A method for detecting defects in horizontal mold seals of glass containers.

By using a combination of light source and image capture camera on glass containers to analyze the contour deviation of the bottle mouth edge, the problem of difficulty in efficiently detecting defects in horizontal mold seals in existing technologies is solved, and high-speed and reliable defect detection and quality control are achieved.

CN116802482BActive Publication Date: 2026-11-10TIAMA SOCIETE ANONYME
View PDF 2 Cites 0 Cited by

Patent Information

Application Number
CN202180086201.6
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Priority Date
2020-12-22
Filing Date
2021-12-09
Publication Date
2026-11-10
Estimated Expiration
2041-12-09

AI Technical Summary

Technical Problem

Existing technologies struggle to reliably detect small defects in horizontal mold seals of glass containers at high speeds, and cannot effectively distinguish between different types of defects.

Method used

By placing a glass container between a light source and an image-capturing camera, with the camera's observation optical axis perpendicular to the container's vertical axis, multiple rotating incremental images are captured. The deviation between the bottle mouth edge contour and the reference contour is analyzed to detect defects in the horizontal mold seal.

Benefits of technology

It enables efficient and reliable detection of various defects in horizontal mold seals on high-speed production lines, including ejection defects, flange defects, and overhang/overfit defects, ensuring that container quality meets standards.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN116802482B_ABST
    Figure CN116802482B_ABST
Patent Text Reader

Abstract

This invention relates to a method for detecting defects in a horizontal mold seal (JH) on the neck (B) of a container (R), the method comprising the steps of: - arranging the container (R) between a light source and a camera; - ensuring that the container (R) itself rotates according to one full rotation; - acquiring images by the camera at each rotation increment of the container, such that the number of images per full rotation is greater than 36; - analyzing the captured images for each container such that: * the contour of the bottle neck edge is detected in each image; * the contour of the bottle neck edge is compared with a reference contour of the bottle neck edge to detect differences between the contours; * and when at least one image of the container has a deviation, defects in the horizontal mold seal (JH) for the container are detected.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] This invention relates to the technical field of objects to be inspected, hollow articles, or generally transparent or translucent containers (such as bottles, jars, or flasks made of glass).

[0002] The present invention relates more specifically to the field of inspecting such glass containers, with the aim of detecting the presence of defects in the horizontal mold seal at the finish of such containers. Background Technology

[0003] Typically, containers have a bottom from which a vertical wall rises and terminates at a part called the neck. Depending on the closure system provided, the neck is of different types, and as... Figure 1 As shown, at the top, it includes an annular bottle neck surface S and at the bottom, it includes a counter-finish bottle neck CB. Figure 1 This describes a spiral-shaped bottle neck, also including a mesh located on a columnar vertical portion. Glass containers are known to be manufactured using a forming machine called an IS machine, which consists of different independent forming sections supplied with malleable glass droplets. Each forming section is equipped with at least one blanking chamber containing a blank mold, and an equal number of finishing chambers, each receiving a blow molding die in which the container acquires its final shape at high temperature.

[0004] Conventionally, the bottle neck is formed in a preform mold. When the preform is conveyed to the blow molding die, the bottle neck is already formed and the preform is held by the bottle neck. For this purpose, the bottle neck is formed by a bottle neck mold, which consists of two half-molds forming the vertical walls of the bottle neck and a mold component called a ring that forms the bottle neck surface or sealing surface. The preform mold also includes two preform body half-molds to form the walls of the preform body. In the press-blow molding process, a punch pushes the glass onto the preform body half-molds. In the blow-blow molding process, the punch is shorter and enters the bottle neck mold; however, compressed air pushes the glass onto the preform body half-molds. Therefore, the outer vertical walls of the bottle neck are formed by the two bottle neck half-molds, the upper surface of the bottle neck is formed by the ring, and the interior of the bottle neck is formed by the punch. Thus, as... Figure 1 As shown, each container neck R includes a vertical mold seal JV and a horizontal mold seal JH on either side of its vertical wall. The vertical mold seal JV corresponds to the interface between the two neck halves, and the horizontal mold seal JH corresponds to the interface between the neck mold component, referred to as the ring, and the two neck halves. These neck mold seals are more or less marked on the container or are visible on the container, depending on the adjustment of the mold's parts that deteriorate with use.

[0005] It should be considered that the horizontal mold seal JH is located slightly below the bottle neck surface S of the container. Furthermore, there is a need to detect defects in the horizontal seal of the mold affecting the container, in order to eliminate containers with defects that may affect their aesthetic properties or, more seriously, pose a real danger to the user.

[0006] In existing technology, AGR International has proposed the possibility of detecting defects in horizontal mold seals, referred to as knock-outs and flanges, using its DSG400 laboratory monitoring machine, which precisely measures the dimensional characteristics of containers. This machine specifically includes a light source located on one side of the container and a camera located on the other side. The container is driven to rotate around its vertical axis while images are being captured. This machine does not appear to be designed to detect all defects in horizontal mold seals. Furthermore, the machine has a limited monitoring rate, which does not allow for monitoring containers at a known production rate on a container production line.

[0007] Document WO2013 / 128538 describes a method for detecting bottle mouth defects in horizontal mold seals on glass containers. The method aims to ensure the container itself is rotated and that light is projected vertically onto the bottle mouth. A linear camera collects the light reflected from the bottle mouth. The method involves analyzing the reflected light to infer defects based on variations in the reflected light's profile. This method is only capable of detecting flanges whose shape causes them to reflect light along the camera's direction, and it does not allow for the quantification of defect size or the detection of different types of defects in the horizontal seals of the container's mold, while possessing a very high level of reliability for detecting small defects.

