JP6203538B2 - Product inspection method for packaging machines - Google Patents

Description

  The present invention relates to a method for inspecting products in packaging machines in the field of tobacco. In particular, the method includes a machine for wrapping cigarette packages, wherein a group of cigarettes wrapped in a wrapping material is a product, or a group of cigarette packages, ie a carton of packages. In a packaging machine, it consists of a product inspection method in which a group of cigarette packages wrapped in a packaging material is a product.

  Automatic machines for packaging groups of cigarettes and packages of cigarettes are known. In the automatic machine, each group is advanced along the package forming path by conveyor means. In the wrapping path, each group is contained in one or more wrapping materials and / or wrapped with and / or receives the wrapping material.

  Such packaging typically comprises an inner packaging component and one or more outer packaging components. The inner packaging component is for stabilizing a group of cigarettes or a group of cigarette packages. In addition, the outer packaging component is for protection against moisture in a packaged and / or packaged document for the user. For example, in a cigarette package, the inner packaging component is a sheet of aluminum foil or metallized paper for wrapping a group of cigarettes, whereas a cellophane sheet is in both the package and the carton of the package. For outer packaging components. The outer wrapping component wraps the product wrapped in the inner wrapping component (a group of cigarettes or a group of cigarette packages), and a cigarette package and / or a soft carton of a package, or a cigarette package and / or Or it may be a sheet of printing paper and / or a stamped blank (which is also printed), each for obtaining a hard carton of the package.

  Typically, optical inspections are performed in the course of quality control procedures for groups of cigarettes and / or groups of cigarettes. This is to discriminate between undesirable tears and dirt that can occur in the packaging itself, incorrect packaging forms of the packaging, or (if applicable) misapplication of additional packaging elements.

  For example, the top and bottom surfaces of a group of cigarettes wrapped in a metal-coated paper or aluminum foil sheet are properly folded without extra folds and / or undesired wrinkles on the cellophane overwrap surface Be present and / or that the revenue stamp (for example, affixed on the larger back in a conventional rigid package) is in its proper position, ie its longitudinal axis overlaps the hinge line of the lid Optical inspection is performed to verify

  Such optical inspection is usually performed by an optical inspection unit on a product-by-product basis (i.e. every machine cycle). The optical inspection unit comprises an optical acquisition device and an illuminator associated with such device so that an image of the product can be acquired as the product is advanced by the conveyor of the packaging machine. It has become.

  The term “optical device for image acquisition” is capable of acquiring an image of an object and in particular a characteristic of interest from such an image (eg, geometric and / or shape characteristics of such an object). Means an optoelectronic device capable of processing the image to extract. The acquired images can be color or black-and-white images, from which information about color (tone, saturation, etc.) or gray level (density) and luminous intensity is extracted. Can do.

  The optical device comprises a main body on which an electronic sensor is arranged, and special light-receiving optical means fixed to the main body. The electronic sensor on the main body is, for example, an array of photoelectric elements of linear type or matrix type two-dimensional type, for example, CCD type or CMOS type, that is, an array. The optical receiving means is, for example, an objective mirror composed of one or more lenses, through which the sensor is suitable for receiving light scattered by the object to be acquired. The number of photoelectric elements of the electronic sensor corresponds to the number of dot-like elements constituting a raster image or bitmap display in the pixel, that is, the memory of the optical device. Through a one-time acquisition using a two-dimensional sensor, ie a matrix-type [n × m] photoelectric sensor, or n consecutive times using a linear sensor with m photoelectric elements. It will be appreciated that through acquisition (also referred to as scanning) an image having a resolution of [n × m] pixels can be obtained. For example, a linear sensor-type optical device is advantageously used to acquire an image of a product that is advanced by a conveyor, in which case the optical device is placed in a fixed reading position and the fixed device is fixed. It is known to perform a continuous scan as the product travels through the read position. Each successive scan corresponds to a very thin "line" or "column" in the overall image. In this case, a complete image of the product is reconstructed by arranging together all the stored (stored) columns after each scan.

  To acquire images, issue commands to start the illuminator, and (in some applications) to process the image for the purpose of extracting the characteristics of interest from the acquired image, and to send the results of the optical inspection to an external A control device for communicating to the control system is provided in the optical device.

  Such an optical device is known as a linear type or matrix type video camera or camera. If the cameras can process the acquired images to analyze the information of interest, they are also called “smart cameras”.

  The result of the optical inspection is transmitted to an external control system, such as a packaging machine control system, through an Ethernet communication network having a high data transmission rate or another type of network. Alternative communication means may be provided that are implemented through a set of input and output digital signals from the optical device, each associated with similar digital signals at the output and input of the control system of the packaging machine.

  Thus, the control system of the packaging machine directly excludes an inspected product that has been determined not to meet the required quality requirements as soon as the product reaches the exclusion station (or an external operation that performs the exclusion operation). Some of the defect information can be communicated to the device).

  The optical device is usually used in products having a parallelepiped shape in terms of the product advancement method, the type of conveyor, the positioning of the optical device relative to the surface of the product to be inspected, and the packaging material surrounding the product. It will be appreciated that any surface of interest, i.e., the side, front, back, top, or bottom, can be contained in a frame (frame for imaging range).

  When forming a product with a group of cigarettes to be wrapped with a sheet of aluminum foil or metal-coated paper, the conveyor is usually implemented as a packaging wheel that can rotate about a rotational axis. The wrapping wheel includes compartments or pockets for housing each group of cigarettes. Similar observations can be made for other types of packaging conveyors found in packaging machines.

