a match or sending (212) a negative quality indication (148) in case of a mismatch.

Description

[Technical field]

[0001] The present invention relates generally to food package quality assessment, and more particularly to a food package quality assessment method, a packaging system, and a non-transitory computer readable storage medium having stored thereon program code portions for implementing the method. [Background technology]

[0002] It is now commonly known to have a quality assessment station as part of a food packaging system. Such a station may, for example, comprise a camera and a control unit configured to capture image data depicting the caps of the food packages, compare the image data with reference image data, and notify an operator that the cap application process needs to be examined if a deviation between the captured image data and the reference image data is identified. Being able to detect quality issues can reduce the risk of producing packages that do not meet food safety standards, thereby reducing the risk of distributing unsafe food to consumers.

[0003] In addition to ensuring food safety, quality evaluation stations are also closely related to reducing food waste. Being able to quickly identify quality problems in food packaging allows food packaging systems to be stopped quickly, which results in fewer food packages being produced with quality issues. This is beneficial in that fewer food packages are discarded, which in turn results in less food waste. Thus, installing quality evaluation stations for food packaging systems also has the advantage of improving environmental impacts such as CO2 emissions.

[0004] Several techniques are known for evaluating image data in the packaging industry, for example to evaluate cap applications. The development of cameras has been remarkable in the last decades, and the image quality of today's cameras offers possibilities that did not exist 10-20 years ago. The development of image data processing equipment has likewise progressed well in the last decades. As a result, the quality assessment of food packaging based on image data has improved in the last years.

[0005] Although image-based quality assessment stations can efficiently and reliably detect deviations, quality assessment needs to be further improved to provide better food safety and less food waste. Summary of the Invention [Problem to be solved by the invention]

[0006] The object of the present invention is to at least partially overcome one or more of the above identified limitations of the prior art, in particular to provide a method for evaluating the quality of food packages, which allows to further reduce the risk of producing food packages that do not meet pre-set quality standards.

[0007] According to a first aspect, there is provided a method for assessing the quality of food packages in a packaging system, the method comprising: capturing first image data depicting the food package at a first station in the packaging system, the first station including a first image capture device; capturing second image data depicting the food package at a second station of the packaging system, the second station comprising a second image capture device, the second station being located at a different position on the package transport path P of the packaging system than the first station, the second image capture device being differently positioned compared to the first image capture device, whereby the food package is represented differently in the second image data compared to the first image data; Processing the first and second image data to determine the characteristics; Compare the features to a reference data If there is a match, we will send you a positive quality indication. If there is no match, a negative quality indication will be sent. This includes:

[0008] The features may be interpreted as single or composite features of the package being evaluated. For example, the features may include the distance between the cap placed on the top of the package and the shortest distance to the edge of the top of the package. Such features are referred to herein as single features. If the features include a relationship, i.e., a quotient, between the shortest distance between the cap and the edge of the top and the longest distance between the cap and the edge of the top, such features are referred to herein as composite features. In other words, the features may be interpreted as package details that may vary from package to package, such that they can be used to identify the package or to determine whether the package deviates from other packages being manufactured, thereby detecting whether the package does not meet a pre-set quality standard. The features may include pre-set quality standards, thereby making the comparison a direct evaluation of the pre-set quality standard, while the features may include features that are indirectly related to the pre-set quality standards, thereby making the comparison an indirect evaluation of the pre-set quality standard.

[0009] The reference data may in format correspond to features extracted from the first and second image data sets, but instead of being linked to the package to be evaluated, the reference data is linked to a package that is deemed to meet a pre-defined quality criterion. Thus, in case of a match, i.e. if the features match the reference data, the package is deemed to meet the pre-defined quality criterion and a positive quality indication may be output. Conversely, in case of a mismatch, a negative quality indication may be output.

[0010] To improve the reliability of the evaluation, the reference data may be based on packages that are deemed to meet and not meet pre-defined quality criteria. Having information on both packages that are deemed to meet and not meet pre-defined quality criteria may provide more reliable reference data.

[0011] The advantage of having two separate stations is that combining the image data captured at the two stations may provide more complete information regarding the condition of the packages.

