Method, system and storage medium for quality detection and rejection control of cigarette filter rods

Abstract

The application provides a control method, system and storage medium for quality detection and rejection of a cigarette filter rod, which comprises the following steps: S1: setting a target detection mode under a conveying belt transmission process inside a quality detection device, the target detection mode comprising a transmission speed, a transmission rotational inertia and a sensing acquisition mode, the transmission rotational inertia being used for filter rod rotation surface control; S2: as the target cigarette filter rod is transmitted, target scanning sampling is performed through the target detection mode control to determine N two-dimensional contour point clouds; S3: the N two-dimensional contour point clouds are transmitted to a quality detection module for three-dimensional fitting reconstruction and geometric evaluation of the contour point clouds to determine a quality detection result; and S4: according to the quality detection result, the target cigarette filter rod is output at a first sorting end and / or a second sorting section of the quality detection device. The application improves the quality detection precision and accuracy of the cigarette filter rod, thereby improving the product quality stability.

Description

Technical Field

[0001] This application belongs to the field of cigarette testing, specifically relating to a control method, system, and storage medium for quality testing and rejection of cigarette filter rods. Background Technology

[0002] Precise inspection of the appearance, shape, dimensions, and surface quality of cigarette filter rods during production, and rejection of substandard products, is a crucial step in ensuring the stability of cigarette quality.

[0003] Existing methods for inspecting cigarette filter rod quality primarily rely on mechanical inspection or single-sensor devices, such as photoelectric sensors, laser rangefinders, or mechanical gauges, to statically or dynamically inspect the key dimensions of the filter rod. These methods determine the pass / fail status of the filter rod based on preset inspection thresholds and sort out substandard filters. However, on the one hand, the unstable posture of the filter rod during dynamic transport and the insufficient control precision of the mechanical transmission equipment can easily lead to inconsistent data acquisition, thus affecting the reliability of the inspection results. On the other hand, existing inspection equipment has a single acquisition dimension, only acquiring geometric data of a local area or a single angle of the filter rod, and cannot accurately capture the comprehensive geometric features of the filter rod. This limitation results in insufficient inspection accuracy, a high rate of missed detections and false detections, and consequently affects the stability of product quality. Summary of the Invention

[0004] In view of this, the purpose of this application is to provide a control method, system and storage medium for quality inspection and rejection of cigarette filter rods. This method solves the technical problems of insufficient dynamic control precision of cigarette filter rods and single data acquisition dimension in the prior art, which leads to the inability to accurately collect comprehensive geometric feature data of cigarette filter rods, resulting in insufficient quality inspection precision and high rates of missed and false detections. This method improves the precision and accuracy of cigarette filter rod quality inspection, thereby enhancing the stability of product quality.

[0005] To solve the above-mentioned technical problems, this application adopts the following technical solution: In a first aspect, this application provides a control method for quality inspection and rejection of cigarette filter rods. The control method includes: Step S1: Setting a target detection mode under the conveyor belt transmission process inside the quality inspection equipment. The target detection mode includes transmission speed, transmission rotational inertia, and sensor acquisition mode. The transmission rotational inertia is used for filter rod rotation control; Step S2: As the target cigarette filter rod is transmitted, target scanning sampling is performed through the target detection mode control to determine N two-dimensional contour point clouds; Step S3: The N two-dimensional contour point clouds are transmitted to the quality inspection module for three-dimensional fitting reconstruction and geometric evaluation of the contour point clouds to determine the quality inspection result; Step S4: Based on the quality inspection result, the target cigarette filter rod is output to the first sorting end and / or the second sorting section of the quality inspection equipment.

[0006] Further, step S1 includes: Step S11: Setting the transmission speed of the conveyor belt; Step S12: Calculating the transmission rotational inertia based on the transmission speed, constrained by the length of the conveyor belt in the detection area, and requiring the tangential periodic rotation of the cigarette filter rod; Step S13: Performing uniform sensor configuration on the conveyor belt in the detection area, and calculating the sampling detection sequence triggered by the sensor based on the transmission speed and the transmission rotational inertia, as the sensor acquisition mode; Step S14: Integrating the transmission speed, transmission rotational inertia, and sensor acquisition mode to determine the target detection mode.

[0007] Furthermore, the construction steps of the quality detection module include: Step M1: Building a three-dimensional reconstruction unit based on point cloud feature recognition and reconstruction based on two-dimensional relative planes; Step M2: Constructing a first reference branch based on the standard geometric data and standard surface features of the interactive cigarette filter rod; Step M3: Establishing a parallel connection between the first reference branch and the second real-time branch to generate the quality detection unit, wherein the second real-time branch is used to receive the output of the three-dimensional reconstruction unit, and the first reference branch and the second real-time branch establish a lateral interaction channel; Step M4: Post-connecting the quality detection unit to the three-dimensional reconstruction unit, performing sample-driven training until convergence, and obtaining the quality detection module.

[0008] Further, in step S31: the N two-dimensional contour point cloud items are identified by their relative planar relationships; in step S32: the N two-dimensional contour point clouds are transmitted to the three-dimensional reconstruction unit to identify the point cloud features of the target cigarette filter rod, wherein the point cloud features include ranging, position coordinates, and surface features, and the ranging is the spatial relationship between the sensor configuration position and the scanning point cloud position; in step S33: based on the ranging, position coordinates, and relative planar relationships, the three-dimensional position coordinates are converted and determined; in step S34: the three-dimensional contour point cloud is reconstructed according to the three-dimensional position coordinates; in step S35: based on the surface features, the three-dimensional contour point cloud is identified to determine the three-dimensional reconstruction result.

[0009] Further, step S331: Identify the two-dimensional point cloud at the edge position, and calculate the neighboring point cloud vector based on the distance measurement and position coordinates. The neighboring point cloud is the adjacent first point cloud and second point cloud. The conveyor belt plane is used as the reference plane, and the direction and angle of change of the second point cloud relative to the first point cloud are used as the point cloud vector. Step S332: Calculate the Gaussian curvature of each two-dimensional point cloud at the edge position based on the neighboring point cloud vector. Step S333: Determine whether the Gaussian curvature is consistent and determine the first judgment result. Step S334: Identify the two-dimensional point cloud at the lateral position, perform line type determination of the point cloud distribution, and determine the second judgment result. Step S335: If the first judgment result is inconsistent or the second judgment result is non-linear, mark the target cigarette filter rod as a defective product and terminate the quality inspection process.

[0010] Further, in step S336: if the first judgment result is consistent and the second judgment result is linear, perform three-dimensional point cloud conversion and surface feature identification based on the relative planar relationship to determine the three-dimensional reconstruction result; in step S337: transmit the three-dimensional reconstruction result to the second real-time branch of the quality detection unit; in step S338: based on the lateral interaction channel, perform feature mapping between the first reference branch and the second real-time branch, compare the consistency of the contour geometric features and surface features, and determine the comparison result; in step S339: check the quality accuracy of the cigarette filter rod, and based on the quality accuracy, make a qualified judgment by comparing the results and output the quality detection result.

[0011] Further, step S41: mark the target cigarette filter rod with an electronic tag according to the quality inspection results; step S42: at the output port of the quality inspection equipment, by recognizing the electronic tag, perform output control on the target cigarette filter rod based on the first sorting end and the second sorting end, wherein the first sorting end is the quality qualified output end, and the second sorting end is the quality unqualified rejection end.