  An indication of inadequate quality in the packaged group is given by the darker or brighter parts present in the image acquired by the optical device, and during the formation described above, Permanently associated. This darker or lighter part is due to tears in the aluminum foil mentioned above, oily dirt, unwanted pieces (sticky debris, leaf tobacco fibers for filling cigarettes, already partially glued A piece of torn paper that has been torn). The torn piece of paper may stick, for example, between the overlapping flaps of aluminum foil on the top surface of the product, in which case the packaged group should be excluded. The top surface is actually the surface that consumers see when they first open the product, which is subject to special quality requirements.

  The undesired results of the optical inspection mentioned above are products that are actually of flawless quality (packed cigarette groups or packaged cigarette packages) of poor quality and should therefore be excluded. That is, there is a possibility that it will be regarded as.

  In practice, the unwanted pieces are not permanently bonded to the product, but may be carried by the air and only pass in front of the optical device when acquiring the inspection image. . However, it is actually flawed because it cannot be identified without verification of recognizable shapes that can be associated with dirt and / or tears in the packaging sheet and / or air-borne pieces, respectively. It is impossible to process the acquired image to exclude only the product.

  In order to try to solve this problem, it is known to provide an optical inspection unit with a linear camera and air supply means in a packaging machine. The air flow of the air supply means is at least so as to be sandwiched between the optical means of the optical device and the surface of the product to be housed in the frame, so that there are no small pieces that may be found in the air. Of the surface of the product, the part that is housed in the frame is cleaned by a linear sensor. Such cleaning, which is effective with air flow, however, causes turbulence in the volume surrounding the flow. Within the turbulent flow, arbitrarily present small piece motion can occur in a chaotic manner. Because of these turbulences in the area adjacent to the air boundary, undesired pieces accumulate and can therefore be lined up in an unpredictable manner in parts of the product surface that are not affected by the aforementioned air flow. It means that there is sex.

  An inspection unit that has an optical device with a two-dimensional sensor instead of the air flow and distributes and supplies cleaning air with a dispersed jet is known. However, as is increasingly the case, cleaning the plane of the product surface that is housed in the frame by the two-dimensional sensor is a much more complex task. In fact, since a cleaning operation must be performed on a large area that is equal to the area affected by the image to be acquired, it is not possible to avoid turbulent movement in the air volume affected by the cleaning operation. Nearly possible. Therefore, the stability of the result is not guaranteed.

  The object of the present invention is to overcome the drawbacks of known inspection methods.

  A further object of the present invention is to provide an inspection method in a packaging machine capable of identifying small pieces of foreign material that are carried by air during product image acquisition and only pass in front of an optical device, and It also sorts out permanent stains that indicate defects in the product.

  All of the foregoing and other objectives are achieved by a method for inspecting a product in a packaging machine as defined by claim 1 and the other claims set forth below.

  The present invention will be better understood and practiced with reference to the accompanying drawings, which illustrate exemplary and non-limiting embodiments thereof.

An image of a section of a packaging wheel in a packaging machine containing a group of cigarettes wrapped in aluminum foil or metal-coated paper, inspected by an optical device placed facing the packaging wheel itself The figure which shows the image which is stored in the flame | frame within an area and is acquired at the preset acquisition time. FIG. 3 shows another image of the pocket of FIG. 1 as obtained after acquisition of the image of FIG. 1 is a schematic view of a control system for a packaging machine, comprising a control unit and a product inspection unit connected to each other, and a single optical device provided in such an inspection unit. FIG. 4 is a modified schematic diagram of the inspection unit of FIG. 3, comprising two optical devices both connected to the control unit. A modified model of the inspection unit of FIG. 3 comprising a projector for illuminating the surface of an object with a line of light and such an object being housed in a frame by an optical device arranged facing the object. Figure. The end of the group of cigarettes wrapped in the aluminum foil or metal-coated paper of FIG. 1, as housed in the frame by the light line projector of FIG. 5 and acquired at a preset acquisition time. The figure which shows a perspective image.

  FIG. 3 shows a control system 1 of a packaging machine (not shown). The control system 1 includes a control unit 2 (typically a packaging machine control unit) and an inspection unit 3 for a product 4. The control unit 2 is connected to the inspection unit 3 via the communication means 5. Through the communication means 5, the control unit 2 can receive data and / or commands and / or provide data and / or commands to the inspection unit 3, which is best described herein below. The

  The communication means 5 can be implemented by a high data transmission rate network or by special wiring (digital I / O interface) for input / output digital signals. These input and output digital signals are respectively connected to the output from the control unit 2 of the packaging machine and the similar digital signals at the input to the control unit 2, the digital signals at the input to the inspection unit 3 and the output from the inspection unit 3. A set of

  The inspection unit 3 comprises an optical device 6 for acquiring a digital image, in particular a single first optical device. The first optical device 6 is provided with a main body 7 and optical means, that is, an objective mirror 8. The objective mirror 8 of the optical device 6 has an optical axis A. The optical axis A is such that at least one surface 4a of the product is framed in the inspection area when the product 4 is advanced along the path P in a preset direction D by a conveyor (not shown) of the packaging machine. It is suitable for payment. In the context of the optical device 6, there may optionally be an illuminator (not shown) to illuminate the product 4 encased in the frame during image acquisition. An illuminator that generates scattered light is preferable in order to enable an accurate optical inspection that is not affected by reflection that may occur in the packaging material. That is, it is an illuminator that can be thought of as coming from all directions with respect to each point of the illuminated surface.

