Method and equipment for detecting bead blasting leakage in cigarette filter tip

Abstract

The invention relates to a method and equipment for detecting the leakage of a blasting bead in a cigarette filter tip, and the method comprises the steps: collecting the data information of a cigarette from different angles in the circumferential direction through CT equipment, and constructing a three-dimensional reconstruction model of the cigarette based on the data information; splitting the three-dimensional reconstruction model according to gray difference to obtain a blasting bead area, and forming a point cloud data set of the blasting bead area; performing sphere fitting on the point cloud data set of the bead blasting area, and calculating the diameter and sphericity index of a sphere after fitting; if the diameter of the fitted ball body is larger than a preset blasting bead diameter threshold value, it is determined that the blasting bead leaks; or if the sphericity degree index is smaller than the preset blast bead sphericity degree index threshold value, it is determined that the blast bead leaks, operation is easy, large-scale data operation is not needed, the data source is changed from two-dimensional data to three-dimensional data, data calculation is accurate, and abnormal blast bead detection is sensitive.

Description

TECHNICAL FIELD

[0001] The application relates to a method and a device for detecting leakage of a burst bead in a filter of a cigarette. BACKGROUND

[0002] Cigarette products with burst beads added in filter rods have occupied a large share in the cigarette market. The burst bead in a cigarette refers to a small bead added in the middle of a filter rod, and the burst bead contains various flavors of essence. When the burst bead is crushed, the taste and flavor of the cigarette can be improved. There are many flavors of burst beads in the market, which can provide different smoking experiences for consumers. Due to the special design of the burst bead cigarette, the quality of the burst bead has a direct impact on the smoking experience. The size of the burst bead directly affects the size of the essence diffusion area, which in turn affects the taste experience during smoking. The sphericity index of the burst bead is also an important factor affecting the quality of the burst bead. The regularity of the burst bead surface affects the diffusion of the burst bead when it is crushed. In addition, whether the burst bead leaks in the filter of the cigarette is also an important indicator for detecting the quality of the burst bead.

[0003] To solve the above problems, the patent publication No. CN119006457A discloses a burst bead defect detection method, device, electronic equipment and computer readable medium, wherein first, the feature map extraction and prior box detection processing are sequentially performed on the burst bead image set to generate a target burst bead anchor frame group; the target burst bead anchor frame group set is subjected to burst bead bubble defect detection; then, at least one to-be-detected cigarette burst bead image representing no burst bead bubble defect in the burst bead bubble detection result is subjected to contour extraction, geometric detection, determination of the burst bead inscribed circle area and the burst bead circumscribed circle area of each to-be-detected cigarette burst bead image, and based on the foregoing burst bead roundness defect detection on the at least one to-be-detected cigarette burst bead image; then, at least one burst bead contour representing no burst bead roundness defect in the burst bead roundness detection result is subjected to burst bead size defect detection based on the burst bead area, so as to control the burst bead rejection according to the burst bead bubble detection result, the burst bead roundness detection result and the burst bead size detection result, that is, the burst bead bubble detection is performed on each burst bead, the burst bead roundness detection is performed on the burst bead without the bubble defect, and the burst bead size monitoring is performed on the burst bead without the bubble defect and the roundness defect. Obviously, the burst bead with the roundness defect but without the bubble defect, the burst bead with the size defect but without the bubble / roundness defect, and the like still cannot be accurately rejected, resulting in burst bead leakage in the overall detection process and inaccurate rejection. Moreover, the related data of the burst bead need to be converted and processed in a large scale for each monitoring process, which brings great operation pressure to the monitoring equipment. In addition to ensuring the smooth completion of the detection, unlimited data processing of the burst bead is also required, which is easy to cause data redundancy and long-time inaccurate monitoring. SUMMARY

[0004] The application aims to provide a method and device for detecting leakage of burst beads in a filter of a cigarette, to solve the technical problems of complicated detection process, large data processing amount, and low detection result accuracy.

