Tobacco material air separation and impurity removal process control method and device, and storage medium

Abstract

The present disclosure relates to a tobacco material air separation and impurity removal process control method, device, and storage medium. The tobacco material air separation and impurity removal process control method comprises: using spectral measurement and analysis technology to collect spectral images of the material in the air separation box during the air separation and impurity removal process, wherein the material spectral image includes material distribution status data information; performing data preprocessing and feature extraction on the material spectral image; using a pre-established component determination detection model to determine whether the material in the air separation box contains impurities; and adjusting the air volume by changing the frequency of the air separation fan according to whether the material in the air separation box contains impurities, thereby changing the impurity removal strength. The present disclosure can control the air separation and impurity removal of tobacco shreds (or tobacco sheets) through spectral imaging technology.

Description

Technical Field

[0001] The present disclosure relates to the field of wireless communications, and in particular to a tobacco material air separation and impurity removal process control method and device, and a storage medium. Background Art

[0002] Multiple processes or processing links in the cigarette processing process involve the air separation and impurity removal of materials. Through the air separation and impurity removal process, non-tobacco substances can be removed, and stems (or stem sticks) and wet clumps in tobacco materials can be removed to a certain extent, reducing the stem and stick removal pressure of the cigarette making machine, improving the quality stability of cigarette products, and reducing the quality risk of mold and mildew in rolled cigarettes. Summary of the Invention

[0003] Through research, the inventors found that many cigarette companies have set up an air separation process after the leaf shreds drying process in the related technology. The air separation equipment in the related technology is mostly based on the experience of the operators. After debugging and determining the air separation ratio, a fixed air separation frequency or air door opening is used for air separation. This method lacks the real-time removal of debris and wet clumps.

[0004] In view of at least one of the above technical problems, the present disclosure provides a tobacco material air separation and impurity removal process control method and device, and storage medium, which can control the air separation and impurity removal of tobacco (or tobacco sheets) through spectral imaging technology.

[0005] According to one aspect of the present disclosure, a method for controlling a tobacco material air separation and impurity removal process is provided, comprising:

[0006] Using spectral measurement and analysis technology, collect the material spectrum image in the air separation box during the air separation and impurity removal process, wherein the material spectrum image includes material distribution status data information;

[0007] Performing data preprocessing and feature extraction on the material spectral image;

[0008] Use the pre-established component determination detection model to determine whether there are foreign objects in the material in the air separation box;

[0009] According to whether there are foreign objects in the air separation box, the air volume is adjusted by changing the frequency of the air separation fan to change the strength of removing foreign objects.

[0010] In some embodiments of the present disclosure, the tobacco material air separation and impurity removal process control method further includes:

[0011] Using the pre-established component determination detection model, the spectral image of the material in the air separation box is identified, the material image distribution information is analyzed, and whether the material distribution is uniform is determined;

[0012] According to the uniformity of material distribution, the air distribution mechanism in the air separation box is used to adjust the flow field, change the air flow distribution state in the air separation box, and then adjust the material distribution in the air separation box.

[0013] In some embodiments of the present disclosure, identifying the spectral image of the material in the winnowing box, analyzing the material image distribution information, and determining whether the material distribution is uniform includes:

[0014] Analyze the spectral image of the material in the air separation box, analyze the material image distribution information, and the material concentration per unit area;

[0015] According to the material concentration distribution information, the flow field distribution status in the air separation box is identified;

[0016] The uniformity of material distribution is determined by the coefficient of variation of the material concentration ratio after binarization per unit area.

[0017] In some embodiments of the present disclosure, adjusting the flow field by using the air distribution mechanism in the air separation box according to the uniformity of material distribution includes:

[0018] When the coefficient of variation of the material concentration ratio is greater than a predetermined threshold, the air distribution mechanism is controlled to adjust the flow field by adjusting the orifice openings in different areas of the air distribution plate.

[0019] In some embodiments of the present disclosure, the debris includes at least one of tobacco stems, stem sticks, wet lumps, fibers, plastics, and metals.