[0012] The different locations along the package transport path P should be broadly construed. For example, the first and second stations may be positioned to capture image data that simultaneously depicts a top and bottom of a package. In other words, the different locations need not be at different times.

[0013] The method further comprises: Identifying an identifying mark in the first image data set; Identifying an identifying mark in the second image data set; linking the first and second image datasets to the food package; It may include the following.

[0014] The advantage of having an identification mark is that it allows the two sets of image data to be securely linked, and in addition to linking the two sets of image data, the identification mark may also be used to identify the package at a later time.

[0015] The identification mark may be a mark printed on the food package.

[0016] The second lighting arrangement may include a polarizing filter such that the second light emitted by the second lighting arrangement is converted to polarized light and reflected by the food package before reaching the second image capture device, and the first lighting arrangement is configured without a polarizing filter such that the first light emitted by the first lighting arrangement is reflected by the food package and received as unpolarized light by the first image capture device.

[0017] Using non-polarized light at one of the stations has the advantage that dents can be identified using the reflection of non-polarized light, whereas using polarized light has the advantage that text and other printed information on the package can be more easily recognized. Thus, having a combination of the two, i.e. two stations using both options, allows different aspects to be combined and evaluated at different stations, improving the quality assessment.

[0018] A first set of image data produced by the unpolarized light received by the first imaging device may be used to detect dents in the food package.

[0019] A second set of image data produced by the polarized light received by the second imaging device may be used to identify printed text on the food package.

[0020] The first lighting arrangement may be configured to be physically different compared to the second lighting arrangement such that the first light is reflected differently by the food package compared to the second light.

[0021] The method further comprises: receiving packaging material data and / or recipe data, the packaging material data including information regarding the type of material used in the packaging of the food package, and the recipe data including information regarding the food to be filled into the food package; adjusting the characteristics of the first light and / or the second light based on the packaging material data and / or the recipe data; It may include the following.

[0022] An advantage of taking into account packaging material data and / or recipe data is that quality issues related to specific packaging materials or specific products can be addressed, thereby achieving a more reliable quality assessment of the packages.

[0023] The method further comprises: receiving embellishment data, the embellishment data including information regarding printing applied to the food package; adjusting characteristics of the first light and / or the second light based on the decoration data; It may include the following.

[0024] By having information about the decoration of the packaging material and adapting the lighting depending on the decoration, the generated image data allows for an improved quality assessment.

[0025] The step of determining the characteristic may further include determining a number of measurements associated with the characteristic, and the method may further include transmitting the measurements to the filling machine so that settings can be adjusted accordingly.

[0026] The method further comprises: receiving machine data linked to the food package, the machine data including information regarding machine components and / or machine settings of a packaging system used in producing the food package; and determining the characteristics by processing the first and second image data in combination with the machine data. It may include the following.

[0027] The advantage of taking into account machine data is that features related to components of a filling machine that are soon to be replaced can be specifically addressed. For example, if a filling machine's transverse sealing system is soon to be replaced, features related to the transverse seals can be specifically addressed. In doing so, the risk that a quality problem will be overlooked can be reduced. Furthermore, it also offers the possibility of extending the life of components, since features related to these will be monitored more closely if they are soon to be replaced, or in other words, close to the end of their expected life.

[0028] The first station is located upstream of the cap applicator and the second station is located downstream of the cap applicator, where the characteristic comprises a degree of rotation of a cap applied to the food package by the cap applicator relative to the food package, and the cap is rotationally asymmetric.

[0029] Alternatively or complementary, where the first station is located upstream of the cap applicator and the second station is located downstream of the cap applicator, the characteristics may include prelaminate hole locations determined using the first set of image data and cap locations determined using the second set of image data, thus ensuring that the caps are correctly positioned relative to the prelaminate holes.

[0030] The method further comprises: receiving a feature request from a server including a database that holds reference data; one or more requested features specified in a feature request are determined in a step of processing the first and second image data to determine features; Uploading one or more requested features to a server to fill gaps in the database; It may include the following.

[0031] An advantage of this method is that if some of the reference data has too few data points to make a reliable comparison, a feature request can be made to fill in those gaps. In other words, in addition to capturing data that can be used for a particular package, the first and second stations can capture data that can be used to provide reliable reference data.