[0012] Secondly, this application provides a control system for the quality inspection and rejection of cigarette filter rods. The control system is used to execute the aforementioned control method for the quality inspection and rejection of cigarette filter rods. The control system includes: a detection module setting unit, a scanning sampling unit, a quality inspection unit, and a sorting output unit. The detection mode setting unit is used to set the target detection mode under the conveyor belt transmission process inside the quality inspection equipment. The target detection mode includes transmission speed, transmission rotational inertia, and sensor acquisition mode. The transmission rotational inertia is used for filter rod rotation control. The scanning sampling unit is used to perform target scanning sampling under the target detection mode control as the target cigarette filter rod is transmitted, and to determine N two-dimensional contour point clouds. The quality inspection unit is used to transmit the N two-dimensional contour point clouds to the quality inspection module, perform three-dimensional fitting reconstruction and geometric evaluation of the contour point clouds, and determine the quality inspection result. The sorting output unit is used to output the target cigarette filter rod to the first sorting end and / or the second sorting section of the quality inspection equipment according to the quality inspection result.

[0013] Thirdly, this application provides a computer-readable storage medium storing a computer program / instructions thereon, which, when executed by a processor, implements the steps of the control method for quality detection and rejection of cigarette filter rods described above.

[0014] Fourthly, this application provides a computer program product, including a computer program / instructions, which, when executed by a processor, implement the steps of the above-described control method for quality detection and rejection of cigarette filter rods.

[0015] As can be seen from the above technical solutions, the advantages and positive effects of the control method, system, and storage medium for quality detection and rejection of cigarette filter rods proposed in this application are as follows: This application achieves precise detection and efficient control of cigarette filter rod quality by combining multiple steps, including setting a precise detection mode, efficiently acquiring two-dimensional contour point clouds, performing three-dimensional fitting reconstruction and geometric evaluation, and automated sorting output. It can accurately acquire the geometric feature data of the filter rod, improve detection accuracy, and reduce the rate of missed detections and false detections. At the same time, due to accurate detection and automated sorting, the quality of cigarette filter rods is improved, and the stability of product quality is enhanced. Attached Figure Description

[0016] The above description of this application and the following detailed embodiments will be better understood when read in conjunction with the accompanying drawings. It should be noted that the drawings are merely examples of the claimed technical solutions.

[0017] Figure 1 This is a flowchart of the control method for quality inspection and rejection of cigarette filter rods in this application; Figure 2 This is the detection waveform of the defective filter rod in this application; Figure 3 This is the test waveform of the qualified filter rod of this application; Figure 4 This is a structural diagram of a control system for the quality detection and rejection of cigarette filter rods.

[0018] The reference numerals in the attached figures are explained as follows: Detection mode setting unit: 10; Scanning sampling units: 20; Quality inspection units: 30; Sorting output unit: 40. Detailed Implementation

[0019] The detailed features and advantages of this application are described below in the specific embodiments. The content of this description is sufficient to enable any person skilled in the art to understand the technical content of this application and implement it accordingly. Based on the specification, claims and drawings disclosed in this specification, a person skilled in the art can easily understand the related objectives and advantages of this application.

[0020] The invention will now be described with reference to the accompanying drawings, in which similar reference numerals denote similar elements. While specific structures and arrangements are discussed, it should be understood that this is done merely for illustrative purposes. Those skilled in the art will recognize that other structures and arrangements can be used without departing from the spirit and scope of the invention. It will be apparent to those skilled in the art that the invention can also be used in a variety of other applications.

[0021] In this specification and claims, several terms will be used, and unless otherwise indicated, these terms will be defined to have the following meanings: The singular forms “a” and “the” include their corresponding plural forms. “At least one” means one or more, and “more” means two or more. “At least one of the following” or similar expressions refer to any combination of these items, including any combination of single or plural items. For example, at least one of a, b, or c can be expressed as: a, b, c, ab, ac, bc, or abc, where a, b, and c can be single or multiple.

[0022] All figures used to represent component amounts, properties (e.g., molecular weight), reaction conditions, etc., should be considered to be modified in all cases by the terms "within the unavoidable margin of error" or "about". Therefore, the numerical values ​​set forth herein are approximate and may vary depending on the desired properties sought to be obtained by the present invention. The principles of equivalents, which are applied to a minimum and not intended to limit the scope of the claims, should be applied, for example, each value should be interpreted at least according to the reported significant digits and by applying conventional rounding techniques.

[0023] It should be understood that the term "and / or" in this article is merely a description of the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can represent: A existing alone, A and B existing simultaneously, or B existing alone. A and B can be singular or plural. Additionally, the character " / " in this article generally indicates an "or" relationship between the preceding and following related objects, but it can also represent an "and / or" relationship. Please refer to the context for a more accurate understanding.

[0024] In the description of this embodiment, it should be noted that the terms "upper", "lower", "inner", "bottom", etc., indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings, or the orientation or positional relationship in which the product is usually placed during use. They are only for the convenience of describing this application and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation. Therefore, they should not be construed as limitations on this application.

[0025] Unless otherwise indicated, the following abbreviations have the following meanings, and any other abbreviations used herein but not defined have their generally accepted standard meanings: All other terms used herein for special definition are intended to have the general meaning understood by one of ordinary skill in the art, and in particular, meaning that one of ordinary skill in the art, upon reading the claims, specification and drawings of this patent, can directly and without doubt determine how the technical solution of this patent can be implemented.

[0026] Even if there are incomplete descriptions, omissions, or ambiguities in the grammar, words, punctuation, graphics, symbols, etc. of the claims, specification, and drawings of this patent, a person skilled in the art can still arrive at the only correct understanding by reading the claims, specification, and drawings as a whole without extensive reasoning or experimentation, and effectively exclude various incorrect interpretations that are not aimed at achieving the purpose of this patent.

[0027] Those skilled in the art would first choose to read the claims, specification, and drawings of this patent to reasonably interpret the terms; secondly, they would choose to refer to the relevant definitions in other documents published by the applicant before the filing date to reasonably interpret the terms; thirdly, they would choose the references cited in this patent to reasonably interpret the terms; and finally, they would choose to combine the technical dictionaries, technical manuals, reference books, textbooks, national or industry technical standards, etc., commonly used by those skilled in the art to reasonably interpret the terms.

[0028] To make the objectives, technical solutions, and advantages of this application clearer, the embodiments of this application will be described in further detail below with reference to the accompanying drawings.

[0029] Please refer to Figure 1 This application provides a control method for quality detection and rejection of cigarette filter rods, the specific steps of which include: Step S1: Set the target detection mode under the conveyor belt transmission process inside the quality inspection equipment.

[0030] The target detection mode includes transmission speed, transmission rotational inertia and sensor acquisition mode. The transmission rotational inertia is used for filter rod rotation control.

[0031] Specifically, quality testing equipment is an instrument used to test the quality of cigarette filter rods. It integrates multiple functional modules, such as a conveyor belt transmission system, a sensor acquisition device, a quality testing module, and a sorting end, and can detect and analyze various quality indicators of the filter rods and sort them according to the results.

[0032] Before starting the quality inspection of cigarette filter rods, it is necessary to set the target inspection mode under the conveyor belt transmission process inside the quality inspection equipment. This target inspection mode is configured according to the quality inspection requirements, including the transmission speed, transmission rotational inertia and sensor acquisition mode, etc., to optimize the dynamic operation performance and data acquisition accuracy of the equipment.

[0033] First, the conveyor speed inside the quality inspection equipment is set according to the characteristics of the filter rod, the required detection accuracy, and the equipment's processing capacity. For example, for smaller filter rods with high detection accuracy, a slower conveyor speed is needed to ensure the sensors have sufficient time to acquire data. Second, the transmission rotational inertia is adjusted. This inertia is used for filter rod rotation control, ensuring the filter rod rotates uniformly so that all surfaces can be detected. This can be achieved by adjusting the conveyor belt's friction or the drive system's inertial parameters to prevent the filter rod from tilting or rotating uncontrollably during transport, thus providing stable input for subsequent data acquisition. Finally, the sensor's acquisition mode is set according to the detection requirements, such as the sensor's acquisition position and sequence.