  In FIG. 3, the surface 4 a is the upper surface of the product 4. This surface 4a is placed in the frame by the optical device when such a surface 4a is located in the field of view (view) of the optical device 6 itself. Point out the following: That is, the field of view means an acquisition field (acquisition range) of the optical device 6, that is, a preset area in which an image of the product 4 can be acquired, which is within the focusing range, thereby A preset depth of field can be defined along the optical axis A of the objective mirror 8.

  The optical device 6 is provided with its own control device (not shown) in order to issue commands for image acquisition and optional lighting of the illuminator. The control device is accommodated in the main body 7 of the optical device 6 itself. The control device of the optical device 6 may also be implemented and configured to process the image in order to extract the feature of interest from the acquired image. Alternatively and / or in addition, the inspection unit 3 can be provided with its own control device (not shown) for processing the acquired image. In this case, the control device of the optical device 6 may be simpler with only image acquisition and illuminator control functions. For example, the control device of the inspection unit 3 can be implemented by its own electronic circuit configuration (for example, a DSP board for image processing). In the electronic circuit configuration, the optical device 6 is connected by the communication means 5 in the same manner as described above for the connection between the control unit 2 and the inspection unit 3. Thereby, the dimension of the optical device 6 can be reduced, the optical device 6 is close to the product 4, and an electronic circuit configuration of the inspection unit 3 (also referred to as a vision system). Can be placed at an optimal position sufficiently spaced apart.

  It is also possible that the control unit 2 is in direct contact with the optical device 6 (if the control device of the optical device 6 can process the acquired image) via the communication means 5 (when the acquired image is processed). If the above process is instead left to the control device of the inspection unit 3, it can be configured to communicate with the inspection unit 3. As will be pointed out more precisely in the present specification, the latter inspection unit 3 which is sandwiched between the control unit 2 and the optical device 6 and communicates with both 2 and 6 is represented by the present invention. In some embodiments of the method.

  As already described above, the optical device 6 may include a one-dimensional or two-dimensional CCD or CMOS photoelectric element sensor.

  In the case of a one-dimensional sensor, image acquisition occurs through continuous scanning performed one column at a time. For that acquisition, the optical device 6 is placed so that the linear sensor is transverse to the product 4 advanced along the path P by the conveyor. Each scan can be actuated by the control device of the optical device 6 or the inspection unit 3. Each scan is actuated at time intervals that are continuous with each other and have the same duration (which depends on the forward speed of the product 4) so as to obtain rows that are adjacent to each other at equal intervals in the entire image. . The overall acquired image constituted by a preset number of such columns is stored in the control device of the optical device 6 in order to construct an image of the surface 4a of the product 4.

  In the case of a two-dimensional sensor, instead, a complete image of the surface 4a is acquired by a single scan that is activated at a preset time.

  A communication protocol is prepared for the communication means 5 between the control unit 2 and the control device of the optical device 6 and / or the control device of the inspection unit 3. Through the communication protocol, when the product 4 is placed in a frame by the optical device 6 itself, the control unit 2 can issue a command (command) to the optical device 6 to acquire an image. The communication protocol further allows the optical device 6 to receive additional commands and configuration parameters.

  Through the communication means 5, the optical device 6 is further provided with a set of digital data (eg acquired images and / or data obtained as a result of an optical examination and stored in the control device of the optical device 6 or in an auxiliary storage device). A composite part of the information etc.) can be provided to the control unit 2 or a composite command by the control unit 2 can be received. For example, the complex command is associated with a command code, that is, a command to be executed, and a complex part of data information, for example, a plurality of basic data (for example, values of parameters settable in the optical device 6). With both data structures including

  As described above, the product 4 housed in the frame by the optical device 6 is a group of cigarettes or a group of cigarette packages wrapped in a packaging material. 1 and 2 show the upper surface 4a of the product 4, the product 4 is a group of cigarettes enclosed in an aluminum foil 9 or a sheet of metal-coated paper (in the case of special packages, More suitably, a cigarette complex).

  The product 4 is advanced along a circular package forming path (not shown) having a path P by a packaging wheel (not shown). The packaging wheel comprises a plurality of peripheral compartments 10, one of which appears in the acquired images of FIGS. These peripheral compartments 10 are arranged radially on the outer periphery of the packaging wheel so as to receive the product 4.

  The package forming path described above extends from the loading station to the take-out station. At that loading station, a group of cigarettes is received in one of the packaging wheel compartments 10 after catching each aluminum foil sheet 9. At the take-off station, the group of cigarettes fully packaged to obtain product 4 is transferred to the next packaging wheel.

  Each section 10 has a “U” -shaped receiving profile 11. A holding means comprising a plurality of holding covers 12 is associated with the packaging wheel. One of the retaining covers 12 is shown in FIGS. Each holding cover 12 faces the respective peripheral compartment 10 and holds the product 4 in the peripheral compartment 10 during rotation of the packaging wheel. Alternatively, a fixed holding cover 12 can be provided that faces a preset annular portion of the outer periphery of the packaging wheel.

  According to a known packaging scheme, the top surface 4a and the bottom surface (not shown) of the product 4 both have two end flaps 13 and 13 'folded (particularly having a trapezoidal shape) of aluminum foil 9. ing. Their end flaps 13 and 13 'are already folded during the advancement of the product 4 along the packaging path described above.

  The upper surface 4a is a surface corresponding to the upper end of the product 4 that is opened by the user to take out the contents of the product 4 (as described above, cigarettes and cigarette packages). Accordingly, the upper surface 4a is a subject of special attention in the field of quality control procedures, and any reference to the upper surface will be made in the present specification without losing generality anyway.