[0005] To achieve the above-mentioned purpose, in one aspect, the application provides a method for detecting leakage of burst beads in a filter of a cigarette, comprising:

[0006] acquiring data information of the cigarette from different angles by a CT device, and constructing a three-dimensional reconstruction model of the cigarette based on the data information;

[0007] splitting a burst bead region from the three-dimensional reconstruction model according to a gray difference to form a point cloud data set of the burst bead region;

[0008] performing sphere fitting on the point cloud data set of the burst bead region, and calculating a diameter and a sphericity index of the fitted sphere;

[0009] if the diameter of the fitted sphere is greater than a preset burst bead diameter threshold, it is determined that the burst bead has leaked; or, if the sphericity index is less than a preset burst bead sphericity index threshold, it is determined that the burst bead has leaked.

[0010] Further, the sphericity index is calculated by the following method: the point cloud data of the fitted sphere is processed by using an open3D open source library to obtain a corresponding surface area and volume; based on the surface area and volume of the fitted sphere, the sphericity index of the fitted sphere is calculated.

[0011] Among them, the surface area is calculated by using a triangulation mesh model algorithm; the volume is calculated by using a voxel grid algorithm.

[0012] Further, the sphericity index is calculated by the following formula:

[0013]

[0014] Among them, V is the volume of the burst bead; S is the surface area of the burst bead.

[0015] Further, the point cloud data set of the burst bead region is fitted by the following method:

[0016] selecting data points of a preset point number from the point cloud data set of the burst bead region, and performing sphere fitting on the selected data points;

[0017] judging whether the number of inliers of the fitted sphere is equal to the number of inliers of the point cloud data set of the burst bead region;

[0018] if yes, the fitted sphere of the burst bead region is obtained;

[0019] If not, the iterative algorithm is called to continue selecting unfitted data points from the point cloud dataset of the bursting bead area for sphere fitting until the number of inner points of the fitted sphere is equal to the number of inner points of the point cloud dataset of the bursting bead area.

[0020] Furthermore, the following method is used to split the bead-bursting area from the 3D reconstructed model according to the grayscale difference:

[0021] Based on the data information of the cigarette, training for splitting the bursting bead area according to grayscale differences is performed to generate a training model for extracting the bursting bead area; the training model is called on the three-dimensional reconstructed model to split the bursting bead area.

[0022] Furthermore, a filtered back projection algorithm is used to process the data information of the cigarette to generate a three-dimensional reconstructed model of the cigarette.

[0023] On the other hand, the present invention also proposes a device for detecting leakage of burst beads in cigarette filters, wherein the device includes a processor for executing the above-mentioned method for detecting leakage of burst beads in cigarette filters.

[0024] The beneficial effects of the present invention are as follows: using the data information of cigarettes collected from different circumferential angles by CT equipment, and constructing a three-dimensional reconstruction model of the cigarettes based on the data information; splitting the bursting bead area from the three-dimensional reconstruction model according to the grayscale difference to form a point cloud data set of the bursting bead area; performing sphere fitting on the point cloud data set of the bursting bead area, and calculating the diameter and sphericity index of the fitted sphere; if the diameter of the fitted sphere is greater than the preset bursting bead diameter threshold, it is determined that the bursting bead has leaked; or, if there is a sphericity index less than the preset bursting bead sphericity index threshold, it is determined that the bursting bead has leaked, thereby obtaining the three-dimensional data information of the cigarette and constructing a three-dimensional reconstruction model, and then combining the sphere fitting operation to obtain the best sphere model of the bursting bead, thereby improving the fitting accuracy of the sphere corresponding to the bursting bead; and cleverly using the judgment of the bursting bead diameter parameter and the sphericity index parameter to confirm whether the bursting bead has leaked, which is not only simple to operate and does not require large-scale data operations, but also the data source is broken through from two-dimensional data to three-dimensional data, so that the data calculation is accurate and the abnormal bursting bead detection is sensitive. BRIEF DESCRIPTION OF THE DRAWINGS

[0025] Figure 1 This is a flow chart of a method for detecting leakage of popping beads in cigarette filters proposed by the present invention;

[0026] Figure 2 This is a schematic diagram of the working of a CT device when collecting data information of a cigarette in an actual application scenario of a detection method for burst bead leakage in a cigarette filter proposed by the present invention;

[0027] Figure 3It is a kind of cigarette filter in the method for detecting the leakage of burst bead proposed by the application in actual application scene CT equipment in the principle diagram of scanning process of cigarette;

[0028] Figure 4 It is a kind of cigarette filter in the method for detecting the leakage of burst bead proposed by the application in actual application scene CT equipment in the principle diagram of scanning process of cigarette;

[0029] Reference signs:

[0030] 1-X-ray source;2-Loading platform;3-Sample base;4-Sample holder;5-Detection platform. DETAILED DESCRIPTION

[0031] In order to make the purpose, technical scheme and advantages of the present application more clear and obvious, the present application will be further described in detail below with reference to the drawings and examples.