[0032] According to a second aspect, a packaging system comprises: a first station including a first image capture device configured to capture first image data depicting a food package; a second station including a second image capture device configured to capture second image data depicting the food package; the second image capture device is configured differently from the first image capture device such that the food package is depicted differently in the second image data compared to the first image data; and a controller, the controller comprising: a characteristic determination module configured to process the first and second image data to determine the characteristic; a comparison module, the comparison module comparing the features to reference data; If there is a match, a positive quality indication will be sent. If there is no match, a negative quality indication will be sent.

[0033] The same features and advantages as given in relation to the first embodiment also apply to this second embodiment.

[0034] According to a third aspect, there is provided a non-transitory computer readable storage medium having stored thereon program code portions for performing the method according to the first aspect when executed on a device having processing capability.

[0035] Further objects, features, aspects and advantages of the present invention will become apparent from the following detailed description and drawings. [Means for solving the problem]

[0036] Embodiments of the present invention will now be described, by way of example only, with reference to the accompanying drawings, in which: FIG. [Brief description of the drawings]

[0037] [Figure 1] 1 illustrates a schematic representation of a packaging system. [Diagram 2] FIG. 2 is a top view of the package according to the first embodiment. [Diagram 3] FIG. 11 is a top view of a package according to a second embodiment. [Figure 4] FIG. 11 is a bottom view of a package according to a third embodiment. [Diagram 5] 1 is a flow chart illustrating a method for evaluating the quality of food packages in a packaging system. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0038] Referring to Figure 1, an example packaging system 100 is generally illustrated. In this particular example, food packages 102 are transported on a conveyor band 104 along a package transport path P from a filling machine 106, such as a roll-fed carton packer, to a cap applicator 108 where caps 110 are applied to the food packages 102. As shown, a first station 112 having a first image capture device 114 and a first lighting arrangement 116 is provided downstream of the filling machine 106 and upstream of the cap applicator 108. A second station 118 is provided downstream of the cap applicator 108, with a second image capture device 120 and a second lighting arrangement 122. An advantage of having two different stations for quality assessment of the food packages 102 is that image data can be captured in different ways. For example, the first lighting device 112 and the second lighting arrangement 122 are set to have different angles between the incident light (first light 162 at the first station 112 and second light 164 at the second station 118) and the top surface of the food package 102. The different angles, illustrated as α and β in Fig. 1, may result in different reflections at the first station 112 and the second station 118. This allows the information from the two stations to be combined to provide a more reliable quality assessment.

[0039] As indicated above, the filling machine 106 may be a roll-fed filling machine, such as a Tetra Brik® filling machine sold by Tetra Pak®, also referred to as a packaging machine. To form and fill the packages 102, the filling machine 106 may receive packaging material 124, illustrated here as a reel having a carton-based packaging material, and food product 126. The packaging material 124 may be printed with decoration before being fed to the filling machine 106, or decoration may be printed as part of a step performed by the filling machine 106. That is, part of the decoration is printed before the packaging material 124 is fed to the filling machine 106, and part of the decoration is printed at the filling machine 106. In addition to printing decoration, product-related information such as a best-before date may be printed in or downstream of the filling machine by a separate printer.

[0040] The first station 112 and the second station 118 may be configured differently in many different ways. As alluded to above, the angle of incident light may be different at the first and second stations 112, 188, which will result in different reflections on the package 102. Other possibilities include different spectral characteristics of the incident light, different numbers of light sources used, different arrangements of the light sources, different types of shielding against ambient light, different intensities of the incident light, etc.

[0041] After the first image data set 128 is captured by the first imaging device 114 and the second image data set 130 is captured by the second imaging device 120, they can be transferred to a controller 132. The transfer of image data can be wired or wireless. The controller 132 can be located near the first and second stations 112, 118 or can be located remotely.

[0042] In the controller 132, a feature determination module 134 may be used to determine features 136 in the first and second image data sets 128, 130. Such features may be, for example, how the cap 110 is positioned relative to how the cutouts in the packaging material 12, referred to as prelamination holes, are positioned. Such features may include the reflection of the packaging material 124 from the incident light. Dents and other types of deformations in the package 102 may affect the reflection, and by imaging the reflections from different stations 112, 118, combining the first and second image data 128, 130 from the two may more reliably identify the dents and other deformations compared to using only a single station.