[0034] By setting the target detection mode appropriately, the filter rod can be kept in a stable state during subsequent detection, ensuring that data from all sides can be accurately collected, thus improving the accuracy and comprehensiveness of data collection.

[0035] Specifically, step S1 includes: Step S11: Set the conveyor belt speed.

[0036] Step S12: Based on the transmission speed, with the length of the conveyor belt in the detection area as a constraint and the tangential periodic rotation of the cigarette filter rod as a requirement, calculate the transmission rotational inertia.

[0037] Step S13: Perform uniform sensor configuration on the conveyor belt in the detection area, and calculate the sampling and detection timing of the sensor trigger based on the transmission speed and transmission rotational inertia, which serves as the sensor acquisition mode.

[0038] Step S14: Integrate transmission speed, transmission moment of inertia, and sensor acquisition mode to determine the target detection mode.

[0039] Specifically, the conveyor belt speed is set based on factors such as production needs, required testing accuracy, and equipment performance. For example, if the conveyor belt is 2 meters long and the filter rod is required to pass through the testing area in 0.5 seconds, then the speed should be set to 4 m / s.

[0040] Based on the pre-set transmission speed, the length of the conveyor belt in the detection area is used as a constraint. Since the movement time of the filter rod on the conveyor belt is related to both the conveyor belt length and the transmission speed (time = length / speed), the transmission moment of inertia is calculated using relevant mechanical and kinematic formulas based on the required tangential periodic rotation of the cigarette filter rod. For example, if the mass, radius, and desired rotation period of the filter rod are known, the moment of inertia can be calculated using the formula I = mr. 2(Where m is mass and r is radius), and considering the motion of the filter rod on the conveyor belt, the corresponding transmission moment of inertia is calculated. By adjusting the conveyor belt drive system according to the calculation results, the moment of inertia is adjusted to ensure that the tangential rotation period of the filter rod in the detection area is stable, so that all surfaces of the filter rod can be detected at appropriate positions and times, thus improving the comprehensiveness of the detection.

[0041] Sensors are evenly distributed on the conveyor belt in the detection area to ensure the uniformity and completeness of data collection from the cigarette filter rods. Then, based on the determined transmission speed and moment of inertia, the sampling timing for sensor triggering is calculated—that is, the order in which the sensors collect data—to ensure that each sampling covers the entire or critical features of the filter rod. For example, if the transmission speed is 0.5 m / s and the moment of inertia causes the filter rod to rotate once per second, the sensor can trigger sampling at integer multiples of the second.

[0042] After determining the transmission speed, transmission moment of inertia, and sensor acquisition mode, the control system of the quality inspection equipment integrates these three parameters. For example, the control software of the quality inspection equipment stores and recalls these three parameters as a unified detection mode setting. When the inspection process begins, the quality inspection equipment controls the operation of the conveyor belt, the rotation of the filter rod, and the data acquisition of the sensors according to this integrated target detection mode. This ensures the coordination between the transmission system and the inspection system during the cigarette filter rod inspection process, providing a stable and efficient operating environment for subsequent data acquisition and inspection procedures.

[0043] Step S2: As the target cigarette filter rod is transmitted, target scanning and sampling are performed through target detection mode control to determine N two-dimensional contour point clouds.

[0044] Specifically, when the target cigarette filter (i.e., the cigarette filter to be quality inspected) is transported along the conveyor belt according to the target inspection mode set in step S1, the sensor is activated to scan and sample the target cigarette filter. For example, an optical sensor is used to scan the surface of the filter at certain time intervals or spatial intervals. As the filter moves, the sensor continuously collects data points on the surface of the filter. These data points, based on their positional relationship on the surface of the filter, gradually construct N two-dimensional contour point clouds. Here, N is a positive integer, representing the number of two-dimensional contour point clouds collected from different parts or angles, which varies depending on the complexity of the filter's shape and the inspection requirements. These two-dimensional contour point clouds contain the contour information of the filter on a certain plane (such as a cross-section or longitudinal section).

[0045] Two-dimensional contour point clouds are obtained by target scanning sampling, realizing the conversion from the physical shape of the filter rod to digital data, providing basic data for subsequent three-dimensional fitting and reconstruction, and accurately reflecting the contour features of the filter rod on a certain plane.

[0046] Step S3: Transmit the N two-dimensional contour point clouds to the quality inspection module, perform three-dimensional fitting reconstruction and geometric evaluation of the contour point clouds, and determine the quality inspection results.

[0047] Specifically, the quality inspection module is the control center built into the quality inspection equipment. It contains various algorithms and processing programs (such as shape matching algorithms and size deviation calculation programs) to analyze and process the collected filter rod data (such as two-dimensional contour point clouds) to determine whether the filter rod is of qualified quality.

[0048] The N two-dimensional contour point clouds obtained in step S2 are transmitted to the quality inspection module. In the quality inspection module, a 3D fitting and reconstruction algorithm is used to combine the two-dimensional contour point clouds into a 3D model of the filter rod based on their spatial relationships and order. Then, the reconstructed 3D model is geometrically evaluated. A shape comparison algorithm can be used to compare the reconstructed filter rod model with a standard filter rod model, calculating indicators such as dimensional deviation and surface flatness. For example, a dimensional deviation calculation program can be used to measure the difference between the dimensions of each part of the filter rod model and the standard dimensions to determine whether the filter rod is qualified and generate the quality inspection results.

[0049] By using 3D fitting reconstruction and geometric evaluation, the geometric features of the filter rod can be comprehensively and accurately detected, thereby determining whether the quality of the filter rod meets the requirements. Compared with traditional 2D inspection methods, this 3D inspection can more accurately detect shape defects and dimensional deviations of the filter rod, improving the accuracy of quality inspection.

[0050] Specifically, step S3 may include: Step S31: The N two-dimensional contour point cloud identifiers have relative planar relationships.

[0051] Step S32: Transmit the N two-dimensional contour point clouds to the three-dimensional reconstruction unit to identify the point cloud features of the target cigarette filter rod.

[0052] The point cloud features include ranging, position coordinates, and surface features. The ranging is the spatial relationship between the sensor configuration position and the scanned point cloud position.

[0053] Step S33: Based on the distance measurement, position coordinates and relative plane relationship, convert and determine the three-dimensional position coordinates.

[0054] Specifically, step S33 includes: Step S331: Identify the two-dimensional point cloud at the edge position, and calculate the neighboring point cloud vector based on the distance measurement and position coordinates. The neighboring point cloud is the first point cloud and the second point cloud that are adjacent to each other. The conveyor belt plane is used as the reference plane, and the direction and angle of change of the second point cloud relative to the first point cloud are used as the point cloud vector.

[0055] Step S332: Calculate the Gaussian curvature of each two-dimensional point cloud at the edge position based on the neighborhood point cloud vector; Step S333: Determine whether the Gaussian curvature is consistent and determine the first judgment result.

[0056] Step S334: Identify the two-dimensional point cloud at the lateral position, determine the line type of the point cloud distribution, and determine the second determination result.

[0057] Specifically, before performing the three-dimensional position coordinate transformation, the cigarette filter rod is initially inspected based on the point cloud features of the identified target cigarette filter rod.

[0058] Two-dimensional point clouds at edge positions are identified based on their position coordinates. During the identification process, a certain coordinate range can be set, or the edge position can be determined based on the density changes of the point cloud. Then, based on the ranging (the spatial relationship between the sensor configuration position and the scanned point cloud position) and the position coordinates, a neighborhood point cloud vector is calculated. This neighborhood point cloud vector reflects the local variation trend of the point cloud at the edge position.