[0032] The inventive concept of the present application is that: in order to facilitate and efficiently detect whether the burst bead leaks, a three-dimensional reconstruction model of cigarette is constructed, and the burst bead area is split and spherical fitting is carried out, and the diameter of the fitted sphere and the sphericity index are judged to determine whether the burst bead leaks.

[0033] Method embodiment 1:

[0034] As shown in Figure 1 It is a flowchart of the method for detecting the leakage of burst bead in cigarette filter proposed by the application, which includes steps S11-S14, specifically:

[0035] Before step S11, according to Figure 2 The data information of cigarette is collected, specifically, Figure 2A detection method for leakage of burst beads in a filter of a cigarette according to the present application is shown in the working schematic of the CT device in the actual application scene when collecting data information of the cigarette. The cigarette is fixed by using a sample holder 4, and the position of the cigarette is adjusted to ensure that the cigarette is perpendicular to the sample base 3; the industrial CT device is opened, the sample base 3 is placed on the bayonet of the sample platform 2 in the industrial CT device, the clamping device (sample base 3 and sample holder 4) is fixed on the sample platform 2 to avoid the sample from falling off when the sample platform 2 rotates; the X-ray source tube voltage of the X-ray source 1 is set to 100kv, the X-ray source tube current is set to 70μA, the scanning thickness is set to 0.004mm, the scanning interval is set to 0.004mm, the CT scanning mode is set to cone beam scanning, and the CT scanning mode is set to Normal scanning; the sample platform 2 is moved by the workbench so that the cigarette to be detected is located in the central position of the X-ray scanning range, the sample platform 2 is rotated to ensure that the cigarette is in the X-ray scanning position within a range of 360°; the cigarette is taken out, the CT device is air calibrated, then the central axis calibration rod is placed on the sample platform 2, the central axis is calibrated, and then the cigarette is placed on the sample platform 2, and the CT scanning is started to scan the cigarette; the data information received by the detection platform 5 is transmitted to the computer for storage, the collection of the data information of the cigarette is completed, and the comprehensive collection of the data information of the cigarette without damaging the cigarette is realized, thereby providing comprehensive and accurate data sources for subsequent analysis of the content of the cigarette.

[0036] After the data information of the cigarette is collected by the CT device, step S11 is executed, the data information of the cigarette collected from different angles in the circumferential direction by the CT device is used, and a three-dimensional reconstruction model of the cigarette is constructed based on the data information; here, the axial different angles refer to collecting the data information of the cigarette at each axial angle around the cigarette for one revolution, achieving the acquisition of the data information of the cigarette at all angles, and constructing the three-dimensional reconstruction model of the cigarette based on the comprehensive data information, thereby realizing the high restoration of the three-dimensional reconstruction model.

[0037] It should be noted that in actual application scenarios, the data information of the cigarette is processed by using a filtered back projection algorithm to generate a three-dimensional reconstruction model of the cigarette. When the filtered back projection algorithm (i.e., the FDK algorithm) is used for data information processing, the following steps are specifically implemented: first, the acquired data is preprocessed, wherein the image acquisition is performed according to a collection period of 1° per rotation, and finally 360 cigarette images can be acquired. The two-dimensional projection data collected at each angle is weighted to correct the cone beam. Second, the above corrected and weighted projection data is one-dimensionally filtered along the projection perpendicular to the flat panel detector. Third, the data of the second step is back-projected to reconstruct the image, wherein the filtered data is back-projected along the X-ray direction, and meanwhile, interpolation calculation is required for the Z-axis direction in the above back-projection reconstruction image due to the large interval of the Z-axis direction in the sampling process. In the present application, the bilinear interpolation method is used to complete the interpolation in the X-axis direction and the interpolation in the Y-axis direction. Thus, the three-dimensional reconstruction model of the cigarette can be constructed.