[0043] Once the characteristics 136 are determined, they are transferred to a comparison module 138 and compared to reference data 140. The reference data 140 may be based on a number of packages 102 known to meet quality standards, such as packages without dents or other deformations, or packages in which the cap 110 is aligned with the cutout in the packaging material. The reference data 140 may be obtained from a server 142 that includes a database 144 in which the reference data 140 is stored. By storing the reference data 140 on the server 142, data may be collected from a large number of packaging systems 100, resulting in statistically reliable reference data. Although not shown, operator input may also be received by the server 142 so that the reference data 140 can be generated.

[0044] The comparison module 138 may output a positive quality indication 146 if the features 136, also called feature data, match the reference data 140, i.e. if the package 102 is deemed visually similar to a package known to meet quality standards. Conversely, if there is no match, a negative quality indication 148 may be output. The comparison module 138, like the feature determination module 134 and the server 142, may be based on artificial intelligence. That is, the feature determination module 134 and the comparison module 138 may decide, in whole or in part, which features to evaluate and how to compare the features with the reference data. When artificial intelligence is used, neural networks may be used.

[0045] If there are gaps in the database 144, i.e., if additional information is needed to be able to provide the necessary reference data 140 to perform a reliable assessment of the package 102, for example if there are too few data points related to a certain type of opening device, a feature request 150 may be forwarded from the server 142 to the first station 112 and / or the second station 118. In this way, the two stations 112, 118 may be configured such that the first and second image data sets 128, 130 provide information related to the features specified in the feature request 150. In other words, in addition to collecting data that can be used to assess a particular package, the first and second stations 112, 118 may be used to continually improve the reference data 140 used in the comparison module 138.

[0046] Because different packaging materials 124 may have different specifications, for example the packaging materials 124 may comprise different materials, different thicknesses, and different numbers of layers, information regarding the packaging material 124 used may be provided to the packaging system 100 in the form of packaging material data 152. Based on the packaging material data 152, the settings of the first and second stations 112, 118 may be adapted so that the first and second sets of image data 128, 130 include information that allows a more reliable quality assessment to be made. The packaging material data 152 may be provided to the controller 132 and / or the server 142.

[0047] Similarly, the embellishment data 154 may be provided to the first and second stations 112, 114 as well as to the controller 132 and the server 142. Using the embellishment data 154 providing information regarding the printed pattern provided on the packaging material 124, it is possible to adapt the first and / or second lighting arrangements 116, 122 such that their spectral characteristics match the particular embellishment of the package 102 being evaluated.

[0048] Additionally or alternatively, the machine data 156 may be used as input to the first and second stations 112, 118 and to the controller 132 and / or server 142. The machine data 156 may include information regarding machine components and / or machine settings of the packaging system 100 used when producing the packages 102. This may be advantageous in that certain features may be selected to be able to address effects caused by, for example, machine settings and / or machine components. For example, if the lateral sealing system of the filling machine 106 is about to be replaced, there may be an increased risk that the lateral seals of the packages 102 will not meet quality standards, and as such features related to lateral sealing may be prioritized.

[0049] Recipe data 157, i.e. information related to the food product 126 and the processing of the food product 126, may also be taken into account such that the features are selected in such a way that an improved quality assessment is possible. For example, if the food product 126 contains seeds, there may be an increased risk that the food product 126, especially the seeds, may get caught in the transverse seal and the transverse seal may be adversely affected. The recipe data 157 may be input to the first and second stations 112, 118 as well as to the controller 132 and the server 142.

[0050] As mentioned above, reflections on the package 102 may be used to detect dents or other deformations of the package 102. It is understood that for this purpose unpolarized light 168 may be used. However, polarized light 166 may be used to enable text or other printed information provided on the packaging material 124 to provide image data depicted in such a way that the text can be recognized. As shown, unpolarized light 168 may be used at the first station 112 and polarized light 166 may be used at the second station 118. To form the polarized light 166, a polarizing filter 158 may be placed in front of the second illumination arrangement 122. A polarizing filter 159 may be placed in front of the second imaging device 120 to provide that the polarized light 166 emitted from the second illumination arrangement 122 and passing through the polarizing filter 158 placed in front of the second illumination arrangement reaches the second imaging device 120. As an alternative to the polarizing filter 159 being placed in front of the second imaging device 120, the image sensor of the second imaging device 120 may be configured such that only the polarized light 166 is registered. Such an image sensor is called a polarization image sensor.