  After the end flaps 13 and 13 'have been folded and overlaid, the upper surface 4a can be placed in the frame by the optical device 6 of the inspection unit 3 facing the upper surface 4a in the inspection area, so that it can be removed. Prior to the station, the integrity of the packaged cigarettes and the accuracy of the packaging can be verified.

  In use, the optical device 6 of the inspection unit 3 receives an inspection start command from the control unit 2 through the communication means 5. The reception of this command was achieved by placing the product 4 placed in the peripheral compartment 10 and advanced by the wrapping wheel, in particular the surface 4a of the product 4 facing the optical device 6, in a frame in the field of view by the optical device 6. Sometimes done.

  In this specification, it will be pointed out below that reference will always be made to the acquisition of images through a two-dimensional sensor type digital camera, unless the use of a linear sensor type camera will be clearly indicated. Keep it. Instead, with respect to image processing, the method by which the image is acquired does not affect the processing itself.

  The inspection start command can be a complete inspection command that is left to the optical device 6.

  The control device of the optical device 6 issues a command to acquire an image and optionally turn on the illuminator when such a complete examination start command is received.

  The image thus obtained by the optical device 6 (shown in FIG. 1) is processed by the control device so that it is darker or brighter than the image of the defect-free product 4 that may be present in the image. Is identified in the image. This darker or lighter portion 14 may be associated with dirt and / or tearing of the packaging sheet and / or may be associated with undesired pieces carried by air.

  A black and white image can be considered as a numerical matrix I having a size (type) equal to the product of the number of rows and columns of the image. In that image, each element of matrix I represents one pixel (i.e. a point in the image) and has an associated numerical value (which is equal to luminous intensity). Such numerical values are given in gray scale (density), which depends on the type of electronic sensor used.

  A color image should instead be considered as a set n of matrices In having the same type, one for each base color component to be associated in the display under consideration (RGB, HSV etc.) Is. Each element of the matrix In has an associated numerical value equal to the intensity of the associated base color of the matrix.

  The darker or lighter portions 14 that can be observed by the inspection unit 3 usually have an irregular shape. And for such a goal, the inspection unit 3 is programmed according to known techniques in order to detect several parameters specific to each part 14, in particular its position and size in the acquired image. .

  The position is associated with the coordinates of the barycentric point in the area of the portion 14. Said size is associated with the radius of a circular area, which covers the geometrically most important part of the part 14 itself, centered on the center of gravity.

  Instead of this, the size is determined by the portion 14 with respect to the preset lower and higher light intensity thresholds and with respect to the average value of the corresponding portions in the image of the product 4 without defects. It is measured as the number of pixels having lower luminosity when darker and as the number of pixels having higher luminosity when portion 14 is brighter. This is because the average luminous intensity of a certain image varies depending on the part of the image.

  The pixel of the portion 14 must satisfy the following three conditions. Be adjacent to each other, have a sufficiently large number, i.e., a number that is greater than the acceptable average value of slight variations in luminous intensity, and substantially the same tone (for color images) or the same gray level (for black-and-white images) ). For the purposes this inspection method is to achieve, it would be preferable to use a monochromatic optical device 6 and thus process only black and white images.

  To identify a portion 14 so that the result of image processing is not affected by external illumination conditions (which determine the average brightness of the same image, which is brighter or darker) when the image is acquired Known techniques are as follows. That is, by taking into account the intensity gradient calculated according to known vector calculation rules on the whole image or each part thereof, rather than the intensity of the image after acquisition (i.e. gray level in the case of black and white images) Is to process the image.

  In practice, the light intensity gradient can be defined based on a single matrix I of the black and white image or on each of the n matrices In of the color image.

  The gradient has a characteristic that it does not depend on the absolute value but depends on the change. In this way, it is possible to highlight a set of pixels that represent an outline, i.e. a borderline, and that define areas with different luminosities (where a change in luminosity level occurs). With such a set of contour pixels, darker or brighter portions 14 are identified and defined. By applying a light intensity gradient calculation, it can be seen that the same processing algorithm can be used for both the search for the contour of the darker portion 14 and the contour of the brighter portion 14.

  The position and size of such a portion 14 in the acquired image is calculated and stored (stored) in the manner described above.

  When a darker or brighter portion 14 is present in the acquired image, the optical device 6 processes the second image acquired after the first image.

  The control device of the optical device 6 can issue a command to acquire an image and optionally turn on the illuminator only after the portion 14 has been identified in the first image, The first and second images can always be acquired with a time interval set by a preset delay. Then, only the first image is processed at first, and the second image is processed only when a brighter or darker portion 14 exists in the first image.

  Such a second image is shown in FIG. The second image acquired in this way is compared with the first image acquired previously. Then, when a darker or brighter portion 14 is specified in the second acquired image, such a darker or brighter portion 14 is the same in the first image and the second image. If present at a position (ie, the first and second positions substantially coincide with each other), it is determined that the product 4 is defective and should be eliminated. When the second position is within a circular tolerance zone centered on the first position and having a radius equal to the tolerance threshold radius, the first position is considered to be substantially coincident with the second position. It is pointed out that he is deceived.

  On the other hand, if the position where the darker or lighter portion 14 is present in the first acquired image is different from the position where this portion 14 is present in the second acquired image, the product 4 is defective. It is not judged as a thing. This is because this darker or lighter portion 14 is associated with undesirable pieces that are carried by air.

  In the example of FIGS. 1 and 2, it can be seen that the darker portions 14 are present in both acquired images but at different positions. The product 4 contained in the frame does not have to be regarded as defective. The dark part is obviously because it is an unwanted piece present in the air in the examination area.