[0059] A neighborhood point cloud refers to any two point clouds that are adjacent to each other, namely the first point cloud and the second point cloud. The point cloud vector is a vector formed by the direction and angle of change of the second point cloud relative to the first point cloud, with the conveyor belt plane as the reference surface. This vector reflects the local change trend of the point cloud at its edge position. The calculation of the neighborhood point cloud vector can be performed using vector operations: randomly select two adjacent point clouds from the two-dimensional point clouds at the edge position, namely the first and second point clouds. Using the conveyor belt plane as the reference surface, calculate the direction and angle of change of the second point cloud relative to the first point cloud to form the neighborhood point cloud vector. This vector reflects the local change trend of the point cloud at its edge position. For example, if the coordinates of the first point cloud are (x1, y1) and the coordinates of the second point cloud are (x2, y2), then the component of the point cloud vector in the x-direction is x2 - x1, and the component in the y-direction is y2 - y1. The change angle can be calculated using the arctangent function. Calculation. Calculating the neighborhood point cloud vector can describe the local variation characteristics of the point cloud at the edge position, which helps in subsequent analysis of the shape changes at the edge position and provides local geometric information for judging the quality of the filter rod.

[0060] Based on the calculated neighborhood point cloud vectors, the Gaussian curvature of each 2D point cloud at the edge location is calculated using the Gaussian curvature calculation formula. For discrete 2D point cloud data, a discrete Gaussian curvature calculation method can be used. For example, a local covariance matrix can be constructed based on the neighborhood point cloud vectors, and then the Gaussian curvature can be calculated using the eigenvalues ​​of the covariance matrix. Calculating the Gaussian curvature can quantify the bending characteristics at the edge location. By analyzing the Gaussian curvature values, it can be determined whether the shape of the filter rod edge meets expectations, such as whether there is local excessive bending or unevenness.

[0061] like Figure 2 and Figure 3 As shown, the cross-section of a cigarette should be circular, therefore the Gaussian curvature at all locations should be the same. Any discrepancies indicate potential depressions or convexities. The calculated Gaussian curvatures of the two-dimensional point clouds at each edge location are compared. If the Gaussian curvature values ​​of all point clouds are equal within a certain error range, they are considered consistent, and the first determination result is "consistent." If the Gaussian curvature values ​​of any point cloud exceed the error range, they are considered inconsistent, and the first determination result is "inconsistent." The first determination result is then used to identify the target cigarette filter rod.

[0062] Two-dimensional point clouds in lateral positions can be identified by comparing their coordinates with pre-defined edge positions. Then, the linearity of the point cloud distribution is determined. A fitting algorithm, such as least squares fitting, can be used to fit the point cloud into a curve, and then it is determined whether this curve conforms to preset linear characteristics. If the equation of the fitted curve matches the preset straight line or arc equation within a certain error range, it is determined that a linearity exists (the second determination result is "linearity exists"); otherwise, it is determined that a non-linearity exists (the second determination result is "non-linearity exists"). The second determination result is then used to identify the target cigarette filter. The linearity determination of the point cloud distribution can detect whether the side shape of the filter meets the requirements. If the side shape does not conform to the linear characteristics, it may indicate quality problems such as deformation on the side of the filter.

[0063] If either the first judgment result is "inconsistent" or the second judgment result is "nonlinear existence," the target cigarette filter is marked as defective. In practice, this can be achieved by setting a flag in the detection system or modifying the filter's status flag. Once marked as defective, the quality inspection process is terminated, avoiding unnecessary subsequent inspections of already determined defective filters, thereby improving inspection efficiency.

[0064] By determining the Gaussian curvature and linearity of the point cloud distribution, filter rods that clearly do not meet quality requirements can be promptly identified, reducing the workload of subsequent testing steps and improving the efficiency of the entire quality inspection process. Simultaneously, accurately identifying defective products also aids in the classification and statistical analysis of non-conforming items.

[0065] Step S335: If the first judgment result is inconsistent and the second judgment result is nonlinear, mark the target cigarette filter rod as defective and terminate the quality inspection process.

[0066] Step S336: If the first determination result is consistent and the second determination result is linear, combine the relative planar relationship to perform three-dimensional point cloud conversion and surface feature identification to determine the three-dimensional reconstruction result.

[0067] Step S337: Transmit the 3D reconstruction results to the second real-time branch of the quality inspection unit.

[0068] Step S338: Based on the lateral interaction channel, perform feature mapping between the first reference branch and the second real-time branch, perform consistency comparison between contour geometric features and surface features, and determine the comparison result.

[0069] Step S339: Analyze the quality accuracy of the cigarette filter rod. Based on the quality accuracy, compare the results to determine if the quality is acceptable and output the quality inspection results.

[0070] Specifically, after the initial inspection is passed, the target cigarette filter rod is examined more carefully to identify minor errors.

[0071] When the first judgment result is consistent and the second judgment result is linear, a 3D point cloud transformation is first performed based on the relative plane relationship. Assuming the relative planes are distributed along a certain axis (e.g., the z-axis), and the 2D point cloud is known to be in the xy-plane, the 2D coordinates (x, y) are converted to 3D coordinates (x, y, z) by combining the position information of the relative planes (e.g., their index or distance in the z-axis direction). For surface feature identification, additional attribute fields can be added to the 3D point cloud data structure to store surface feature information. For example, if a structure is used to represent the 3D point cloud, a field can be added to the structure to store the surface roughness value. The 3D point cloud transformation and surface feature identification process is completed by the 3D reconstruction unit. Through 3D point cloud transformation and surface feature identification, a complete 3D reconstruction result can be obtained. This result contains accurate 3D shape information and surface feature information, providing a comprehensive data foundation for subsequent quality inspection.

[0072] The obtained 3D reconstruction results are directly transmitted to the second real-time branch of the quality inspection unit. This process can be achieved through a data transmission interface. For example, in a software system, function calls or message passing mechanisms can be used to transfer the 3D reconstruction results from the 3D reconstruction unit to the second real-time branch module of the quality inspection unit.

[0073] Based on the lateral interaction channel, feature mapping between the first reference branch and the second real-time branch is performed in the quality inspection unit. This can be achieved by establishing a data mapping relationship. For example, if the standard geometric features in the first reference branch are stored in a certain data structure (such as an array or structure), the corresponding actual detected geometric feature data is found in the second real-time branch, and a mapping relationship is established. Then, a consistency comparison is performed on the contour geometric features (such as the diameter and length of the filter rod) and surface features (such as surface roughness) to determine whether the actual detection results are consistent with the standard data, generating a corresponding comparison result, such as the difference between the actual detection results and the standard data. Through feature mapping and consistency comparison, the difference between the actual features and standard features of the cigarette filter rod can be accurately determined, thus providing a basis for judging whether the cigarette filter rod is qualified.

[0074] The quality accuracy of the cigarette filter rod can be determined by reading from a quality standard database or obtaining from a preset parameter file. Then, based on the quality accuracy, a comparison is made to determine whether the result is acceptable, and the quality inspection result is output: if the comparison result meets the quality accuracy requirements, it is judged as acceptable; otherwise, it is judged as unacceptable. For example, if the quality accuracy requirement is that the filter rod diameter error is within ±0.1mm, and the comparison result shows that the diameter error is within this range, it is judged as acceptable; otherwise, it is judged as unacceptable. Finally, the quality inspection result is output, which can be displayed on a display device, stored in a storage device, or interacted with other production control systems. By comparing the results based on quality accuracy and determining whether the result is acceptable, and outputting the quality inspection result, it is possible to accurately determine whether the cigarette filter rod meets the quality requirements, thereby effectively controlling product quality during the production process.