[0020] In some embodiments of the present disclosure, adjusting the air volume by changing the frequency of the air separation fan to change the debris removal intensity includes:

[0021] The force of removing debris is changed by adjusting the air volume by the weight of the debris itself and by changing the frequency of the air separation fan. The air volume adjustment by changing the frequency of the air separation fan includes: reducing the fan frequency or increasing the damper opening to adjust the air separation wind speed.

[0022] In some embodiments of the present disclosure, the performing data preprocessing and feature extraction on the material spectral image includes: preprocessing the material spectral image to obtain normalized two-dimensional image data.

[0023] In some embodiments of the present disclosure, the use of a pre-established component determination detection model to determine whether there are foreign objects in the material in the air separation box includes: calculating a mean spectral vector based on two-dimensional image data; determining a target spectral vector based on the two-dimensional image data and the mean spectral vector; determining a response graph of the entire image based on the target spectral vector and the covariance matrix; comparing the response graph of the entire image with a predetermined threshold to determine a binary graph of the detection result; and determining whether there are foreign objects in the material in the air separation box based on the binary graph of the detection result.

[0024] In some embodiments of the present disclosure, performing data preprocessing and feature extraction on the material spectral image includes:

[0025] Perform black and white frame correction on the collected spectral images;

[0026] Filter the corrected reflectance data of the black and white frames.

[0027] In some embodiments of the present disclosure, the tobacco material air separation and impurity removal process control method further includes:

[0028] A component determination detection model is established in advance, wherein the component determination detection model is an unsupervised model, a semi-supervised model or a supervised model.

[0029] In some embodiments of the present disclosure, when the component determination detection model is a supervised model, pre-establishing the component determination detection model includes:

[0030] Hyperspectral measurement technology is used to collect data on debris during the air separation and impurity removal process;

[0031] Spectral measurement and analysis technology is used to collect data on the material distribution status in the air separation box during the air separation and impurity removal process;

[0032] Perform data preprocessing and feature extraction on debris data information and material distribution status data information;

[0033] Establish a component determination and detection model in advance.

[0034] According to another aspect of the present disclosure, a tobacco material air separation and impurity removal process control device is provided, comprising:

[0035] An image acquisition module is configured to use spectral measurement and analysis technology to acquire a spectral image of the material in the air separation box during the air separation and impurity removal process, wherein the spectral image of the material includes data information on the material distribution state;

[0036] a preprocessing module, configured to perform data preprocessing and feature extraction on the material spectral image;

[0037] The debris identification module is configured to use a pre-established component determination detection model to determine whether there are debris in the material in the winnowing box;

[0038] The rejection control module is configured to adjust the air volume and change the strength of the debris rejection by changing the frequency of the air separation fan according to whether there are debris in the material in the air separation box.

[0039] According to another aspect of the present disclosure, a tobacco material air separation and impurity removal process control device is provided, comprising:

[0040] a memory configured to store instructions;

[0041] The processor is configured to execute the instructions so that the tobacco material air separation and impurity removal process control device performs the operation of the tobacco material air separation and impurity removal process control method as described in any of the above embodiments.

[0042] According to another aspect of the present disclosure, a computer-readable storage medium is provided, wherein the computer-readable storage medium stores computer instructions, and when the instructions are executed by a processor, the method described in any of the above embodiments is implemented.

[0043] The present invention can control the air separation and impurity removal of tobacco shreds (or tobacco sheets) through spectral imaging technology. BRIEF DESCRIPTION OF THE DRAWINGS

[0044] In order to more clearly illustrate the embodiments of the present disclosure or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are only some embodiments of the present disclosure. For ordinary technicians in this field, other drawings can be obtained based on these drawings without any creative work.

[0045] Figure 1 Schematic diagram of some embodiments of the tobacco material air separation and impurity removal process control method disclosed in the present invention.

[0046] Figure 2 Schematic diagrams of other embodiments of the tobacco material air separation and impurity removal process control method disclosed in the present invention.

[0047] Figure 3 The embodiment is a schematic diagram of images collected by data acquisition in some embodiments of the present disclosure.