[0008] Therefore, there is a need for a technology that can detect various defects in the horizontal seals of container molds at high speeds, with a very high level of reliability for detecting small defects (typically less than 1 mm), and avoids considering defective containers, while the horizontal seals of such container molds actually meet the required quality standards. Summary of the Invention

[0009] Therefore, the object of the present invention is to meet this need by providing a method for detecting defects in horizontal mold seals at high speed at the mouth of glass containers. This method has a very high level of reliability in detecting small defects in horizontal mold seals while avoiding consideration of defective containers, and the horizontal seals of the molds for such containers actually meet the required quality standards.

[0010] To achieve this objective, the present invention aims to provide a method for detecting defects in horizontal mold seals at the mouths of glass containers, each having a vertical axis. To detect defects in each container, the method includes the following steps: - The container is positioned between a light source and an image capture camera, the observation optical axis of which is substantially perpendicular to an axis parallel to the vertical axis of the container, and the field of view of which includes at least the left edge or the right edge of the bottle opening, including at least a portion of the horizontal mold seal. - Ensure that the container itself rotates along the vertical axis at least once; - At each rotation increment of the container, images are acquired by the camera, such that the number of images per rotation is greater than 36; - Analyze the captured image for each container so that: *The bottle mouth inspection area, which includes at least a portion of the horizontal mold seal, above the height of the bottle mouth is defined in the image; *Detect the contour of the bottle opening edge in the image inspection area; * Compare the contour of the bottle mouth edge in the image with the reference contour of the bottle mouth edge to detect the deviation between these contours of the bottle mouth edge and the reference contour of the bottle mouth edge. *And when at least one image of the container has a deviation, a defect in the horizontal mold seal used for the container is detected.

[0011] According to an advantageous variation of the embodiment, the camera is arranged such that the field of view includes at least a portion of the horizontal mold seal and a reference whose height position is known relative to the horizontal mold seal, and in order to define the bottle neck inspection area in the image, a reference whose height position is known relative to the horizontal mold seal is found, and the bottle neck inspection area is positioned relative to the reference such that the bottle neck inspection area extends beyond the height including the horizontal mold seal.

[0012] For example, find all or part of the bottle neck or reverse bottle neck surface in the image as a reference.

[0013] Typically, there are image capture cameras with a horizontal width of field of view ranging from 5 mm to 130 mm and a height of field of view ranging from 3 mm to 20 mm.

[0014] For example, there are image capture cameras with a resolution greater than 25 pixels / mm.

[0015] According to a preferred feature of the embodiment, the image capturing camera is positioned such that the observation optical axis of the image capturing camera is substantially tangent to the left or right edge of the bottle opening.

[0016] The method according to the present invention can be used with a camera equipped with a telecentric lens.

[0017] An advantageous feature of the embodiment is that a light source has a vertical illumination size and a horizontal illumination size, which are included in the range of 100% to 200%, preferably 120%, of the size of the camera's field of view multiplied by the distance between the light source and the camera lens and divided by the distance between the axis of the container and the camera lens.

[0018] According to another embodiment, there is a telecentric light source whose illumination field size is greater than or equal to the field of view size of the camera.

[0019] According to a variation of the embodiment, deviations are detected by comparing the contour of the bottle mouth edge of the image with the reference contour of the bottle mouth edge using at least one contour obtained on at least one reference container that is considered to conform.

[0020] According to another variation of the embodiment, deviations are detected by comparing the profile of the bottle mouth edge of the image with the reference profile of the bottle mouth edge by using an average profile calculated over several rotational increments of at least one reference container that is considered to conform.

[0021] According to another variation of the embodiment, by using the contour of the bottle mouth edge with a low-pass filter applied as a reference contour, the contour of the bottle mouth edge in the image is compared with the reference contour of the bottle mouth edge to detect deviation.

[0022] According to another variation of the embodiment, a deviation is detected by comparing the contour of the bottle mouth edge in the image with a reference contour of the bottle mouth edge by applying a high-pass filter to the contour of the bottle mouth edge.

[0023] Advantageously, for each image, the contour of the bottle opening edge and the reference contour of the bottle opening edge are compared with a threshold by comparing at least one area, amplitude and / or slope measurement with a threshold, for at least some heights in the inspection area, and a deviation is detected when at least one of these measurements exceeds the threshold.

[0024] Typically, the method includes detecting at least one horizontal mold seal defect: the at least one horizontal mold seal defect is taken from the following defects: ejection, flange and overfit / overfit.

[0025] According to a preferred embodiment, at least one detection criterion is defined for each of the following defects: demolding, flange, and overrun / overfit, and the method includes distinguishing the defects detected in these three defects by using at least one of these detection criteria.

[0026] According to one feature of the embodiment, for at least one image, the shape of the deviation detected at several heights in the inspection area is selected as the detection criterion.

[0027] According to another feature of the embodiment, the range of viewing angles of the deviation corresponding to the number of consecutive images in which the deviation is detected is selected as the detection criterion.

[0028] To distinguish overhang / overfit defects relative to mold release and flange defects, the shape of the deviation detected at several heights in the inspection area is analyzed.

[0029] To distinguish mold defects relative to flange defects, the observation angle range of the deviation is analyzed by considering that a mold defect is detected when the observation angle range of the deviation is below a fixed maximum limit.

[0030] In order to distinguish mold defects relative to flange defects, the observation angle range of the deviation is analyzed by considering that the flange defect is detected when the observation angle range of the deviation is higher than a fixed minimum limit.

[0031] Typically, for each container with a defective horizontal mold seal, a signal is sent to eject the container from the production line.

[0032] According to an advantageous feature of the embodiment, the container itself rotates along the vertical axis for a maximum duration of 200 ms based on at least one full rotation.

[0033] Another object of the present invention is to provide an inspection apparatus for implementing the method according to the invention, for detecting defects in a horizontal mold seal on the mouth of a glass container. Attached Figure Description

[0034] [ Figure 1 ] Figure 1 It is an enlarged view showing the bottle neck of the glass container, the horizontal mold seal, and each of the vertical mold seals.

[0035] [ Figure 2 ] Figure 2 This is a schematic front view of an apparatus for implementing the method according to the invention, used to detect defects in the horizontal mold seal on the neck of a glass container.