[0075] Step S34: Reconstruct the 3D contour point cloud based on the 3D position coordinates; Step S35: Identify the 3D contour point cloud based on surface features and determine the 3D reconstruction result.

[0076] Specifically, the N two-dimensional contour point cloud identifiers acquired through target scanning sampling have relative planar relationships. These relative planar relationships represent the positional relationships of different two-dimensional contour point clouds within the spatial structure of the target cigarette filter rod, reflecting the relative order and spatial layout of each two-dimensional contour point cloud when constructing the three-dimensional model. For example, for an approximately cylindrical filter rod, the axial positional relationship between two-dimensional contour point clouds of different cross-sections is a relative planar relationship. These two-dimensional contour point clouds are transmitted to a three-dimensional reconstruction unit, which identifies the point cloud features of the target cigarette filter rod. The point cloud features here include three aspects: ranging, position coordinates, and surface features. Ranging represents the spatial relationship between the sensor's location and the scanned point cloud location, i.e., the distance between the sensor's location and the acquired point cloud in space. Position coordinates refer to the coordinate position of each point cloud in a certain coordinate system. This coordinate system can be a relative coordinate system defined by the position coordinate device, which accurately describes the position of the point cloud in space. Surface features are the characteristics of the cigarette filter rod surface reflected by the point cloud, such as surface flatness and roughness.

[0077] In the 3D reconstruction unit, for distance measurement identification, the spatial relationship between the sensor configuration position and the scanned point cloud position can be determined using sensor parameters (such as the angle and intensity of emitted light) and the time of receiving reflected light, along with algorithms such as triangulation. For position coordinate identification, the position coordinates of each point cloud can be determined based on the sensor's layout on the conveyor belt and the scanning order, combined with a pre-defined coordinate system. For surface feature identification, image processing techniques can be used, such as analyzing the grayscale value changes of the point cloud to infer features like surface smoothness. By identifying point cloud features, especially distance measurement, position coordinates, and surface features, the position and characteristics of each point cloud in 3D space can be determined more accurately, providing fundamental data for subsequent 3D fitting and reconstruction, and contributing to the construction of a more accurate 3D model.

[0078] Based on the identified distance measurement, position coordinates, and relative planar relationships, coordinate transformation is performed to determine the three-dimensional position coordinates. For example, if a coordinate system is established with the conveyor belt's running direction as the x-axis, the direction perpendicular to the conveyor belt as the y-axis, and the direction perpendicular to the conveyor belt plane as the z-axis, given the position coordinates of the two-dimensional contour point cloud in the xy-plane, the two-dimensional coordinates are converted into three-dimensional coordinates using distance measurement information (such as the distance to the z-axis obtained through triangulation) and relative planar relationships (such as the position along the filter rod's axis) and spatial geometric calculation formulas.

[0079] Based on the determined 3D position coordinates, a 3D contour point cloud is reconstructed using point cloud processing algorithms. For example, an interpolation algorithm can be used to interpolate between point clouds based on their known 3D position coordinates, filling in any gaps and making the point clouds denser and more continuous, thus reconstructing the 3D contour point cloud. Alternatively, a surface fitting algorithm can be used to fit the point cloud to an approximate 3D surface, thereby obtaining the 3D contour point cloud. The obtained 3D contour point cloud is a 3D representation of the 2D contour point cloud obtained by determining its 3D position coordinates, and it more accurately reflects the actual shape of the target cigarette filter rod compared to a 2D contour point cloud.

[0080] Based on the identified surface features, the reconstructed 3D contour point cloud is labeled. For example, if a certain area is identified as having high surface roughness, it can be marked in the corresponding area of ​​the 3D contour point cloud. This labeling can be achieved by adding additional attributes or marker bits to the point cloud data structure. After labeling the surface features of each part of the 3D contour point cloud, the result containing the complete shape and surface feature information of the cigarette filter is the 3D reconstruction result. The 3D reconstruction result provides the quality inspection module with complete 3D information of the target cigarette filter, including its shape and surface features, enabling the quality inspection module to more comprehensively and accurately assess the quality of the filter.

[0081] The construction steps of the quality inspection module include: Step M1: Construct a 3D reconstruction unit by recognizing point cloud features and reconstructing based on two-dimensional relative planes.

[0082] Step M2: Construct the first reference branch based on the standard geometric data and standard surface features of the interactive cigarette filter rod.

[0083] Step M3: Establish a parallel connection between the first reference branch and the second real-time branch to generate a quality detection unit. The second real-time branch is used to receive the output of the 3D reconstruction unit, and the first reference branch and the second real-time branch establish a lateral interaction channel.

[0084] Step M4: Connect the quality detection unit to the 3D reconstruction unit, perform sample-driven training until convergence, and obtain the quality detection module.

[0085] Specifically, the quality inspection module consists of a 3D reconstruction unit and a quality inspection unit connected together. The 3D reconstruction unit generates a 3D geometric model of the cigarette filter rod from 2D point cloud data, capturing complete geometric features and surface information to provide an actual 3D model of the filter rod for quality inspection. Its output is connected to the second real-time branch of the quality inspection unit. The quality inspection unit compares the 3D geometric model of the cigarette filter rod with standard data to complete the evaluation of the filter rod's geometric parameters and surface quality. The quality inspection unit includes a first reference branch and a second real-time branch. The first reference branch stores standard geometric and surface feature data as a benchmark for evaluating filter rod quality. The second real-time branch receives the real-time data output from the 3D reconstruction unit and compares it with the standard data from the first reference branch. The first reference branch and the second real-time branch are connected in parallel and have a lateral interaction channel, forming a dynamic comparison mechanism to compare and analyze the actual inspection results (the 3D model of the cigarette filter rod output by the 3D reconstruction unit). Through comparison and interaction, the quality inspection unit can obtain the quality inspection results of the filter rod.

[0086] The 3D reconstruction unit is constructed from point cloud feature recognition and reconstruction techniques based on two-dimensional relative planes. Point cloud feature recognition uses image processing algorithms (such as edge detection and contour extraction) to identify key features in the point cloud, such as the edges and centers of cigarette filters, and maps them to multiple two-dimensional planes. Reconstruction based on two-dimensional relative planes uses relative plane reconstruction algorithms (such as Poisson reconstruction or Delaunay triangulation) to stitch these two-dimensional plane point cloud data into a complete 3D model.

[0087] The quality inspection unit comprises a first reference branch and a second real-time branch. The first reference branch stores standard geometric and surface feature data as a benchmark for evaluating filter rod quality. The second real-time branch receives real-time data output from the 3D reconstruction unit and compares it with the standard data from the first reference branch. The first reference branch and the second real-time branch are connected in parallel and have a lateral interaction channel, forming a dynamic comparison mechanism to compare and analyze the actual inspection results (the 3D model of the cigarette filter rod output by the 3D reconstruction unit).

[0088] The construction process of the first reference branch is as follows: Standard geometric data and standard surface features of cigarette filter rods are collected from production specifications or quality standard manuals. The standard geometric data refers to the pre-defined geometric shape data of the cigarette filter rod under ideal conditions, including geometric parameters such as the length, diameter, and cross-sectional shape of the filter rod. These data represent the geometric characteristics that a filter rod meeting quality requirements should possess. Standard surface features refer to the surface condition of the cigarette filter rod under ideal conditions, such as surface flatness and smoothness. Then, these standard geometric data and standard surface features are integrated to construct a data structure and stored in a database or dedicated storage unit, forming the first reference branch.