  It can be seen that a more accurate embodiment of the inspection method provides the following. That is, the first position and the second position of the portion 14 in the first and second images are respectively represented by the first size and the second corresponding to the portion 14 calculated in the first and second images, respectively. It is to analyze in relation to the size of. In practice, the air-borne piece may move closer to or away from the optical sensor while keeping its orientation and position constant. That is, when the first and second positions substantially coincide with each other, and when the first and second sizes also substantially coincide with each other, the product 4 Is determined to be defective. On the other hand, the first position and the second position substantially coincide with each other, but when the first and second sizes are different from each other, the small piece is carried by air and is optically He was approaching and moving away from the sensor.

  Again, the first magnitude is considered to be substantially the same as the second magnitude within a preset tolerance threshold.

  At the end of the complete inspection, the control device of the optical device 6 transmits a complete inspection end response to the control unit 2 through the communication means 5. The response indicates whether the product 4 contained in the frame meets the required quality requirements, or whether the product 4 should be considered defective.

  Thus, the control unit 2 rejects the inspected product 4 when it has been determined that it is actually defective and reaches the nearest exclusion station associated with the packaging wheel. be able to.

  The delay in the acquisition of the second image relative to the acquisition of the image is at most equal to 50 milliseconds and is preferably in the interval between 30 and 40 milliseconds. In fact, when considering a production rate of at least 1000 packages per minute, the theoretical maximum inspection time for each product 4 is 60 milliseconds. Therefore, the optical device 6 must be able to secure an image acquisition frequency of at least 20 frames / second, that is, one frame every 50 milliseconds. It is clear that the image acquisition frequency of each image is strictly related to the machine manufacturing speed.

  However, the optical device 6 must ensure a sufficiently short processing time for each acquired image. That is, a complete inspection of the product 4 passing in front of the optical device 6 results in the sequential acquisition of two images, but if darker or brighter portions 14 are present in the first image, This is because both processing of the image is also brought about.

  Furthermore, it can be seen that there is a non-zero time interval between the end of inspection of one product 4 and the moment when the next product 4 to be advanced by the conveyor is to be framed in front of the optical device 6. . This time unavailable for optical inspection further reduces the time useful for a complete inspection and increases the image acquisition speed required for the optical device 6. That is, the optical device 6 must be able to quickly acquire and process two images for each product 4 to be inspected. Therefore, the optical device 6 must be selected from high performance devices. The optical device 6 must have high compactness and a reduced overall size, and furthermore, the processing time required for the control device of the optical device 6 itself must be shortened.

  According to the embodiment of the inspection method described above, the control unit 2 sends an inspection start command through the communication means 5 (in a manner similar to that described above), so that the optical device 6 is encased in the frame. Acquire the first image of the surface 4a of the product 4 and optionally turn on the associated illuminator. The inspection start command is a “simple” inspection command.

  The first image so acquired by the simple inspection command is not processed by the optical device 6, but during the response of the end of the simple inspection, together with the response, through the communication means 5 through the control unit 5 2 is transmitted.

  Thus, the first image is processed by the control unit 2, which controls the optical device 6 if the darker or brighter portion 14 is present in the received first image. Send a further simple test start command to receive the second acquired image. Then, as described above, the control unit 2 is a case where a darker or brighter portion 14 is specified in the second acquired image, and the portion 14 is the first image and the second image. If the images are at the same position, the product 4 contained in the frame is evaluated as having a defect.

  In other words, according to this embodiment, the optical device 6 is solely responsible for acquiring images and does not belong to the processing of those images. That means that the optical device 6 must ensure an appropriate speed for image acquisition and data transmission to the control unit 2. That is, the optical device 6 must only be able to acquire two images for each product 4 to be inspected, and therefore the optical device 6 should be selected from among medium performance devices. Can do it. In this way, the optical device 6 can have high compactness and reduced overall dimensions. However, on the other hand, considering the higher data processing load on the control unit 2, it is required that the control unit 2 has an image processing time shorter than a preset maximum response time.

  The image acquisition process executed by the optical device 6 with a preset acquisition frequency proceeds not only with the processing process executed at the next moment by the control unit 2 but also with the processing process executed with a different processing frequency. be able to. With currently available technology, the acquisition time interval between two consecutive images can be reduced to 10 milliseconds. The remaining constraint is the maximum time allowed for processing these two images (referred to as the maximum response time). The process must be terminated within the duration of the machine cycle, i.e. before the next product 4 arrives in front of the optical device 6.

  As described above, when the control device of the optical device 6 has only the functions of image acquisition and illuminator control, the inspection unit 3 is provided with its own control device for processing the acquired image. Can do.

  According to a further embodiment of the control method of the present invention, the inspection unit 3 receives a complete inspection start command from the control unit 2 through the communication means 5. Upon receiving such a complete inspection start command, the inspection unit 3 sends a simple inspection start command to the optical device 6 to order acquisition of the first image. The inspection unit 3 also optionally performs a further simple inspection to receive a second image from the optical device 6 if a darker or lighter portion 14 is present in the received first image. A start command is transmitted to the optical device 6.

  In other words, the image processing is executed by the control device of the inspection unit 3 through the DSP board for image processing. The DSP board sends a complete test end response to the control unit 2 at the end of the complete test. The response indicates whether the product 4 contained in the frame meets the required quality requirements, i.e. whether the product 4 should be considered defective.

  Also in this embodiment, it can be seen that the optical device 6 has only an image acquisition function. Thus, with regard to the speed of image acquisition and data transmission, what has been said above about the technical characteristics that the optical device must have. The optical device 6 can be selected from medium performance devices.