[0036] [ Figures 3A-3B ] Figures 3A-3B These are, respectively, a side view of the bottle neck edge and a top view of the bottle neck of a glass container with a defective horizontal mold seal.

[0037] [ Figures 4A-4B ] Figures 4A-4B These are, respectively, a side view of the bottle mouth edge and a top view of the bottle mouth of a glass container with a flanged horizontal mold seal defect.

[0038] [ Figures 5A-5B ] Figures 5A-5B These are, respectively, a side view of the bottle neck edge and a top view of the bottle neck of a glass container with a horizontal mold seal defect of overfit / overmatch type.

[0039] [ Figure 6 ] Figure 6 This is an example image of the right edge of the neck of a glass container with a defect in a horizontal mold seal.

[0040] [ Figure 7 ] Figure 7 This is an illustration of extracting the outline of the edge of the bottle mouth of a container with defects in a horizontal mold seal from an image.

[0041] [ Figure 8 ] Figure 8 This is a diagram illustrating how to extract the baseline profile of the bottle opening edge from a container in an image.

[0042] [ Figure 9 ] Figure 9 This is a diagram illustrating a comparison between the profile of the mouth edge of a container with a flange or a defective horizontal mold seal and the reference profile of the mouth edge of the container.

[0043] [ Figure 10 ] Figure 10 This is a diagram illustrating a comparison between the profile of the bottle neck edge of a container with a drooping / overfitting horizontal mold seal defect and the reference profile of the bottle neck edge of the container.

[0044] [ Figure 11 ] Figure 11 This is an illustration of extracting the reference profile of the bottle neck edge in the form of a straight line for the threaded bottle neck from an image.

[0045] [ Figure 12 ] Figure 12 This is an example image of the right edge of the neck of a container with a molding defect, showing the container's outline and reference outline. Detailed Implementation

[0046] like Figure 1 and Figure 2 More specifically, the object of the present invention relates to an apparatus 1 for implementing a method according to the invention for detecting defects in a horizontal mold seal JH on a glass container R. Conventionally, the container R has a bottom F from which a vertical columnar wall V rises along a vertical axis Z and terminates in a portion referred to as the neck component B. In the case of a bottle-shaped container R, the vertical columnar wall V extends from the bottom F, and the portion forming the bottle body is connected to the neck ring C via a shoulder E.

[0047] like Figure 1 As shown, each container R specifically includes a horizontal mold seal JH, which corresponds to the interface between the mold component, referred to as the ring, and the two half-molds forming the vertical columnar wall V of the container. By definition, the horizontal mold seal JH is located slightly below the bottle neck surface S, corresponding to the planar surface used to seal the container. The horizontal mold seal JH extends over the entire circumference of the bottle neck.

[0048] According to the present invention, the method includes detecting defects in the horizontal mold seal JH on each container, the defects being taken from the following defects: demolding JHK, flange JHF, and overfit / overfit JHO.

[0049] like Figure 3A and Figure 3B More specifically, the ejector JHK defect corresponds to a glass flange at one of the T-shaped intersections of the horizontal mold seal JH and the vertical mold seal JV, the vertical mold seal located between the mold component, referred to as the ring, and the two half-molds forming the vertical wall. This defect corresponds to a glass flange resulting from a blunt ridge at the interface of the three molds or from poor alignment between them. The ejector JHK takes the form of a pointed or needle-shaped (typically triangular cross-section) glass protrusion with a small thickness along the vertical axis Z and a small horizontal extent, typically less than 1 mm. Therefore, it has a small circumferential angular range θ around the vertical axis Z. Typically, the ejector JHK has a small circumferential angular range θ of less than a few degrees.

[0050] like Figure 4A and Figure 4B More specifically, the flange JHF defect corresponds to excess glass caused by poor alignment between the mold component, known as the ring, and the two half-molds forming the vertical wall. This flange JHF takes the form of a thin sheet of glass along the vertical axis Z. Furthermore, this flange JHF has a significant angular range θ around the vertical axis Z. Typically, the flange JHF has a significant angular range θ greater than or equal to 10°, but quite easily greater than 30°. It can be much wider, up to 180°. More rarely, the flange can be present across the entire circumference.

[0051] like Figure 5A and Figure 5B As shown more specifically, the overhang / overfit JHO defect corresponds to excess glass caused by the displacement of a mold component known as a ring and the two half-molds forming the vertical wall. This overhang / overfit JHO takes the form of a very thick glass protrusion along the vertical axis Z, with a significant angular range θ around the vertical axis Z. Typically, the ring offset may occur at more than 180°, but the resulting step is usually significant over an angular range θ greater than or equal to 90°.

[0052] As will be described in detail below, the object of the present invention is to detect a container with defects in the horizontal mold seal JH from any one or two of the defects of the ejection JHK, flange JHF, and overhang / overfit JHO. According to an advantageous variation of the embodiments, the object of the present invention is to identify defects in the horizontal mold seal JH from the defects of the ejection JHK, flange JHF, and overhang / overfit JHO.

[0053] Therefore, the apparatus 1 for implementing the method for detecting defects in a horizontal mold seal JH includes a fixed light source 3 and a fixed image capture camera 4. The fixed light source 3 is arranged on one side of the container R, and the fixed image capture camera 4 is arranged on the other side of the container R. The camera 4, including a lens 4a, is adapted to capture an image of at least a portion of the horizontal seal JH of the mold of the container R. The camera 4 is connected to an analysis and processing unit 5, which is configured to analyze the captured image and detect defects in the horizontal seal JH of the mold of the container R.

[0054] The image of container R is captured as container R rotates at least one revolution along its vertical axis Z, allowing the entire horizontal mold seal JH to move in front of the camera. Each container R is supported by a rotation system 6. For example, the rotation system 6 includes a bottom F for container R and a sliding or placement plane 7 for all types of drive systems 8 known per se. The drive system 8 guides each container R to remain between the light source 3 and the camera 4 for the time required to perform at least one revolution, during which time the image is captured by the camera 4, as will be explained later in the specification.