[0089] Building the second real-time branch primarily involves establishing a data reception and processing channel. Using programming techniques or specialized inspection system software, an interface is created to receive the output data from the 3D reconstruction unit. This interface processes the 3D model data, converting it into a format comparable to the standard data in the first reference branch. A parallel connection is established between the first reference branch and the second real-time branch; that is, two parallel threads are created to process data from the first reference branch and the second real-time branch respectively. Simultaneously, a lateral interaction channel is established between the two branches, connecting them to form a complete quality inspection unit. For example, shared memory or message queues can be used to allow the two branches to exchange and adjust data in real-time as needed during the inspection process.

[0090] The quality detection unit is then connected to the 3D reconstruction unit to begin sample-driven training. First, a large amount of cigarette filter sample data is collected, including both qualified and unqualified samples. This sample data is then input into the quality detection unit, which compares and analyzes the data against standard data from the first reference branch and actual data from the second real-time branch to calculate the detection error. As sample data is continuously input and processed, the parameters in the quality detection unit (such as weights in the comparison algorithm) are adjusted until the detection error gradually decreases and converges, thus obtaining a high-precision and stable quality detection module. This module is then deployed to the control center of the quality detection equipment, ensuring fully automated quality control of cigarette filters.

[0091] The quality inspection module constructed through the above steps achieves precise detection from point cloud data acquisition on the filter rod surface to 3D reconstruction, and then to geometric parameters and surface quality. The 3D reconstruction unit lays the foundation for comprehensively capturing the geometric features of the filter rod; the parallel architecture of reference branch and real-time branch ensures dynamic comparison and efficient detection; sample-driven training further optimizes the detection accuracy and the model's generalization ability. The quality inspection module can significantly improve the accuracy of cigarette filter rod quality inspection, reduce the false negative and false positive rates, and provide important technical support for improving the stability of cigarette product quality.

[0092] Step S4: Based on the quality inspection results, output the target cigarette filter rod to the first sorting end and / or the second sorting section of the quality inspection equipment.

[0093] Specifically, the quality inspection results are categorized into two types: qualified and unqualified. If the quality inspection is qualified, the internal sorting mechanism of the quality inspection equipment will guide the filter rods to the first sorting end. For example, the conveying path of the filter rods can be changed through a mechanical or electromagnetic control device to direct them to the first sorting end. If the quality inspection result is unqualified, the same sorting mechanism will guide the filter rods to the second sorting end.

[0094] By outputting qualified and unqualified filter rods from different sorting ends, a closed-loop automated control of detection and rejection is achieved, ensuring that only qualified filter rods enter the next production process, thereby improving the overall quality of cigarette filter rods and reducing product quality problems caused by unqualified filter rods being mixed in.

[0095] Specifically, step S4 includes: Step S41: Based on the quality inspection results, mark the target cigarette filter rod with an electronic tag.

[0096] Step S42: At the output port of the quality inspection equipment, the target cigarette filter rod is controlled based on the first sorting end and the second sorting end by identifying the electronic tag. The first sorting end is the quality qualified output end, and the second sorting end is the quality unqualified rejection end.

[0097] Specifically, after obtaining the quality inspection results, specialized electronic tag writing devices or software modules are used to electronically tag the target cigarette filter. For example, using radio frequency identification (RFID) electronic tags, the quality inspection result information (such as "pass" or "fail" identification codes) is written into the storage area of ​​the electronic tag via an RFID reader. If a QR code electronic tag is used, the inspection result information is encoded into a QR code pattern using QR code generation software, and then printed or marked on the cigarette filter. Electronic tagging gives each cigarette filter its own quality status, facilitating subsequent identification and classification operations in the production process, and improving the automation and intelligence of the production process.

[0098] The quality inspection equipment has two output ports: a first sorting end and a second sorting end. The first sorting end outputs qualified cigarette filter rods, while the second sorting end outputs unqualified cigarette filter rods, removing them from the production process. At the output port of the quality inspection equipment, an electronic tag reader reads the electronic tags on the cigarette filter rods. Based on the quality inspection result information stored in the electronic tags, if the result is "qualified," the cigarette filter rod is guided to the first sorting end (qualified output end); if the result is "unqualified," the cigarette filter rod is guided to the second sorting end (unqualified rejection end). This process can be achieved through the coordinated operation of mechanical devices (such as conveyor belts, sorting pushers, etc.) and electronic control systems (such as PLC combined with sensors, etc.). For example, when the electronic tag reader reads a "qualified" tag, the electronic control system controls the conveyor belt to send the cigarette filter rod to the collection container at the first sorting end; when a "unqualified" tag is read, the electronic control system controls the sorting pusher to push the cigarette filter rod to the waste collection device at the second sorting end. By using electronic tag-based sorting control, qualified and unqualified cigarette filter rods can be separated accurately and efficiently, ensuring the stability of product quality on the production line, while improving production efficiency and reducing manual intervention and errors.

[0099] In summary, the intelligent control method for quality detection and rejection of cigarette filter rods provided in this application has the following technical effects: By setting the conveyor belt's transmission speed, moment of inertia, and sensor acquisition mode, stable transmission and accurate data acquisition of the filter rods during the inspection process were ensured. Next, point cloud feature recognition and 3D reconstruction technology were used to convert the 2D contour point cloud data into a high-precision 3D model, providing accurate geometric and surface feature information for subsequent quality assessment. During this process, by calculating the neighborhood point cloud vector and Gaussian curvature, the edge and side features of the filter rods were analyzed in detail, quickly identifying defective products that did not meet quality standards and terminating their inspection process, thus saving time and resources. For qualified filter rods, 3D point cloud conversion and surface feature identification were further performed based on relative planar relationships to generate a complete 3D reconstruction result, which was then transmitted to the second real-time branch of the quality inspection unit. Through a lateral interaction channel, the features in the real-time branch were mapped and compared with the standard features in the first reference branch. Finally, the results were compared according to the quality accuracy requirements to determine compliance, and the quality inspection results were output. In the final sorting stage, the filter rods were electronically tagged according to the quality inspection results, and automated sorting was achieved by recognizing the electronic tags, with qualified and defective products output through the first and second sorting ends, respectively.

[0100] Overall, the embodiments of this application realize intelligent control of cigarette filter rod quality detection and rejection through multi-level calculation and real-time judgment. It can accurately collect the geometric feature data of the filter rod, improve detection accuracy and reduce the rate of missed detection and false detection. At the same time, due to accurate detection and automated sorting, the quality of cigarette filter rods is improved, the stability of product quality is enhanced, and strong technical support is provided for quality management in the cigarette production process.

[0101] Please refer to Figure 4 Based on the same inventive concept, this application provides a control system for quality detection and rejection of cigarette filter rods. The control system includes: a detection mode setting unit 10, a scanning sampling unit 20, a quality detection unit 30, and a sorting output unit 40.

[0102] The detection mode setting unit 10 is used to set the target detection mode under the conveyor belt transmission process inside the quality detection equipment. The target detection mode includes transmission speed, transmission rotational inertia and sensor acquisition mode. The transmission rotational inertia is used for filter rod rotation control.

[0103] The scanning sampling unit 20 is used to perform target scanning sampling under target detection mode control as the target cigarette filter rod is transmitted, and to determine N two-dimensional contour point clouds.

[0104] The quality inspection unit 30 is used to transmit N two-dimensional contour point clouds to the quality inspection module, perform three-dimensional fitting reconstruction and geometric evaluation of the contour point clouds, and determine the quality inspection results.

[0105] The sorting output unit 40 is used to output the target cigarette filter rod to the first sorting end and / or the second sorting section of the quality inspection equipment according to the quality inspection results.