  According to a further embodiment of the above-described inspection method shown in FIG. 2, the inspection unit 3 is used for image acquisition in order to acquire the second image in the manner described above with the surface 4a of the product 4 in a frame. A second optical device 6 ′ is provided.

  The control unit 2 in the control system 1 of the packaging machine transmits a simple inspection start command to the first optical device 6 through the communication means 5 to command acquisition of the first image. Also, if the first image received through the simple examination end response has a lighter or darker portion 14, the control unit 2 will retrieve the second image from the second optical device 6 '. To receive, a further simple test start command is sent to the optical device 6 '.

  If the inspection unit 3 is provided with its own control device for processing the acquired image, the control unit 2 may send a complete inspection start command to the inspection unit 3. Continues to wait for a complete test end response from its test unit 3.

  The inspection unit 3 transmits a simple inspection start command for instructing acquisition of the first image to the first optical device 6 this time. And if the first image received through a simple test end response has a lighter or darker portion 14, then the test unit 3 is further instructed to acquire the second image. A simple inspection start command is transmitted to the second optical device 6 ′.

  According to this further embodiment, the image processing is performed by the control unit 2 and / or the inspection unit 3, and each optical device 6 or 6 'acquires only one image in each machine cycle. That is, the first optical device 6 and the second optical device 6 'must be able to acquire only one image for each product 4 to be inspected. Thus, these optical devices 6, 6 'may have less cost and reduced overall dimensions. Therefore, the delay of the acquisition of the second image relative to the acquisition of the first image is not more bound by the performance of the individual optical device (ie the camera), and the delay optimizes the selection of moving pieces. You can choose to do.

  The presence of the two optical devices 6 and 6 'makes it possible to implement further embodiments of the inspection method. The embodiment always provides for the acquisition of the first and second images spaced by a preset delay and processing only the first image in the initial step of the method. is there. On the other hand, the processing of the second image to detect possible defects is performed only if a brighter or darker portion 14 is present in the first image.

The forward speed of the product 4 in the direction D along the path P is denoted by ν, the width of the product 4 (equal to the long side in the rectangular part of the parallelepiped) is denoted by l (el), and the duration of the machine cycle is T The following relationship applies to the duration τ of the acquisition interval (acquisition suspension period).
(L / 2ν) <τ <T

  It can be seen that this further embodiment can also lead to the use of two linear sensor type optical devices 6 and 6 '. In this case, the two acquisitions are spaced by a delay set in advance according to only the acquisition speed of each column. A typical delay between two consecutive acquisitions is 0.05 milliseconds. Similar to what has been described above, the acquisition of the first and second images that are spaced apart from each other and the processing of the first image are always provided.

  The optical device 6 of FIG. 3 and the first and second optical devices 6, 6 ′ of FIG. 4 are defined by the surface 4a of the product 4 to be placed in front of the product 4, ie, to be placed in the frame. It can be seen that the optical axis A and the optical axis A ′ are substantially perpendicular to the plane.

  Such an arrangement of the optical device 6 with respect to the surface 4a of the product 4 is not always possible. This is due to the overall dimensions of the optical device 6 and the accommodated position at a preset location of the packaging machine. Alternatively, such an arrangement is not always desirable for ambient lighting conditions present in the examination area.

  However, the first single optical device 6 having the optical axis A is preset with respect to an axis B perpendicular to the plane defined by the surface 4a of the product 4 to be received in the frame (for example from 10 °). Advantageously, the optical axis A can be arranged to be inclined by an angle α (in the range between 30 °). What has just been described applies to the second optical device 6 'having the optical axis A' as well, and is tilted by an angle α 'if applicable.

  Thus, thanks to the present invention, it is possible to trust small pieces of foreign material carried by the air and passing through the examination area without adding expensive cleaning devices to the optical device 6 and / or (optionally existing) 6 '. An inspection method that can be specified in a sexual manner can be provided. In that way, it is possible to identify real defects in the product 4 that are to be eliminated, without having to provide a cleaning air flow to the inspection unit 3.

  In addition, the control unit 2 and each optical device 6, (optionally present in the examination area), so as to send and / or receive commands and / or responses for starting both full examinations and simple examinations. ) 6 ′, the inspection method during operation of the first and / or second optical device 6, 6 ′ can be changed, so that (both) the optical devices 6, 6 ′ themselves It is possible to ensure flexibility in use without requiring a complicated configuration.

  According to a further embodiment of the optical inspection method of the present invention, the inspection unit 3 is provided with an optical assembly 19 for three-dimensional scanning of the surface 4 a of the product 4. As shown in FIG. 5, an optical assembly 19 for three-dimensional scanning includes a 3D profile optical device 6 ″ capable of detecting a three-dimensional profile (three-dimensional contour shape) and a single light line 16 (preferably a laser beam). In fact, when a surface with a single line of light illuminates a surface, if the illuminated surface is flat, the projected line is a straight line; If there is a part, the projected line is a curve; if there is a corner, the projected line is a bent line, if the surface illuminated by the line of light is a surface in space The projected line is a mixed bent line with a straight line (segment), a range of curves, and a bent line.The application of them is as an application of 3D vision (stereoscopic view) Is also represented.

  With such an optical assembly 19 for three-dimensional scanning, it is possible to obtain a map of the distance of each point in the object contained in the frame with respect to the plane of the electronic sensor of the optical device 6 ″ itself. The product 4 being advanced along the path P is illuminated by a light line 16. The light line 16 is a surface of light in a direction transverse to the inspection surface defined by the surface of the surface 4a of the product 4 to be inspected. It will be seen that the line of light 16 is also in a direction transverse to the path P of the product 4 parallel to such an inspection surface.