[0038] In step S12, the blasting bead region is split from the three-dimensional reconstruction model according to the gray difference to form a point cloud data set of the blasting bead region. Here, the blasting bead region and / or the filter region are split from the three-dimensional reconstruction model according to the gray difference by the following method: training of the blasting bead region splitting according to the gray difference based on the data information of the cigarette to generate a training model for extracting the blasting bead region; calling the training model on the three-dimensional reconstruction model to split the blasting bead region, and using multiple splitting training of the blasting bead region according to the gray difference to form an accurate splitting model, thereby improving the accuracy of splitting the blasting bead region from the three-dimensional reconstruction model and laying an accurate data foundation for subsequent data processing.

[0039] Specifically, the splitting training is performed by the following method: the computer image processing system is used to perform image segmentation processing on the images containing the blasting beads in the collected two-dimensional sequence images. The blasting beads and the filter are marked in the three-view interface of the image processing software according to the different gray values of the blasting beads and the filter (to ensure the accuracy of the classification, multiple two-dimensional images need to be marked in this process). When the blasting bead region is split on the three-dimensional reconstruction model, an ROI (Region of Interest) based on the gray value is created in the image processing system, the filter region is separately split out of the obtained three-dimensional reconstruction model, and a new ROI is formed. Then, a new ROI for extracting the blasting bead region is formed in the ROI of the filter region; and the blasting bead region is split out of the filter region according to the new ROI.

[0040] In step S13, the point cloud data set of the burst ball region is subjected to sphere fitting, and the diameter and sphericity index of the fitted sphere are calculated. It should be noted that the sphericity index refers to a parameter for characterizing whether the shape of the burst ball in the filter is round. In actual application scenarios, the sphericity index can be calculated according to various equations of the sphericity index. The point cloud data set of the burst ball region is subjected to sphere fitting by the following method: a preset number of data points are selected from the point cloud data set of the burst ball region, and the selected data points are subjected to sphere fitting. It is determined whether the number of inliers of the fitted sphere is equal to the number of inliers of the point cloud data set of the burst ball region. If yes, the fitted sphere of the burst ball region is obtained. If no, an iterative algorithm is called to continue selecting un-fitted data points from the point cloud data set of the burst ball region for sphere fitting until the number of inliers of the fitted sphere is equal to the number of inliers of the point cloud data set of the burst ball region. Here, the preset number of data points refers to the number of basic data points required for fitting a sphere. In actual application scenarios, the number of data points is set according to the requirements of different cigarettes for burst balls. In the preferred embodiments of the present application, the number of data points is preferably 4.

[0041] At the same time, in actual application scenarios, the fitting parameters of the fitted sphere are also calculated. When the fitting parameters of the fitted sphere reach the preset fitting parameter threshold, the best sphere model is obtained. The fitting parameter threshold refers to a parameter for measuring whether the best sphere model is obtained in the sphere fitting process, so as to ensure that all inliers contained in the fitted sphere are integrated in each sphere fitting process, avoiding the occurrence of inlier fitting misplacement or inlier non-fitting.

[0042] In step S14, if the diameter of the fitted sphere is greater than the preset burst ball diameter threshold, it is determined that the burst ball has leaked. Or, if the sphericity index is less than the preset burst ball sphericity index threshold, it is determined that the burst ball has leaked. Here, the preset burst ball diameter threshold can be set as the diameter value of the burst ball required by the cigarette, or as the diameter value of the non-leaked burst ball. Similarly, the preset sphericity index threshold can be set as the sphericity index of the burst ball required by the cigarette, or as the sphericity index of the non-leaked burst ball. At the same time, it should be noted that when the burst ball leaks, due to its initial spherical shape, the internal components of the burst ball will spread to the surrounding, causing the diameter of the burst ball to increase, so the leakage of the burst ball can be directly detected.