[0055] For example, the rotation system 6 includes wheels or rotators 8 as a drive system, which drive the container to rotate via friction, while the container is fixedly supported on at least two free casters or rollers 9. The rollers, for example, are formed as part of a transport star wheel 10 for transporting the container R on a circular path, such that they are sequentially supported and slide on a placement plane 7 in front of the inspection device 1. Advantageously, the transport star wheel 10 forms part of an inspection machine that inspects more than 150 containers per minute, including one or more stations for inspecting containers running at the exit of the production line. In other words, the inspection device 1 of the present invention can be implemented outside of inspection stations conventionally placed for inspecting containers on a production line without reducing the rate of container operation at these stations, which integrate the rotation of the containers for inspection.

[0056] According to the features of the invention, the online inspection machine inspects 50 to 500 containers per minute, typically 150 to 450 containers per minute. For each container, the method includes a transport step comprising bringing the container R to and arranging it between the light source 3 and the image capture camera 4, and ensuring that the container R itself rotates along the vertical axis Z according to at least one full rotation, during which an image is acquired. It should be noted that, using the transport star wheel 10, previously inspected containers are brought to the inspection station simultaneously with the new containers to be inspected. Therefore, the online inspection machine is equipped with a high-speed processing device. For a rate of 150 containers per minute, the duration of the rotation and inspection steps is at most 200 ms. According to a preferred feature of the invention, the complete rotation of the container lasts at most 100 ms. It is also often chosen that the container rotates 1.5 full rotations.

[0057] In other words, the method according to the invention includes arranging a container R between a light source 3 and an image capturing camera 4, the observation optical axis of which is substantially perpendicular to an axis parallel to the vertical axis Z of the container, and its field of view includes at least the left or right edge of the bottle opening, which includes at least a portion of a horizontal mold seal, ensuring that the container R itself rotates along the vertical axis Z according to at least one rotational revolution in less than 200 ms, and acquiring an image at each rotational increment of the container by the camera 4, such that the number of images per rotational revolution is greater than 36.

[0058] According to one feature of the invention, the guiding camera 4 captures images at each rotation increment of the container, thereby resulting in more than 36 images per rotation. In other words, the purpose of the method according to the invention is to acquire at least one image for every 10° rotation of the container R. For example, the number of images of the container R over 360° includes between 36 and 96. The rotation increment of the container between each captured image represents the angular sector traversed by the bottle neck, which ranges, for example, from 10° to less than 1°. Of course, a fast camera with an integration time of less than 500 μs and a readout time of 0.5 ms can acquire 400 images per rotation within 200 ms. The increased acquisition frequency leads to higher accuracy, especially in the estimation of the radial length of the mold. This results in a higher system cost due to the price of the fast camera 4 and the computational power required by the analysis and processing unit 5.

[0059] According to the invention, the light source 3 and the camera 4 are adapted such that the camera can acquire image I during rotation of the container about its vertical axis Z, in which at least a portion of the horizontal seal JH of the mold of the container R appears. Figure 2 As shown, by convention, the vertical axis Z of the container is considered to be parallel to the vertical direction z of the orthogonal reference system x, y, z. The vertical direction z is perpendicular to the horizontal direction y passing through the light source 3 and the camera 4, while the lateral direction x is perpendicular to both the horizontal direction y and the vertical direction z.

[0060] Container R is positioned between light source 3 and camera 4, whose observation optical axis Y extends parallel to the lateral direction y, that is, along a direction substantially perpendicular to the vertical direction z. Camera 4 has a field of view extending laterally along a lateral direction x orthogonal to the vertical direction z and the observation optical axis Y. Therefore, the field of view of camera 4 extends in a plane defined by the vertical direction z and the lateral direction x. According to a preferred example, camera 4 is positioned such that its optical axis is substantially tangent to the left or right edge of the bottle opening.

[0061] According to the vertical direction z, the field of view of camera 4 includes at least the horizontal mold seal JH. The field of view of camera 4 includes at least the left edge or the right edge of the bottle mouth B. According to a preferred variant of the embodiment, for good resolution, the field of view of camera 4 includes only the left edge or the right edge of the bottle mouth B. Of course, the method according to the invention can be implemented using a camera whose field of view is not limited to the right or left edge. This preferred variant is particularly suitable for high speeds because it is known that in-line inspection machines rotate the container at least one revolution, so unilateral observation is sufficient to ensure that defects in the horizontal seal JH of the mold of the bottle mouth are observed. Furthermore, unilateral observation allows for optimal use of the camera's field of view and its optical axis to be positioned tangent to the observed edge of the bottle mouth.

[0062] For each rotation increment, the camera 4 (such as a matrix camera) transmits the horizontal projection of the bottle opening to the analysis and processing unit 5, in which at least one edge of the bottle opening (referred to as the left edge or right edge) is highlighted, taking into account their position in each image captured by the camera. Figure 6 Therefore, the qualification of left or right is given from the observation point of each shot, because strictly speaking, the rotating container does not include a left or right side. Figure 6 In the example shown, image I corresponds to the right edge of container R.

[0063] Therefore, in each image captured by the camera, at least the left edge or the right edge of the bottle neck B appears, allowing at least a portion of the horizontal mold seal JH and at least the outline P of the bottle neck edge to be seen. In other words, as Figure 6 As shown, each image I must include at least a portion of the shape or contour P of the bottle neck edge at an elevation or height taken along a direction parallel to the vertical direction z, which includes at least a portion of the horizontal mold seal JH. Therefore, an inspection area ZI corresponding to a horizontal bar in each image is defined, the horizontal bar having a defined height to include a portion of the contour P of the bottle neck edge, which includes a portion of the horizontal mold seal JH.