[0051] The identification mark 160 can be used to identify the package 102 and thereby provide for linking the first image data set 128 and the second image data set 130 related to the same package 102. The identification mark 160 can be a printed mark such as a QR code or Data Matrix, or a mark integrated into the packaging material 124. For example, magnetic particles can be incorporated into the packaging material 124 to provide a unique identification for the package 102. It is also possible to track the package 102 throughout the packaging system 100, and tracking the package 102 can link the first and second image data sets 128, 130 even if the package 102 does not have an identification mark 160. Furthermore, a combination of the two approaches, namely providing the package 102 with an identification mark 160 and tracking the package 102, can provide a more reliable process for linking the two sets of image data.

[0052] The positive and negative quality indications 146, 148 may include measurements 170, i.e., numerical data related to the characteristics. For example, the distance between the lateral seal and the cap 110 may be determined and form part of the measurements 170. This information may be fed back to the filling machine 106 and / or the cap applicator 108 (not shown) to provide for settings to be adjusted to meet set quality criteria. For example, if the distance between the lateral seal and the cap 110 is greater than a set value, but still within an acceptable range, the cap applicator may be adjusted to provide that the upcoming cap 110 is applied closer to the lateral seal.

[0053] FIG. 2 illustrates in further detail the top 200 of the package 102 by way of example. In this particular example, the cap 110 has not yet been applied, and this example may constitute an example of the package 102 as it passes through the first station 112. The illustrated package 102 is a carton-based package with a longitudinal seal 202. The longitudinal seal 202 may be formed by placing one side of the packaging material 124 on top of the other side and then heat sealing them together. A transverse seal 204 may be provided as well. As shown, a top fin with a transverse seal 204 may be folded onto the top surface 200 of the package 102 such that a substantially flat top surface of the package 102 is achieved. Printed information 206 may be provided on the package 102 to identify the package 102 and / or to provide information related to the manufacturing process of the package 102 and / or the food product 126.

[0054] In the illustrated example, the cap 110 is applied at a later stage. The carton layer cutout 208 can be provided not only in the packaging material 124 but also in the top surface 200 of the package 102 formed by the packaging material 124. As the name suggests, the carton layer cutout 208 can be a part of the top 200 of the packaging material 124 where the carton layer has been removed. In this way, the cutting element of the applied cap 110 can penetrate the packaging material 124. The effect of removing the carton layer is that the aluminum layer 210 becomes visible. The reflection 212 by the first light 162 from the first lighting arrangement 116 can occur not only in the aluminum layer 210 but also in other parts of the package 102 where the carton layer has not been removed. However, due to the different properties of aluminum and carton, the reflection 212 is different in the area where the carton layer has been removed, i.e. the carton layer cutout 208, and in other areas where the carton layer is present. By having two stations, it is possible to address different needs in an improved way. Fold lines 214 , also referred to as lines of weakness, may be present on the apex 200 and may be used to determine the features 136 and / or measurements 170 .

[0055] Figure 3 shows another example of the top 200 of the package 102. Unlike the example shown in Figure 2, the example illustrated in Figure 3 shows the top surface 200 with the cap 110 applied. The aluminum layer 210 is not visible in this example because the cap 110 is placed over the carton layer cutout 208. When using a packaging system 100 arranged according to the example shown in Figure 1, the example shown in Figure 3 is an example of how the package 102 may be arranged as it passes through the second station 118, i.e., downstream of the cap applicator 108.

[0056] The cap 110 can have a cap base 216 that attaches to the top 200 of the package 102 and a lid 217 that screws onto a spout that forms part of the cap base 216. To ensure that the cap 110 is properly attached to the top 200, the angle between the lateral seal 204 and the side of the cap base 110 can be determined and used as one of the features 136. The angle between the lateral seal 204 and the side of the cap base 110 can be used as one of the measurements 170.