[0043] Through steps S11-S14, the burst ball region is split from the three-dimensional reconstruction model of the cigarette constructed, and the point cloud data set of the burst ball region is subjected to sphere fitting, so as to calculate and determine the diameter and sphericity index of the fitted sphere.

[0044] In a preferred embodiment of the present application, the cigarette A to be detected is arranged on a sample holder in the CT device, and data information of the cigarette A is collected from each axial angle through scanning of X-rays in the CT device; and a three-dimensional reconstruction model Model of the cigarette A is constructed based on the data of the cigarette A using the FDK algorithm. According to a pre-trained training model for splitting according to grayscale differences, the blasting bead region and the filter region are split from the three-dimensional reconstruction model Model to form a point cloud data set M1 of the blasting bead region, and a sphere fitting is performed on the power data set M1 of the blasting bead region. When the number of inliers of the fitted sphere is equal to the number of inliers of the point cloud data set M1 of the blasting bead region, the best sphere model of the fitted sphere is obtained, and the diameter d and the sphericity index q of the fitted sphere are calculated. The diameter d is compared with the diameter D' (a pre-set blasting bead diameter threshold), and the sphericity index q is compared with the sphericity index Q'. When d>D', or when qQ', the blasting bead in the filter of the cigarette A has a leakage phenomenon.

[0045] Method embodiment 2:

[0046] In the method for detecting leakage of a blasting bead in a filter of a cigarette provided in the present application, the sphericity index is calculated by the following method: the point cloud data of the fitted sphere is processed using the open3D open source library to obtain the corresponding surface area and volume; and the sphericity index of the fitted sphere is calculated based on the surface area and the volume of the fitted sphere. Here, the open3D open source library is a high-efficiency and easy-to-use tool for 3D data processing.

[0047] Then, according to the above embodiments of the present application, the following operations are performed on the fitted sphere: the surface area of the fitted sphere is calculated using the Open3D open source library, and preferably a triangulation network model algorithm is used to calculate the surface area. Specifically, the point cloud data of the fitted sphere is triangulated to obtain a triangulation mesh model; the triangulation mesh model is converted into a point cloud model; and the surface area S of the fitted sphere is calculated using the built-in method of Open3D. Preferably, a speed-up mesh algorithm is used to calculate the volume. Specifically, a voxel mesh is created, and the point cloud data of the fitted sphere is converted into a voxel mesh through voxelization operation; the diameter of the voxel mesh is calculated using the built-in method of Open3D; and the volume of the voxel mesh, i.e., the volume V of the fitted sphere, is further calculated based on the diameter of the voxel mesh. According to the surface area S and the volume V of the fitted sphere, the sphericity index formula is called to obtain the sphericity index q of the blasting bead corresponding to the fitted sphere.

[0048] Meanwhile, in actual application scenarios, the sphericity index is calculated by the following formula:

[0049]

[0050] wherein V is the volume of the burstable balloon; and S is the surface area of the burstable balloon.

[0051] Method embodiment 3:

[0052] In the method for detecting leakage of a burstable balloon in a filter of a cigarette rod provided by the present application, whether the burstable balloon leaks or not is monitored by performing the following steps: S1, collecting data information of a cigarette rod sample by using a CT device; S2, processing the obtained cigarette rod data by using a relevant algorithm (such as a filtered back projection algorithm (FDK)); S3, performing image segmentation on the burstable balloon and the filter rod; and S4, processing point cloud data. Step S4 specifically includes S41-S44.

[0053] S41, extracting the burstable balloon region from a three-dimensional reconstruction model of the cigarette rod sample to form a point cloud data set.

[0054] S42, processing the point cloud data of the burstable balloon portion to calculate the diameter of the burstable balloon. Specifically, a least square method is used to perform a sphere fitting operation on the point cloud data points in the point cloud data set of the burstable balloon region. The key steps are as follows:

[0055] The sphere equation is:

[0056] (x-a) 2 +(y-b) 2 +(z-c) 2 =r 2

[0057] The above formula can be expanded and arranged as follows:

[0058] -2xa-2yb-2zc+1*(a 2 +b 2 +c 2 -r 2 )=-x 2 -y 2 -z 2

[0059] Let d=(a 2 +b 2 +c 2 -r 2 ):