[0064] According to an advantageous feature of the embodiment, the camera 4 is arranged such that its field of view includes at least a portion of the horizontal mold seal JH and a reference whose height position relative to the horizontal mold seal JH is known. Therefore, after finding said reference in the image, the bottle neck inspection area ZI can be positioned relative to this reference, such that the bottle neck inspection area extends beyond the height that undoubtedly includes the horizontal mold seal.

[0065] Of course, while taking into account the positioning deviation of container R relative to camera 4 and the manufacturing tolerances of container R, an additional margin is allowed when performing this image capture to visualize the reference and a portion of the horizontal mold seal JH. This margin is, for example, at least + / - 2 mm when the height of the container can vary by + / - 2 mm.

[0066] It must be understood that the reference presented by container R consists of salient elements such as contours, angles, and shapes that can be located in the image. This salient element is also visible in all images captured during the rotation of the item, and its vertical position in each image can be definitively determined. Its vertical distance Dv to the horizontal mold seal JH is known in each image, and therefore preferably constant. The position of this reference in the image allows for the assignment of a positional reference system to each image, which can be used for image analysis.

[0067] Typically, for example when the container comprises one, the entire or partial bottle neck S or the reverse bottle neck CB surface can be used as a reference. With the bottle neck surface S as the reference, camera 4 is adjusted such that the height of its field of view includes the bottle neck surface S plus a margin to account for container positioning deviations or manufacturing tolerances. Similarly, camera 4 is adjusted such that the height of its field of view includes the horizontal mold seal JH plus a margin.

[0068] The horizontal width of the field of view of camera 4 is adapted to include the profile P of the bottle neck edge with a margin and a reference, i.e., a portion of the bottle neck surface in the considered example. According to a preferred variation, the width of the field is chosen to include only the bottle neck edge plus the necessary margin. For example, camera 4 has a horizontal width preferably including a field of view between 5 mm and 80 mm and a height including a field of view between 3 mm and 20 mm. This allows for optimization of the camera's field, i.e., limiting unnecessary areas of the sensor. According to another variation, both the left and right edges of the bottle neck are included in the field of view, and the field of view width can reach 130 mm to observe a bottle neck with a diameter of 120 mm and a margin of 10 mm.

[0069] For example, camera 4 is a matrix camera combined with an optical lens, which can observe the inspection area at a resolution greater than 25 pixels / mm.

[0070] The light source 3 is made in any suitable manner to ensure that the backlighting of the container R is suitable for the image captured by the camera 4.

[0071] According to a preferred feature of the embodiment, the light source 3 has defined vertical and horizontal illumination dimensions. Typically, the vertical dimension is the dimension measured along the vertical direction z, while its horizontal dimension is parallel to the lateral direction x. In a non-telecentric optical system, the dimension DL of the light source 3 is of the type: DL = CH x DI / DC; where the CH field of view of the camera 4, the DI distance between the light source 3 and the camera lens 4a, and the DC distance between the inspection area and the camera lens are more specifically considered as the distance between the camera lens (e.g., the optical center of a non-telecentric lens) and the focal plane, which contains the vertical axis Z of the container and the point tangent to the edge of the bottle opening through which the light rays pass through the optical center. The vertical and horizontal illumination dimensions are the dimensions of the useful area of ​​the light source 3 observable by the camera 4.

[0072] According to one feature of the embodiment, the light source 3 has a vertical illumination size and a horizontal illumination size, which are included in 100% to 200% and preferably 120% of the size of the field of view of the camera 4, multiplied by the distance between the light source 3 and the lens 4a of the camera and divided by the distance between the vertical axis Z of the container and the lens 4a of the camera.

[0073] According to a variation of the embodiment, the light source 3 can be a telecentric light source, whose illumination field size is greater than or equal to the field of view size of the camera 4. Therefore, the light source 3 generates a beam of light that extends into the camera's field of view, the average of which is parallel to the camera's optical axis. In this case, it should be noted that the camera lens 4a is a telecentric lens.

[0074] Images captured by camera 4 for each container R are analyzed by analysis and processing unit 5, which is configured to detect defects in the horizontal seal JH of the mold of container R. The method for detecting defects in the horizontal seal JH includes: analyzing the captured images for each container, particularly analyzing, in each image, a bottle mouth inspection area ZI above the bottle mouth height, comprising at least a portion of the horizontal seal. This analysis method includes detecting the contour P of the bottle mouth edge in the image inspection area and comparing the contour P of the bottle mouth edge in the image with a reference contour Pf of the bottle mouth edge to detect deviations between these contours P and the reference contour Pf. When at least one image of container R shows a deviation, a defect in the horizontal seal of the container is detected. For each container with a horizontal seal defect JH, analysis and processing unit 5 sends a signal notifying the container of the defect quality, which controls an ejector to eject the container from the production line.

[0075] The detection of the contour P of the bottle neck edge in each captured image can be performed using any appropriate image processing method.

[0076] As described above, the image analysis is preferably limited to the bottle neck inspection area ZI, where a portion of the bottle neck profile P, including the horizontal mold seal, appears. Advantageously, the inspection area is positioned relative to a reference point of the container appearing in the image, and its height position relative to the horizontal mold seal is known. A variation of this embodiment limits the analysis of the bottle neck profile to only the portion necessary to characterize the horizontal mold seal, while eliminating other profile portions that might be confused with the horizontal mold seal, such as those including mesh.

[0077] According to this method, a reference with a known height position relative to the horizontal mold seal is found in each image, and the bottle mouth inspection area is positioned relative to the reference such that the bottle mouth inspection area extends beyond the height of the horizontal mold seal JH.

[0078] Figure 7 The detection of the contour P of the right edge of the bottle opening in image I is illustrated by way of example, in which the bottle opening surface S appears as a reference. The first step involves positioning in height, that is, positioning the position of the bottle opening surface S along the vertical direction z. For example, along a vertical line following the left edge of the image, the vertical gray level transition corresponding to the bottle opening surface S is detected. Subsequently, the inspection area ZI can be positioned according to this reference, so that the inspection of the image can take into account the horizontal mold seal JH.