[0106] The detection mode setting unit 10 is also used to set the transmission speed of the conveyor belt; based on the transmission speed, with the length of the conveyor belt in the detection area as a constraint and the tangential periodic rotation of the cigarette filter rod as a requirement, the transmission rotational inertia is calculated; the conveyor belt in the detection area is uniformly configured with sensors, and the sampling detection sequence triggered by the sensors is calculated based on the transmission speed and the transmission rotational inertia as a reference, which serves as the sensor acquisition mode; the transmission speed, transmission rotational inertia, and sensor acquisition mode are integrated to determine the target detection mode.

[0107] The quality detection unit 30 is also used to identify the relative planar relationships of N two-dimensional contour point clouds; transmit the N two-dimensional contour point clouds to the three-dimensional reconstruction unit to identify the point cloud features of the target cigarette filter rod, wherein the point cloud features include ranging, position coordinates and surface features, and ranging is the spatial relationship between the sensor configuration position and the scanning point cloud position; based on the ranging, position coordinates and relative planar relationships, the three-dimensional position coordinates are converted and determined; based on the three-dimensional position coordinates, the three-dimensional contour point cloud is reconstructed; step S35: based on the surface features, the three-dimensional contour point cloud is identified to determine the three-dimensional reconstruction result.

[0108] The quality inspection unit 30 is also used to identify the two-dimensional point cloud at the edge position. Based on the distance measurement and position coordinates, it calculates the neighboring point cloud vector, where the neighboring point cloud is the first point cloud and the second point cloud that are adjacent. The conveyor belt plane is used as the reference plane, and the direction and angle of change of the second point cloud compared with the first point cloud are used as the point cloud vector. Based on the neighboring point cloud vector, the Gaussian curvature of each two-dimensional point cloud at the edge position is calculated. It is determined whether the Gaussian curvature is consistent, and the first judgment result is determined. Step S334: Identify the two-dimensional point cloud at the lateral position, perform line type determination of the point cloud distribution, and determine the second judgment result. Step S335: If the first judgment result is inconsistent, or the second judgment result is non-linear, either of these conditions is met. The target cigarette filter rod is marked as a defective product, and the quality inspection process is terminated.

[0109] The quality inspection unit 30 is also used to determine the three-dimensional reconstruction result by combining the relative planar relationship to perform three-dimensional point cloud conversion and surface feature identification if the first judgment result is consistent and the second judgment result is linear; step S337: transmit the three-dimensional reconstruction result to the second real-time branch of the quality inspection unit; perform feature mapping between the first reference branch and the second real-time branch based on the lateral interaction channel, compare the consistency of the contour geometric features and surface features, and determine the comparison result; interact with the quality accuracy of the cigarette filter rod, and make a qualified judgment based on the comparison result according to the quality accuracy, and output the quality inspection result.

[0110] The sorting output unit 40 is also used to mark the target cigarette filter rod with an electronic tag according to the quality inspection results; at the output port of the quality inspection equipment, the target cigarette filter rod is output controlled based on the first sorting end and the second sorting end by recognizing the electronic tag, wherein the first sorting end is the quality qualified output end and the second sorting end is the quality unqualified rejection end.

[0111] It is understood that the control system for quality detection and rejection of cigarette filter rods provided in this application corresponds to the control method for quality detection and rejection of cigarette filter rods provided in this application. For the sake of brevity, the same or similar parts can be referred to the content of the control method for quality detection and rejection of cigarette filter rods, and will not be repeated here.

[0112] The various modules in the aforementioned control device for quality inspection and rejection of cigarette filter rods can be implemented entirely or partially through software, hardware, or a combination thereof. These modules can be embedded in the processor of the server in hardware form or independent of it, or stored in the server's memory in software form, so that the processor can call and execute the corresponding operations of each module. The processor can be a central processing unit (CPU), a microprocessor, a microcontroller, etc.

[0113] Based on the same inventive concept, this application provides a computer-readable storage medium storing computer-readable instructions, which, when executed by a processor, implement the steps in the above-described control method for quality detection and rejection of cigarette filter rods.

[0114] The memory in this application embodiment can be volatile memory or non-volatile memory, or it can include both volatile and non-volatile memory. The non-volatile memory can be read-only memory (ROM), programmable read-only memory (PROM), erasable programmable read-only memory (EPROM), electrically erasable programmable read-only memory (EEPROM), or flash memory. The volatile memory can be random access memory (RAM), which is used as an external cache. By way of example, but not limitation, many forms of random access memory (RAM) are available, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate synchronous DRAM (DDR SDRAM), enhanced synchronous DRAM (ESDRAM), synchronous linked DRAM (SLDRAM), and direct rambus RAM (DR RAM).

[0115] The above embodiments can be implemented, in whole or in part, by software, hardware (such as circuits), firmware, or any other combination thereof. When implemented using software, the above embodiments can be implemented, in whole or in part, as a computer program product. A computer program product includes one or more computer instructions or computer programs.

[0116] When computer instructions or computer programs are loaded or executed on a computer, all or part of the processes or functions according to the embodiments of this application are generated. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, computer instructions can be transmitted from one website, computer, server, or data center to another website, computer, server, or data center via wired (e.g., infrared, wireless, microwave, etc.) means.

[0117] Computer-readable storage media can be any available medium that a computer can access, or a data storage device such as a server or data center that includes one or more sets of available media. Available media can be magnetic media (e.g., floppy disks, hard disks, magnetic tapes), optical media (e.g., DVDs), or semiconductor media. Semiconductor media can be solid-state drives (SSDs).

[0118] It should be understood that in the various embodiments of this application, the order of the above-mentioned processes does not imply the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of this application.

[0119] Those skilled in the art will recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0120] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working processes of the devices, apparatuses, and units described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here.

[0121] In addition, the functional units in the various embodiments of this application can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit.

[0122] If a function is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or a part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods of the various embodiments of this application. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0123] In this specification, references to "an embodiment" or "a specific implementation" mean that a particular feature, structure, or characteristic described in connection with that embodiment / specific implementation is included in at least one embodiment / specific implementation of the invention. Therefore, the phrase "in one embodiment / specific implementation" appearing in various places in this specification does not necessarily refer to the same embodiment / setting, but rather to potentially different embodiments. Furthermore, specific features, structures, or characteristics may be combined in one or more embodiments / settings in any suitable manner, as will be apparent to those skilled in the art from this disclosure.

[0124] Similarly, it should be understood that in the above description of exemplary embodiments / specific implementations of the invention, various features of the invention are sometimes combined in a single embodiment / specific implementation or its figures and description, with the aim of simplifying the disclosure and aiding in the understanding of one or more of the various aspects of the invention. However, the method of description in this patent should not be construed as reflecting an intention that the claimed features of the invention are more than those expressly stated in each claim, except where explicitly stated otherwise or in obvious technical contradiction or exclusion. Rather, the inventive aspect reflected in the claims lies in not all the features of a single foregoing disclosed embodiment / specific implementation. Therefore, the claims following the detailed description are expressly incorporated herein by reference, each claim existing independently as a separate embodiment / specific implementation of the invention.

[0125] Furthermore, while some embodiments / specific implementations described herein include, but are not limited to, other features included in other embodiments / specific implementations, combinations of features from different embodiments / specific implementations are intended to be within the scope of the invention and form different embodiments / specific implementations, as will be understood by those skilled in the art. For example, in the following claims, embodiments / specific implementations of any claim can be used in any combination.

[0126] The terms and expressions used in this specification are for illustrative purposes and not for limitation. In using these terms and expressions, it is not intended to exclude any equivalents of the features or portions thereof shown and described, but rather to recognize that various modifications may be possible within the scope of the invention.