[0057] Figure 4 illustrates a third example of package 102. Unlike the examples illustrated in Figures 2 and 3, the example illustrated in Figure 4 is the bottom 218 of package 102. In this example, as in the examples illustrated in Figures 2 and 3, package 102 is a carton package, and more specifically, package 102 is formed from a web of packaging material 124 that is comprised of multiple layers, at least one of which is a carton layer.

[0058] In the top 200 shown in Figures 2 and 3, the flaps are folded over the side panels of the package 102 and are not visible in the illustrated example. However, in the bottom 218, the first and second bottom flaps 220, 222 may be folded inward and toward the lateral sealing provided in the bottom fin 224, thereby providing a more stable base for the package 102. The angle of the first and second bottom flaps 220, 222 as well as the distance between them may form part of the features 136 and the numerical data associated therewith may form part of the measurements 170. Printed reference data 226 may be provided in the bottom 218 and may form the basis of the features 136 and / or measurements 170 by itself or in combination with, for example, the first and second bottom flaps 220, 222.

[0059] 5 is a flow chart illustrating a method 200 for evaluating the quality of food packages 102 in a packaging system 100. The method 200 includes: A first set of image data 128 depicting the food package 102 is captured (202) at a first station 112 in the packaging system, where the first station 112 comprises a first image capture device 114; capturing (204) a second set of image data 130 depicting the food package at a second station 118 in the packaging system, the second station 118 including a second image capture device 120, the second station 118 being disposed at a different location from the first station 112 on the package transport path P of the packaging system 100, the second image capture device 120 being differently positioned compared to the first image capture device 114, such that the food package 102 is represented differently in the second image data 130 compared to the first image data 128; determining (206) features 136 by processing the first and second image data 128, 130; Comparing 208 the features 208 with reference data 140; If there is a match, a positive quality indication 146 is sent (210); If there is no match, a negative quality indication 148 is sent (212); It may include the following.

[0060] Optionally, the method further comprises: Identifying 214 an identifying mark 160 in the first image data set 128; Identifying 216 the identifying mark 160 in the second image data set 130; Linking (218) the first and second image data sets 128, 130 to the food package 102; It may include the following.

[0061] Optionally, the method further comprises: receiving (220) packaging material data 152 and / or recipe data 157, where the packaging material data 152 includes information regarding the types of materials used in the packaging material 124 of the food package 102, and the recipe data 157 includes information regarding the food 126 to be filled into the food package 102; adjusting (222) the characteristics of the first light 162 and / or the second light 164 based on the packaging material data 152 and / or the recipe data 157; It may include the following.

[0062] Optionally, the method further comprises: receiving (224) embellishment data 154, the embellishment data 154 including information regarding printing provided on the food package 102; adjusting (226) a characteristic of the first light 162 and / or the second light 164 based on the embellishment data 154; It may include the following.

[0063] Optionally, the method further comprises: receiving (228) machine data linked to the food package 102, the machine data 156 including information regarding machine components and / or machine settings of the packaging system 100 used in producing the food package 102; The step of determining (206) the characteristics 136 is performed by processing the first and second image data 128, 130 in combination with the machine data 156. It may include the following.

[0064] Optionally, the method further comprises: receiving (230) a feature request 150 from a server 142 that includes a database 144 that holds reference data 136; uploading (232) one or more requested features 136 to the server 142, thereby filling gaps in the database 144; It may include the following.

[0065] From the foregoing description, while various embodiments of the present invention have been described and illustrated, the present invention is not limited thereto and may be embodied in other ways within the scope of the subject matter defined in the following claims.

Claims

1. A method (200) for evaluating the quality of a food package (102) in a packaging system (100), the method (200) comprising: capturing (202) a first image data set (128) depicting the food package (102) at a first station (112) in the packaging system, the first station (112) comprising a first image capture device (114); capturing (204) a second image data set (130) depicting the food package (102) at a second station (118) within the packaging system (100), the second station (118) comprising a second image capture device (120), the second station (118) being located at a different position on a package transport path (P) of the packaging system (100) from the first station (112), the second image capture device (120) being differently positioned compared to the first image capture device (114), and the food package (102) being represented differently in the second image data (130) compared to the first image data (128); determining (206) characteristics (136) by processing the first image data (128) and the second image data (130); Comparing (208) the features with reference data (140); If there is a match, send (210) a positive quality indication (146); If there is no match, send (212) a negative quality indication (148); The method (200) includes:

2. Identifying (214) an identifying mark (160) in the first image data set (128); Identifying (216) the identifying mark (160) in the second image data set (130); linking (218) the first image data set (128) and the second image data set (130) to the food package 102; [0023] 2. The method of claim 1 (200).