[0079] Of course, the methods used to search for the benchmark can be different. For example, in the case where the benchmark is an inverted bottle neck, a search for the maximum two-dimensional correlation can be performed between the image and a benchmark window containing a learned model of the inverted bottle neck or the right or left visible edge of the inverted bottle neck.

[0080] The second step involves extracting the outer contour or profile of the bottle neck edge P, for example, from the bottle neck surface S to the lower limit of the inspection area ZI located below the bottle neck surface and exceeding the horizontal mold seal ZH. Figure 7 As shown, the bottle neck edge profile P includes a horizontal mold seal JH. Searching for a first profile point on the bottle neck edge P can include horizontally browsing the image at a given height relative to a reference until a horizontal grayscale transition is detected.

[0081] In each image, any image processing method can be used to extract the shape or contour of the bottle neck edge P. For example, it can be set in the inspection zone ZI to search for the outer edge of the bottle neck for each height along the vertical direction z by searching for a black / white transition Ti or searching for a white / black transition from the outside. The location of this transition Ti along the lateral direction x is determined relative to the vertical direction z. The search for this transition Ti is performed at each height and across the entire height of the inspection zone ZI. Figure 7 As shown on the right, for each image, the contour P of the bottle neck edge can be obtained in the form of a curve referenced in the plane z, x. For example, the contour P consists of all transformations Ti, for each transformation Ti, the coordinates x and z are known.

[0082] The next step involves comparing the profile of the bottle neck edge P with a reference profile Pf of the bottle neck edge obtained by various possible means. According to a preferred variation of the invention, such as... Figure 8 As shown, the reference profile Pf is extracted from the edge of the bottle mouth in a manner similar to... Figure 7 The extraction of the contour of the bottle neck edge P of the container shown is performed. For this purpose, for a reference container including a reference contour Pf of the bottle neck edge, comprising a portion of a defect-free horizontal reference mold seal, an image is captured such that the reference contour Pf of the bottle neck edge and the bottle neck surface serving as a reference are visible. This reference image is advantageously formed using an inspection device by positioning the reference container between the camera 4 and the light source 3 under the same conditions as the image used for the container to be inspected. Analysis of this image according to the above method allows the extraction of the reference contour Pf of the bottle neck edge, and the defect-free horizontal mold seal appears within this contour.

[0083] According to another variation of the embodiment, a reference profile Pf for the bottle neck edge is obtained by learning from several images of the container or from images of several containers considered to be conforming. For example, any type of mathematical operation can be performed on the image or the profile extracted from the image. Therefore, as the reference profile Pf, one or more profiles obtained on or on several reference containers considered to be conforming can be selected. Similarly, as the reference profile Pf, an average profile calculated on several rotational increments of reference containers considered to be conforming or several reference containers considered to be conforming can be selected. According to another variation of the embodiment, the reference profile Pf for the bottle neck edge can also be obtained from the manufacturing plane of the container or any geometric model.

[0084] according to Figure 11 Another variation of the invention shown, in example Figure 1In the case of the screw-shaped bottle neck shown, the bottle neck edge includes a columnar portion, and therefore the profile of the bottle neck edge includes the right portion PD. In this case, the reference profile Pf of the bottle neck edge can be a simple reference straight line D, which at most passes through the straight portion PD of the profile of the bottle neck edge P, and thus through the point Ti that passes through this portion of the bottle neck shape, or a straight line parallel to it.

[0085] according to Figure 12 In another variation of the illustrated embodiment, the invention aims to use the profile of the bottle neck edge to which a low-pass filter is applied as a reference profile Pf. In other words, the reference profile Pf corresponds to the profile of the bottle neck edge, to which filtering or smoothing is applied. Figure 12 In the example, the contour P (white) of the bottle neck edge corresponds to the original contour extracted from the image, while the baseline contour Pf (gray) corresponds to the moving average of contour P. Of course, different types of low-pass filters can be applied, such as average, Gaussian, median, or phase-shifted median filters.

[0086] According to another variation of the embodiment, it should be noted that a high-pass filter can be envisioned applied to the contour of the bottle neck edge to compare the contour of the bottle neck edge P in the image with a reference contour Pf of the bottle neck edge. For example, the application of a gradient-type high-pass filter can highlight deviations, that is, strong or rapid changes in the derivative of the bottle neck edge contour, reflecting the presence of local irregularities. The presence of such deviations corresponds to defects in the horizontal mold seal.

[0087] For each image, a comparison step between the contour of the bottle neck edge P in the image and the reference contour Pf of the bottle neck edge results in the detection of whether there is a deviation between these contours of the bottle neck edge and the reference contour of the bottle neck edge. When at least one image of the container has a deviation, a defect in the horizontal mold seal of the container is detected.

[0088] This comparison, which determines whether or not defects in the horizontal mold seal are observed, can be performed using various analytical methods.

[0089] For each image, for at least several heights within the inspection area ZI, and preferably for all heights, the contour of the bottle neck edge P and the reference contour of the bottle neck edge Pf are compared. This step aims to compare at least one area, amplitude, and / or slope measurement with a threshold, and detect a deviation when at least one of these measurements exceeds the threshold. Typically, the area measurement may correspond to the area between the contour of the bottle neck edge P and the reference contour Pf of the bottle neck edge. The amplitude measurement may correspond to the difference between the contour of the bottle neck edge P and the reference contour Pf of the bottle neck edge along the lateral direction x. The slope measurement may correspond to the measurement of the slope of the curve obtained by subtracting the contour of the bottle neck edge P and the reference contour Pf of the bottle neck edge. Another possible measurement is the difference between two amplitudes of the bottle neck edge contour P and the reference contour Pf of the bottle neck edge, obtained at equal or different heights.