[0127] Therefore, it should be understood that although the invention has been specifically disclosed through preferred embodiments, exemplary embodiments and optional features, those skilled in the art may take variations or modifications of the concepts disclosed herein, and such variations and modifications are therefore considered to be within the scope of the invention as defined by the appended claims.

[0128] The specific embodiments given in this specification are examples of useful implementations of the present invention. It will be apparent to those skilled in the art that the present invention can be implemented using many variations of the devices, device components, and method steps disclosed in this specification.

[0129] The foregoing description of specific embodiments fully discloses the general features of the present invention, enabling others to easily modify and / or adapt such specific embodiments for various applications by applying knowledge within the scope of the art, without conducting excessive experimentation and without departing from the general concept of the present invention.

[0130] Therefore, based on the teachings and guidance provided herein, it is intended that such modifications and alterations be included within the meaning and scope of equivalents of the disclosed embodiments. It should be understood that the wording or terminology used herein is for descriptive purposes and is not intended to be limiting; thus, the wording or terminology in this specification will be interpreted by those skilled in the art based on the foregoing teachings and guidance.

[0131] Furthermore, the scope of the invention should not be limited to any of the exemplary embodiments described above, but is defined solely by the appended claims and their equivalents.

Claims

1. A method for quality inspection and rejection of cigarette filter rods, characterized in that, The control method includes: Step S1: Set the target detection mode under the conveyor belt transmission process inside the quality inspection equipment. The target detection mode includes transmission speed, transmission rotational inertia and sensor acquisition mode. The transmission rotational inertia is used for filter rod rotation control. Step S2: As the target cigarette filter rod is transmitted, the target is scanned and sampled through the target detection mode control to determine N two-dimensional contour point clouds; Step S3: Transmit the N two-dimensional contour point clouds to the quality inspection module, perform three-dimensional fitting reconstruction and geometric evaluation of the contour point clouds, and determine the quality inspection results; Step S4: Based on the quality inspection results, output the target cigarette filter rod to the first sorting end and / or the second sorting section of the quality inspection equipment.

2. The method for quality inspection and rejection of cigarette filter rods according to claim 1, characterized in that, Step S1 includes: Step S11: Set the conveyor belt speed; Step S12: Based on the transmission speed, with the length of the conveyor belt in the detection area as a constraint and the tangential periodic rotation of the cigarette filter rod as a requirement, calculate the transmission rotational inertia; Step S13: Perform uniform sensor configuration on the conveyor belt in the detection area, and calculate the sampling and detection timing of the sensor trigger based on the transmission speed and the transmission rotational inertia, as the sensor acquisition mode. Step S14: Integrate the transmission speed, transmission moment of inertia, and sensor acquisition mode to determine the target detection mode.

3. The method for quality inspection and rejection of cigarette filter rods according to claim 1, characterized in that, The construction steps of the quality inspection module include: Step M1: Construct a 3D reconstruction unit using point cloud feature recognition and reconstruction based on 2D relative planes; Step M2: Construct the first reference branch based on the standard geometric data and standard surface features of the interactive cigarette filter rod; Step M3: Establish a parallel connection between the first reference branch and the second real-time branch to generate a quality detection unit, wherein the second real-time branch is used to receive the output of the 3D reconstruction unit, and the first reference branch and the second real-time branch establish a lateral interaction channel; Step M4: Connect the quality detection unit to the 3D reconstruction unit, perform sample-driven training until convergence, and obtain the quality detection module.

4. The method for quality inspection and rejection of cigarette filter rods according to claim 3, characterized in that, Step S3 includes: Step S31: The N two-dimensional contour point cloud identifiers have relative planar relationships; Step S32: Transmit the N two-dimensional contour point clouds to the three-dimensional reconstruction unit to identify the point cloud features of the target cigarette filter rod, wherein the point cloud features include ranging, position coordinates and surface features, and the ranging is the spatial relationship between the sensor configuration position and the scanning point cloud position. Step S33: Based on the distance measurement, position coordinates, and relative plane relationship, convert and determine the three-dimensional position coordinates; Step S34: Reconstruct the three-dimensional contour point cloud based on the three-dimensional position coordinates; Step S35: Based on the surface features, identify the three-dimensional contour point cloud and determine the three-dimensional reconstruction result.

5. The method for quality inspection and rejection of cigarette filter rods according to claim 4, characterized in that, Step S33 includes: Step S331: Identify the two-dimensional point cloud at the edge position, and calculate the neighboring point cloud vector based on the distance measurement and position coordinates. The neighboring point cloud is the first point cloud and the second point cloud that are adjacent to each other. The conveyor belt plane is used as the reference plane, and the direction and angle of change of the second point cloud relative to the first point cloud are used as the point cloud vector. Step S332: Based on the neighborhood point cloud vector, calculate the Gaussian curvature of each two-dimensional point cloud at the edge position; Step S333: Determine whether the Gaussian curvatures are consistent, and determine the first determination result; Step S334: Identify the 2D point cloud at the lateral position, determine the line type of the point cloud distribution, and confirm the second determination result: Step S335: If the first determination result is inconsistent and the second determination result is nonlinear, the target cigarette filter rod is marked as defective and the quality inspection process is terminated.

6. The method for controlling the quality inspection and rejection of cigarette filter rods according to claim 5, characterized in that, Step S33 further includes: Step S336: If the first determination result is consistent and the second determination result is linear, perform three-dimensional point cloud conversion and surface feature identification in combination with the relative planar relationship to determine the three-dimensional reconstruction result; Step S337: Transmit the three-dimensional reconstruction result to the second real-time branch of the quality inspection unit; Step S338: Based on the lateral interaction channel, perform feature mapping between the first reference branch and the second real-time branch, perform consistency comparison between contour geometric features and surface features, and determine the comparison result; Step S339: Analyze the quality accuracy of the cigarette filter rods. Based on the quality accuracy, determine the pass / fail status of the comparison results and output the quality test results.

7. The method for quality inspection and rejection of cigarette filter rods according to claim 1, characterized in that, Step S4 includes: Step S41: Based on the quality inspection results, mark the target cigarette filter rod with an electronic tag; Step S42: At the output port of the quality inspection equipment, by identifying the electronic tag, output control is performed on the target cigarette filter rod based on the first sorting end and the second sorting end, wherein the first sorting end is the quality qualified output end and the second sorting end is the quality unqualified rejection end.

8. A control system for quality detection and rejection of cigarette filter rods, characterized in that, The control system is used to execute the control method for quality detection and rejection of cigarette filter rods according to any one of claims 1-7. The control system includes: a detection module setting unit, a scanning sampling unit, a quality detection unit, and a sorting output unit. The detection module setting unit is used to set the target detection mode under the conveyor belt transmission process inside the quality detection equipment. The target detection mode includes transmission speed, transmission rotational inertia and sensor acquisition mode. The transmission rotational inertia is used for filter rod rotation control. The scanning sampling unit is used to perform target scanning sampling under the control of the target detection mode as the target cigarette filter rod is transmitted, and to determine N two-dimensional contour point clouds; The quality detection unit is used to transmit the N two-dimensional contour point clouds to the quality detection module, perform three-dimensional fitting reconstruction and geometric evaluation of the contour point clouds, and determine the quality detection results. The sorting output unit is used to output the target cigarette filter rod to the first sorting end and / or the second sorting section of the quality testing equipment according to the quality test results.

9. A computer-readable storage medium having a computer program / instructions stored thereon, characterized in that, When the computer program / instructions are executed by the processor, they implement the steps of the control method for quality detection and rejection of cigarette filter rods as described in any one of claims 1-7.

10. A computer program product comprising a computer program / instructions, characterized in that, When the computer program / instruction is executed by the processor, it implements the steps of the control method for quality detection and rejection of cigarette filter rods as described in any one of claims 1-7.

Images