[0060] -2xa-2yb-2zc+1*d=-x 2 -y 2 -z 2

[0061] Using an iterative algorithm, the coordinates of 4 points in the point cloud data set of the burst bead are randomly selected and substituted into the above formula to fit the sphere. For the fitted sphere, the number of included points is calculated, and the number of included points is compared with the number of included points in the burst bead point cloud data set. When the number of included points is the same as the number of included points in the burst bead point cloud data set, the iteration ends. The obtained sphere model is the best model. The center coordinates R_0(x, y, z) of the sphere and the radius R of the burst bead fitting sphere can be obtained from the best model of the previous sphere.

[0062] S43, calculate the burst bead sphericity index. Define Q (burst bead sphericity index) to represent whether the shape of the burst bead in the filter rod is round. Its calculation formula is as follows:

[0063]

[0064] Where, V represents the volume of the burst bead; S represents the surface area of the burst bead.

[0065] Wherein, the above-mentioned sphericity index is calculated by the following content.

[0066] The open source library is used to calculate the surface area of the burst bead point cloud data. Specifically, the point cloud data of the fitted sphere is triangulated to obtain a triangulation mesh model; the triangulation mesh model is converted into a point cloud model; and the built-in method of Open3D is used to calculate the surface area S of the fitted sphere.

[0067] The open source library is used to calculate the volume of the burst bead point cloud data. Specifically, a voxel grid is created, and the point cloud data of the fitted sphere is converted into a voxel grid through voxelization operation; the built-in method of Open3D is used to calculate the diameter of the voxel grid; and the diameter of the voxel grid is further calculated to obtain the volume of the voxel grid, i.e. the volume V of the fitted sphere.

[0068] S44, determine whether the burst bead leaks. For whether the burst bead leaks, there are two cases.

[0069] The first case is that the burst bead is accidentally crushed during production, and the essence in the burst bead is completely leaked. In this case, the sphere range of essence diffusion will be much larger than the diameter of the burst bead itself. Therefore, the diameter threshold of the burst bead is set to judge the diameter of the burst bead. When the diameter of the fitted sphere of the detected burst bead exceeds the threshold of the diameter of the burst bead, it is determined that the burst bead leaks.

[0070] The second scenario involves the formation of cavities on the bead surface, leading to partial leakage of the flavoring inside. In this case, the diffusion of the flavoring can cause the point cloud model of the bead obtained by CT scanning to exhibit irregular shapes. Therefore, a bead sphericity index threshold can be set to determine the sphericity index of the fitted sphere of the tested bead. When the sphericity index of the fitted sphere of the tested bead falls below the threshold, it is considered that the bead has partially leaked.

[0071] Method Example 4:

[0072] like Figure 3 The figure shows a schematic diagram of the principle of a method for detecting burst bead leakage in a cigarette filter proposed by the present invention during a CT scan of a cigarette in an actual application scenario, wherein the cigarette to be detected is set at point O, and images are collected by X-rays with an acquisition cycle of 1° rotation. The two-dimensional projection data collected at each angle are weighted using a virtual detector to achieve the purpose of correcting the cone beam; the corrected weighted projection data are one-dimensionally filtered using a flat-panel detector, thereby collecting cigarette data information at all angles and constructing a three-dimensional reconstructed model of the cigarette.

[0073] Method Example 5:

[0074] like Figure 4 The figure shows a schematic diagram of the bursting bead measurement results at different viewing angles of the same bursting bead three-dimensional reconstruction model in an actual application scenario using a method for detecting bursting bead leakage in a cigarette filter proposed by the present invention, wherein the upper left is a top-down viewing angle of the bursting bead of the three-dimensional reconstructed model, the upper right is a left-viewing viewing angle of the bursting bead of the three-dimensional reconstructed model, the lower left is a frontal viewing angle of the bursting bead of the three-dimensional reconstructed model, and the lower right is a stereoscopic viewing angle of the bursting bead of the three-dimensional reconstructed model, and the caliper data at each viewing angle represents the length from the center of the bursting bead on the cross section where the center of the bursting bead is located to the edge of the cigarette paper.