[0090] According to a variation of the embodiment, in order to compare the contour P of the bottle neck edge and the reference contour Pf of the bottle neck edge, the actual position of the bottle neck edge is taken into account, including tilt in the image related to the tilt of the bottle neck relative to the axis Z due to handling hazards during monitoring or due to lack of verticality of the container. For example, Figure 11 The image shows this tilt of the bottle neck. It should be noted that in this simplified variant, when the reference profile of the bottle neck edge Pf is a straight line D that passes through the straight section PD of the bottle neck profile, the deviation can be determined by measuring the distance from a point on the bottle neck profile P to the reference straight line D.

[0091] exist Figure 9 In the example shown, curve CEc is the curve corresponding to the deviation between the profile of the bottle neck edge P and the reference profile of the bottle neck edge Pf. This curve evolves along the vertical axis z, that is, as a function of the height in the inspection area. Different examples of possible measurements of the deviation curve CEc can estimate the magnitude and / or shape of the deviation, and simultaneously estimate the presence of a defect, possibly its type and its severity. For example, the count of points of maximum amplitude of the curve can be taken, corresponding to the end of the protrusion furthest from the vertical axis Z of the container. This maximum amplitude corresponds to the height of the peak value of the deviation curve CEc, which gives a correct estimate of the radial length of the ejector or flange, constituting a very good physical criterion for the severity of the defect. This amplitude can be compared to a threshold, and if the measurement exceeds the threshold, a deviation (or defect) is detected.

[0092] The threshold can be adjustable to parameterize the severity of the inspection; for example, the operator can therefore decide to tolerate slightly marked demolding.

[0093] If the permissible measurement for deviation is the surface area obtained by subtracting the bottle neck edge profile from the bottle neck edge reference profile, then the measurement is, for example, the integral of the deviation curve CEc. If the measurement is a slope, then it can be the maximum slope of the deviation curve CEc, or the deviation between two consecutive slopes on either side of the peak or threshold intersection of the deviation curve CEc.

[0094] As can be clearly seen from the foregoing description, the method according to the invention includes detecting at least one horizontal mold seal JH defect from at least the following defects: demolding JHK, flange JHF, and overfit / overfit JHO.

[0095] According to an advantageous variation of the embodiments, the method according to the invention is capable of identifying defects detected among these three defects: demolding defect JHK, flange defect JHF, and overhang / overfit defect JHO. To this end, at least one inspection criterion is defined for each defect (i.e., demolding defect, flange defect, and overhang / overfit defect), and at least one of these inspection criteria is used to distinguish the defect. Therefore, it becomes possible to identify the type of defect, thereby potentially improving the inspection of containers.

[0096] According to an exemplary embodiment, for at least one image, the shape of the deviation detected at several heights within the inspection area is selected as the detection criterion. For example, the shape of the deviation can be described in different ways, such as by measurements of slope or amplitude obtained at several consecutive heights. This shape can be observed on the deviation curve CEc. Figure 9 In the middle, the deviation curve CEc has a peak value characterizing the mold or flange, but if the defect is similar to Figure 5A If there is a droop / overfit, the deviation curve CEc will have the following characteristics: Figure 10 The trapezoidal shape shown.

[0097] According to another exemplary embodiment, the observation angle range of the deviation corresponding to the number of consecutive images in which the deviation is detected is selected as the detection criterion; that is, the angular range of the deviation observed at the rotation angle of the container in the image where the deviation is detected is used as the detection criterion. For flange or overhang / overfit defects, the observation angle range of the deviation is very close to the circumferential angular range of the defect as seen from above, such as... Figure 3B , 4B As clearly seen in 5B. For mold ejection, the viewing angle range is greater than its circumferential angle range. If, for example, the rotation increment between two image acquisitions is 4°, and the mold ejection has a very small circumferential angle range of 2°, and if it is long enough, it can be seen in several consecutive images (two or three respectively) that are separated by one or more rotation increments in time, and therefore its viewing angle range is 8° or 12° respectively.

[0098] Of course, each of these standards can be implemented independently of each other, or in combination one after another. One of these standards can be applied before another, and vice versa.

[0099] According to one example of the embodiment, the shape of deviations detected at several heights in the inspection area is analyzed to distinguish overhang / overfit JHO defects from those of the demolding JHK and flange JHF defects. In fact, from Figure 5A It can be clearly seen that, compared to the ejection mechanism characterized by sharp protrusions (…), Figure 3A ) and flange ( Figure 4A The overhang / overfit JHO defect is particularly distinguished by its trapezoidal or serrated shape.

[0100] According to another example of the embodiment, considering that a demolding defect is detected when the observation angle range of the deviation is below, for example, a maximum limit fixed at 30°, the observation angle range of the deviation is analyzed to distinguish the demolding JHK defect from a flange JHF defect. In other words, a demolding JHK defect is detected when the number of consecutive images of detected deviation corresponds to an observation angle range below, for example, a maximum limit fixed at 30°.

[0101] According to another example of the embodiment, considering that a flange defect is detected when the observation angle range of the deviation is greater than, for example, a minimum limit fixed at 30°, the observation angle range of the deviation is analyzed to distinguish the mold JHK defect from the flange JHF defect. In other words, a flange JHF defect is detected when the number of consecutive images of detected deviation corresponds to an observation angle range greater than, for example, a minimum limit fixed at 30°.

[0102] According to the present invention, the threshold applied to the measured value to detect deviation, the discrimination criteria such as the discrimination criteria used in the comparison of contour shapes, and / or the discrimination criteria used as the minimum or maximum value of the observation angle range are stored in the analysis and processing unit 5, and preferably can be adjusted by the operator according to the container being inspected and the required quality.