  The 3D profile optical device 6 ″ has its own optical axis A ″ in the range between 30 ° and 60 ° with respect to the axis B perpendicular to the inspection plane, not on the plane of the line of light 16 The 3D profile optical device 6 '' detects the irradiation line profile 17 generated on the upper surface 4a of the product 4 by the light line 16 in this way. And in each acquisition, such a linear irradiation profile 17 can be transmitted to the inspection unit 3 or the control unit 2. By comparing such a linear illumination profile 17 with the profile obtained in relation to the reference surface 4a of the product 4 without defects, and based on the known trigonometric formula, the inspection unit 3 The data of the distance from the surface 4a of the product 4 at each point of the linear irradiation profile 17 is derived.

  During the advancement of the product 4 along the path P, the 3D profile optical device 6 "can continuously scan the linear profile 17. The scan is a linearly spaced, adjacent interval of the whole image. In order to obtain a profile, it can be carried out in time intervals with the same duration (determined by the advancement speed of the product 4) in succession to each other: a complete three-dimensional profile of the surface at the surface 4a of the product 4, i.e. A complete map of the distance of each point on the surface 4a of the product 4 is obtained by placing together a preset number of adjacent linear profiles (which are stored after each acquisition) side by side. Reconfigured.

  Such a reconstruction of a complete three-dimensional profile starting from the individual linear profiles 17 acquired and stored by the optical device 6 "can be performed by the control device of the optical device 6". I understand. It can also be seen that the complete three-dimensional profile thus reconstructed can be transmitted to the inspection unit 3 and the control unit 2 as a composite part of the data information. Alternatively, the reconfiguration can be performed by the inspection unit 3 or the control unit 2 that receives the individual linear profiles 17.

  The 3D profile optical device 6 ″ is a high performance optical device with a two-dimensional sensor to obtain a linear illumination profile 17 even in the presence of a curved area or a bent line, ie a continuous The optical device 6 ″ is in communication with the inspection unit 3 or the control unit 2 in the same manner as described above for the optical device for image acquisition.

  According to a further embodiment of this optical inspection method, in use, the inspection unit 3 acquires and processes at least one acquired image from the optical device 6 and processes the light intensity gradient as illustrated above.

  If there is a darker or brighter portion 14 that may be present in the image received from the optical device 6, the optical assembly 19 provides a complete three-dimensional profile, ie a map of the distance of each point on the surface of the surface 4a. Perform a 3D scan to detect. If the brighter or darker portion 14 corresponds to an object far from the surface of the surface 4a, ie a piece 18 carried by air, the portion 17 'of the linear illumination profile 17 exceeds a preset threshold. And away from the surface of the surface 4a. Even when darker or brighter portions 14 are present, the product 4 is in this case free of defects. In particular, the first position in the acquired image where the brighter or darker portion 14 is located is first identified. In the three-dimensional profile, the portions 17 ′ are positions that are separated from the surface of the surface 4 a. If each of the first position and the position in the profile are substantially equal to each other or are within a preset allowable threshold interval, it is determined that the product 4 is free of defects.

  On the other hand, if there are tears and / or dirt on the surface of the surface 4a and thus the product 4 is to be determined to be defective and therefore to be eliminated, it has been detected. A complete three-dimensional profile substantially corresponds to a similar profile of a defect-free product, except for preset tolerance threshold variations.

  The processing of the complete three-dimensional profile also allows additional observations in the identification of airborne pieces. The complete three-dimensional profile of the surface at the surface 4a of the product 4 makes it possible to sort out the folds of the end flaps 13 and 13 'present on the surface 4a of the product.

  By comparing the three-dimensional profile detected through a three-dimensional scan of the surface 4a of a product 4 with the three-dimensional profile of a reference product without defects, undesired misalignments and folds that may be present in their flaps 13, 13 ' Can stand out. Undesirable folds do not have any color (or gray level) that is clearly different from the color (or gray level) of the defect-free product 4, so they will be difficult to detect by image processing. I understand that.

  An image acquired by an optical device 6 or 6 '(which makes the portion 14 stand out, suitable for detecting gradients in color and gray levels) and a 3D profile optical device 6 "(which In the linear irradiation profile 17, the portion 17 ′ that is distant from the surface of the surface 4 a is made to stand out and is suitable for detecting the profile of the surface. As a result, the acquisitions can be started almost simultaneously.

  When the optical device 6 includes a two-dimensional sensor, it is necessary to turn off the projector 15 when the optical device 6 acquires an image of the surface 4a. That is, it is necessary to complete acquisition of the image of the surface 4a before acquiring a complete three-dimensional profile. A typical delay between acquiring those images and acquiring a profile is 5 milliseconds.

  On the other hand, if the optical device 6 includes a linear sensor, it is sufficient that the light line 16 does not hit the column housed in the frame by the linear sensor at that moment. Thus, the light line 16 always precedes or immediately follows the scan performed by the linear sensor. The two acquisitions are separated from each other by a delay that depends only on the acquisition speed of the individual columns or individual linear illumination profiles 17 (which will constitute the acquired image and the distance map, respectively). Is freed up. A typical delay between two acquisitions is 0.05 milliseconds.

  According to an example of a further embodiment of the optical inspection method, the inspection unit 3 comprises an optical assembly 19 for three-dimensional scanning that also integrates various image acquisition functions. In other words, the optical device 6 for image acquisition and the 3D profile optical device 6 ″ are integrated into a single optical device. This single optical device is a line of laser light on the product 4. As well as information about the color of each point in the line of light (backscattered light) generated by 16, and a scan train (such as distance information from them) such as a linear illumination profile 17 generated by the line 16 of laser light Both the intensity value for each point in the irradiation line produced by the line 16 of laser light reflected from the product 4 can be obtained.