3. The identification mark (160) is a printed mark provided on the food package (102).

3. The method of claim 2 (200).

4. a second lighting arrangement (122) comprising a polarizing filter (158), wherein the second light (164) emitted by the second lighting arrangement (122) is converted into polarized light (166) and reflected by the food package (102) before reaching the second image capture device (120); a first lighting arrangement (116) that does not include a polarizing filter, and a first light (162) emitted by the first lighting arrangement (116) is reflected by the food package (102) and received by the first image capture device (114) as unpolarized light (168); 2. The method of claim 1 (200).

5. a first set of image data (128) generated by the unpolarized light (168) received by the first imaging device (114) is used to detect dents in the food package (102); 5. The method of claim 4 (200).

6. a second set of image data (130) generated by the polarized light (166) received by the second imaging device (120) is used to check printed text on the food package (102); 5. The method of claim 4 (200).

7. the first lighting arrangement (116) is configured to be physically different than the second lighting arrangement (122) such that the first light (162) is reflected differently by the food package (102) compared to the second light (164); 5. The method of claim 4 (200).

8. receiving (220) packaging material data (152) and / or recipe data (157), the packaging material data (152) including information regarding types of ingredients used in the packaging material (124) of the food package (102) and the recipe data (157) including information regarding a food product (126) to be filled into the food package (102); adjusting (222) characteristics of the first light (162) and / or the second light (164) based on the packaging material data (152) and / or the recipe data (157).

2. The method of claim 1 (200).

9. receiving (224) embellishment data (154), the embellishment data (154) including information regarding printing provided on the food package (102); adjusting (226) characteristics of the first light (162) and / or the second light (164) based on the decoration data (154).

2. The method of claim 1 (200).

10. The step of determining (206) the characteristic (136) further includes determining (234) a plurality of measurements (170) associated with the characteristic (136); The method comprises: and sending said measurements (170) to a filling machine (106) so that settings can be adjusted.

2. The method of claim 1 (200).

11. receiving (228) machine data linked to the food package (102), the machine data (156) including information regarding machine components and / or machine settings of a packaging system (100) used in producing the food package (102); and wherein determining (206) the characteristics (136) is performed by processing the first image data (128) and the second image data (130) in combination with the machine data (156).

2. The method of claim 1 (200).

12. the first station (112) is disposed upstream of a cap applicator (108), and the second station (118) is disposed downstream of the cap applicator (108); The features (136) include a rotation angle of a cap (110) applied by the cap applicator (108) onto the food package (102) relative to the food package (102), the cap (110) being rotationally asymmetric.

2. The method of claim 1 (200).

13. receiving (230) a feature request (150) from a server (142) that includes a database (144) that holds the reference data (136); one or more requested features (136) specified in the feature request (150) are determined in a step (206) of determining the features (136) by processing the first image data (128) and the second image data (130); uploading (232) the one or more requested features (136) to the server (142) to fill gaps in the database (144); [0023] 2. The method of claim 1 (200).

14. A packaging system (100), comprising: a first station (112) including a first image capture device (114) for capturing a first set of image data (128) depicting the food package (102); a second station (118) including a second image capture device (120) configured to capture a second set of image data (130) depicting the food package (102), the second image capture device (120) configured differently from the first image capture device (114), the second image data (130) depicting the food package differently compared to the first image data (128); A controller (132); Equipped with The controller: a feature determination module (134) configured to process the first image data (128) and the second image data (130) to determine features (136); comparing said characteristics (136) with reference data; If there is a match, a positive quality indication (146) is sent; a comparison module (138) for sending a negative quality indication (148) if there is no match. A packaging system (100).

15. A non-transitory computer readable storage medium having stored thereon program code portions for performing the method of claim 1 when executed on a device having processing capabilities.

Images