Claims

1. A method for detecting defects in a horizontal mold seam (JH) at the mouth (B) of a glass container (R), each of the glass containers having a vertical axis (Z), the method comprising the steps of: - The container (R) is arranged between the light source (3) and the image capture camera (4), the observation optical axis of the image capture camera being substantially perpendicular to an axis parallel to the vertical axis (Z) of the container, and the field of view of the image capture camera including at least the left edge or the right edge of the bottle opening, including at least a portion of the horizontal mold seam; - Ensure that the container (R) itself rotates along the vertical axis (Z) at least one full rotation; - At each rotation increment of the container, an image is acquired by the image capture camera (4) such that the number of images per rotation is greater than 36; - Analyze the captured image for each container, such that: The bottle mouth inspection area (ZI) above the height of the bottle mouth (B), which includes at least a portion of the horizontal mold seam, is defined in the image; The contour (P) of the bottle mouth edge is detected in the bottle mouth inspection area (ZI) of the image; The contours (P) of the bottle mouth edge in the image are compared with the reference contours (Pf) of the bottle mouth edge to detect the deviations between these contours (P) of the bottle mouth edge and the reference contours (Pf) of the bottle mouth edge. And when at least one image of the container has a deviation, a defect in the horizontal mold seam (JH) of the container is detected.

2. The method according to claim 1, wherein, The image capture camera (4) is arranged such that the field of view includes at least a portion of the horizontal mold seam (JH) and a reference, the height position of which is known relative to the horizontal mold seam, and in order to define the bottle neck inspection area (ZI) in the image, the reference whose height position is known relative to the horizontal mold seam is found, and the bottle neck inspection area (ZI) is positioned relative to the reference such that the bottle neck inspection area (ZI) extends beyond the height including the horizontal mold seam.

3. The method according to claim 2, wherein, Find a reference point in the image where all or part of the bottle neck surface (S) or the reverse bottle neck (CB) surface is located.

4. The method according to claim 1, wherein, An image capture camera (4) exists with a horizontal width of field of view ranging from 5 mm to 130 mm and a height of field of view ranging from 3 mm to 20 mm.

5. The method according to claim 1, wherein, There exists an image capture camera with a resolution greater than 25 pixels / mm (4).

6. The method according to claim 1, wherein, The image capture camera (4) is positioned such that the observation optical axis (Y) of the image capture camera is substantially tangent to the left or right edge of the bottle opening (B).

7. The method according to claim 1, wherein, An image capture camera with a telecentric lens exists (4).

8. The method according to claim 1, wherein, There is a light source (3) having a vertical illumination size and a horizontal illumination size, which are included in 100% to 200% of the following values: the size of the field of view of the image capture camera (4) multiplied by the distance between the light source and the lens (4a) of the image capture camera and divided by the distance between the axis of the container and the lens of the image capture camera.

9. The method according to claim 1, wherein, A telecentric light source exists, and the size of the illumination field of the telecentric light source is greater than or equal to the size of the field of view of the image capturing camera.

10. The method according to claim 1, wherein, Deviations are detected by comparing the contour (P) of the bottle mouth edge of the image with the reference contour (Pf) of the bottle mouth edge using at least one contour obtained on at least one reference container that is considered to conform, as a reference contour (Pf).

11. The method according to claim 1, wherein, Deviations are detected by comparing the profile (P) of the bottle mouth edge of the image with the reference profile (Pf) of the bottle mouth edge using the average profile calculated over several rotation increments of at least one reference container that is considered to conform.

12. The method according to claim 1, wherein, By using the contour of the bottle mouth edge with a low-pass filter applied as a reference contour (Pf), the contour (P) of the bottle mouth edge in the image is compared with the reference contour (Pf) of the bottle mouth edge to detect deviations.

13. The method according to claim 1, wherein, By applying a high-pass filter to the contour of the bottle neck edge, the contour (P) of the bottle neck edge in the image is compared with the reference contour (Pf) of the bottle neck edge to detect deviations.

14. The method according to claim 1, wherein, For each image, the contour (P) of the bottle mouth edge and the reference contour (Pf) of the bottle mouth edge are compared with a threshold by comparing at least one area, amplitude and / or slope measurement value with a threshold, and a deviation is detected when at least one of these measurements exceeds the threshold.

15. The method according to claim 1, wherein, The method includes detecting at least one horizontal die joint (JH) defect, which is derived from the following defects: die ejection (JHK), flange (JHF), and overhang / overfit (JHO).

16. The method according to claim 1, wherein, For each of the following defects, namely JHK (Jmold Release), JHF (Jmold Flange), and JHO (Jmold Overfit), at least one inspection criterion is defined, and the method includes distinguishing the defects detected in the three defects by using at least one of these inspection criteria.

17. The method according to claim 16, wherein, For at least one image, the shape of the deviation detected at several heights in the bottle neck inspection area is selected as the detection standard.

18. The method according to claim 16, wherein, The range of observation angles corresponding to the number of consecutive images in which the deviation was detected was selected as the detection standard.

19. The method according to claim 17, wherein, In order to distinguish the overhang / overfit (JHO) defect from the ejection defect (JHK) and the flange defect (JHF), the shape of the deviation detected at several heights in the bottle neck inspection area is analyzed.

20. The method according to claim 18, wherein, By considering the detection of a demolding defect when the observation angle range of the deviation is below a fixed maximum limit, the observation angle range of the deviation is analyzed to identify the demolding (JHK) defect relative to the flange (JHF) defect.

21. The method according to claim 18, wherein, By considering the detection of a flange defect when the observation angle range of the deviation is higher than a fixed minimum limit, the observation angle range of the deviation is analyzed to identify the ejection defect (JHK) relative to the flange (JHF) defect.

22. The method according to claim 1, wherein, For each container with a horizontal mold seam defect, a signal is sent to eject the container from the production line.

23. The method according to claim 1, wherein, The container (R) itself rotates along the vertical axis (Z) for a maximum duration of 200 ms based on at least one full rotation.

24. An inspection apparatus for implementing the method according to any one of claims 1 to 23, for detecting defects in a horizontal mold seam (JH) on the mouth (B) of a glass container.

Citation Information

Patent Citations

  • Method and apparatus for inspecting flaw on lip section of mouth of glass-container

    WO2013128538A1

  • Optical inspection of container finish dimensional parameters

    CN1131740A