  Basically, the three-dimensional scanning optical assembly 19 is provided with highly advanced technology, ie, a high dynamic range CMOS type optical device. The optical device can detect images of objects that are not very strongly illuminated and / or illuminated by a laser light source having a high luminous intensity.

  Finally, the inspection is aimed at sorting out defects on the product 4, where the defects mean a visible imperfection having a size greater than a preset minimum threshold. . The control system 1 compares such imperfections with a deviation exceeding a preset threshold when compared with the corresponding reference image of the product 4 without defects in the image of the product 4 (or at least a portion 14 thereof). Detect as.

  Defects are sorted out as portions 14 that are lighter or darker than the color (or gray tone) in the corresponding image portion showing the product 4 with no defects. The darker portion 14 of the image is not only due to defects on the package, but also passes through the framed area to detect dirt, debris, leaf tobacco fibers and other objects detected in the image by the control system 1. It can be darker or brighter.

  The present invention aims to avoid the influence of these small pieces 18 that move around the front of the product 4 housed in the frame by the optical device 6. The method which is the subject of the present invention eliminates the irregular effects of such small pieces 18 and reduces or eliminates products that are accidentally eliminated in cigarette packages and cigarette package cartons. It aims to improve the efficiency of the. The present invention further makes it possible to implement a simple and inexpensive control system 1 that processes the first and second images acquired successively without pneumatic means.

  By associating a three-dimensional scan of the profile on the surface 4a of the product 4 with the acquisition of the image, and by processing the profile in the context of the image acquired by the optical device 6, The pieces 18 moving around can be selected with higher accuracy (in this case, a portion 17 ′ of the linear irradiation profile 17 is separated from the surface of the surface 4 a by exceeding a preset threshold value).

  Finally, a three-dimensional scan of the profile on the product surface 4a can also be used advantageously to sort out defective products 4. That is when the darker or lighter portion 14 on the wrapper wrapped around the product 4 is associated with a profile that substantially corresponds to the complete three-dimensional profile of the product 4 without defects. It is. Certainly, in this case, the packaging itself may be damaged, or air-borne pieces may be permanently attached to the product 4.

Claims (7)

A method for inspecting a product (4) in a packaging machine, wherein the product (4) is a group of cigarettes wrapped in a wrapping material and the machine is a machine for packaging a cigarette package Alternatively, the product (4) is a group of cigarette packages wrapped in a packaging material, and the machine is a machine for packaging the group of cigarette packages, and the inspection method comprises the machine A step of advancing the product (4) on a conveyor of the image and processing an acquired image of one side (4a) of the product (4) advanced by the conveyor to at least detect defects that may be present in the image a free product corresponding image as compared to brighter or darker portion (4) (14), comprising the steps of: identifying a portion (14) corresponding to the pieces airborne, the In the,
The method comprises performing a three-dimensional scan of the surface (4a) of the product (4) when the image has the brighter or darker portion (14); and the surface of the surface (4a) Detecting a complete three-dimensional profile, and a portion (17 ′) away from the surface of the surface (4a) in the three-dimensional profile, and the first image and the first image In the three-dimensional profile, when the lighter or darker part (14) and the remote part (17 ′) of the surface (4a) are substantially at the same position, the product (4 ) Is evaluated as being free of defects, and
The step of performing the three-dimensional scanning includes a step of providing an optical assembly (19), and the optical assembly (19) includes a projector (15) of a light line (16) and a linear irradiation profile (17). A 3D profile optical device (6 ″) suitable for detection,
The step of detecting the complete three-dimensional profile comprises scanning the linear illumination profile (17) continuously during advancement of the product (4) on the conveyor of the machine, and after each scan Storing the irradiation linear profile (17); and reconfiguring the complete three-dimensional profile by arranging together a predetermined number of adjacent linear irradiation profiles. and that, wherein the. Identify the first position where the brighter or darker part (14) is located in the first image, and identify the position where the part (17 ') away from the surface exists in the profile with, when the the first position located substantially coincident in the profile, and evaluating as said product (4) there is no defect, including a method of claim 1, wherein. 3. The method of claim 1, further comprising providing an optical device (6) for acquiring an image of the surface (4a) and another 3D profile optical device for acquiring the complete three-dimensional profile. The method described. The optical device (6) for image acquisition comprises a two-dimensional sensor,
The method, during advancement of the product (4) includes the step of terminating the acquisition of the image prior to the continuous scanning of the linear illumination profile (17), according to claim 3 the method of. The optical device (6) for image acquisition comprises a linear sensor, the step of acquiring the image comprises the step of performing continuous acquisition of the columns of images and storing each column And reconstructing the image by arranging together a pre-set number of acquired columns together, the method comprising a sequence of images and a linear illumination profile (17 4. The method of claim 3 , further comprising the step of acquiring at a time, particularly at intervals of 0.05 milliseconds. Providing a single optical device (6 ″) for acquisition of both the luminous intensity of the image of the surface (4a) and the linear illumination profile (17) produced by the line of light (16). The method according to any one of claims 1 to 5 , further comprising: The 3D profile optical device (6 ″) having the optical axis (A ″) is configured so that the optical axis (A ″) is preset with respect to the plane of the surface (4a) of the product (4) (α ”), And the angle (α ″) is defined between the optical axis (A ″) and an axis (B) perpendicular to the plane, in particular 30 ° from an angle ranging between 60 °, the method according to any one of claims 1 6.

Images