[0075] Device Example:

[0076] The present invention also proposes a device for detecting leakage of burst beads in cigarette filters, which includes a processor for executing the above-mentioned method for detecting leakage of burst beads in cigarette filters. Here, for the embodiments of the detection device, please refer to the embodiments (1-5) of the above-mentioned method for detecting leakage of burst beads in cigarette filters, which will not be repeated here.

[0077] In summary, the detection equipment for leakage of the burst ball in the filter tip of a cigarette provided by the application utilizes the industrial CT technology to scan the filter tip of the cigarette containing the burst ball, and obtains the three-dimensional reconstruction model of the sample. The filter tip part and the burst ball part are divided based on the gray value, so as to obtain the point cloud data set of the burst ball part respectively. Through the processing of the point cloud data set, the quality parameters such as the diameter size of the burst ball, the burst ball sphericity index and whether the burst ball leaks are accurately measured, the application solves the problems of low efficiency and insufficient precision in the current burst ball quality detection process of the cigarette, improves the detection precision and efficiency, can bring more economic benefits, at the same time, the mechanical equipment is used to replace the manual detection, solves the problem that the workers need to be exposed to strong light for a long time when detecting the burst ball quality, reduces the harm to the workers' bodies.

Claims

1. A method for detecting leakage of popping beads in cigarette filters, characterized in that: include: Using CT equipment to collect data information of the cigarette from different circumferential angles, and constructing a three-dimensional reconstruction model of the cigarette based on the data information; The bursting bead area is obtained by splitting the 3D reconstructed model according to the grayscale difference, forming a point cloud dataset of the bursting bead area; Perform sphere fitting on the point cloud dataset of the bead-bursting area and calculate the diameter and sphericity index of the fitted sphere; If the diameter of the fitted sphere is greater than the preset bursting bead diameter threshold, it is determined that the bursting bead is leaking; or if the sphericity index is less than the preset bursting bead sphericity index threshold, it is determined that the bursting bead is leaking.

2. The method for detecting leakage of popping beads in cigarette filters according to claim 1, characterized in that: The sphericity index is calculated by the following method: the point cloud data of the fitted sphere is processed using the open3D open source library to obtain the corresponding surface area and volume; based on the surface area and volume of the fitted sphere, the sphericity index of the fitted sphere is calculated; Among them, the triangulated mesh model algorithm is used to calculate the surface area, and the voxel mesh algorithm is used to calculate the volume.

3. The method for detecting leakage of popping beads in cigarette filters according to claim 2, characterized in that: The sphericity index is calculated using the following formula: ； Wherein, V is the volume of the bursting bead; S is the surface area of ​​the bursting bead.

4. The method for detecting leakage of popping beads in cigarette filters according to any one of claims 1 to 3, characterized in that: Perform sphere fitting on the point cloud dataset of the burst bead area using the following method: Select a preset number of data points from the point cloud data set of the bead bursting area, and perform sphere fitting on the selected data points; Determine whether the number of inner points of the fitted sphere is equal to the number of inner points of the point cloud dataset in the bead bursting area; If so, the fitted sphere of the burst bead region is obtained; If not, the iterative algorithm is called to continue selecting unfitted data points from the point cloud dataset of the bursting bead area for sphere fitting until the number of inner points of the fitted sphere is equal to the number of inner points of the point cloud dataset of the bursting bead area.

5. The method for detecting leakage of popping beads in cigarette filters according to any one of claims 1 to 3, characterized in that: The following method is used to split the 3D reconstructed model according to the grayscale difference to obtain the burst bead area: Based on the data information of the cigarette, training for splitting the bursting bead area according to grayscale differences is performed to generate a training model for extracting the bursting bead area; the training model is called on the three-dimensional reconstructed model to split the bursting bead area.

6. The method for detecting leakage of popping beads in cigarette filters according to any one of claims 1 to 3, characterized in that: The data information of the cigarette is processed using a filtered back projection algorithm to generate a three-dimensional reconstructed model of the cigarette.

7. A device for detecting leakage of popping beads in cigarette filters, characterized in that: The device comprises a processor for executing the method for detecting leakage of popping beads in a cigarette filter according to any one of claims 1 to 6.

Images