[0048] Figure 4 Schematic diagram of an image after filtering and denoising in some embodiments of the present disclosure.

[0049] Figure 5 Schematic diagram of an image after background separation and morphological processing in some embodiments of the present disclosure.

[0050] Figure 6 Schematic diagram of an image after instance segmentation in some embodiments of the present disclosure.

[0051] Figure 7 Schematic diagram of some embodiments of the target detection algorithm disclosed in the present invention.

[0052] Figure 8 Schematic diagram of detection results and binarized images in some embodiments of the present disclosure.

[0053] Figure 9Schematic diagram of some embodiments of the tobacco material air separation and impurity removal process control device disclosed in the present invention.

[0054] Figure 10 Schematic diagram of the structure of some embodiments of the tobacco material air separation and impurity removal process control device disclosed in the present invention. DETAILED DESCRIPTION

[0055] The following will be combined with the drawings in the embodiments of the present disclosure to clearly and completely describe the technical solutions in the embodiments of the present disclosure. Obviously, the described embodiments are only part of the embodiments of the present disclosure, rather than all the embodiments. The following description of at least one exemplary embodiment is actually only illustrative and is in no way intended to limit the present disclosure and its application or use. Based on the embodiments in the present disclosure, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of the present disclosure.

[0056] Unless specifically stated otherwise, the relative arrangement of components and steps, the numerical expressions and numerical values ​​set forth in these embodiments do not limit the scope of the present disclosure.

[0057] At the same time, it should be understood that for the convenience of description, the sizes of the various parts shown in the drawings are not drawn according to the actual proportional relationship.

[0058] Technologies, methods and equipment known to ordinary technicians in the relevant art may not be discussed in detail, but where appropriate, such technologies, methods and equipment should be considered part of the authorization specification.

[0059] In all examples shown and discussed herein, any specific values ​​should be interpreted as merely exemplary and not limiting. Therefore, other examples of the exemplary embodiments may have different values.

[0060] It should be noted that like reference numerals and letters refer to like items in the following figures, and therefore, once an item is defined in one figure, it need not be further discussed in subsequent figures.

[0061] Through research, the inventors found that most of the air separation equipment in related technologies is based on the experience of the operators. After debugging and determining the air separation ratio, a fixed air separation frequency or air door opening is used for air separation. This method lacks real-time removal of debris and wet clumps, and process monitoring and control in the air separation box.

[0062] In view of at least one of the above technical problems, the present disclosure provides a tobacco material air separation and impurity removal process control method and device, and a storage medium. The present disclosure is described below through specific embodiments.

[0063] Figure 1Schematic diagrams of some embodiments of the tobacco material air separation and impurity removal process control method disclosed herein. Preferably, this embodiment can be performed by the tobacco material air separation and impurity removal process control device disclosed herein. The method includes at least one of steps 11 to 14, wherein:

[0064] Step 11: Using spectral measurement and analysis technology, collect the material spectrum image in the air separation box during the air separation and impurity removal process, wherein the material spectrum image includes material distribution status data information.

[0065] The spectral measurement and analysis technology employed in this disclosure is based on spectral imaging, which can be used to visualize surface information within biological objects. Spectroscopy combines the advantages of both spectral technology and image processing. Its non-destructive, pollution-free, sample pretreatment-free, rapid, efficient, and highly accurate detection process has made hyperspectral technology a research hotspot in the fields of food, pharmaceuticals, and agricultural products.

[0066] In some embodiments of the present disclosure, hyperspectral images have rich spectral information.

[0067] Step 12: performing data preprocessing and feature extraction on the material spectrum image.

[0068] Step 13: Use the pre-established component determination detection model to determine whether there are foreign objects in the material in the air separation box.

[0069] In some embodiments of the present disclosure, step 13 may include: employing hyperspectral image target detection, utilizing known target prior information to distinguish different substances in the hyperspectral image, and identifying a relatively small number of target pixels from a majority of background pixels. Based on the spectral information of the target of interest, the possible presence of the target in the scene can be detected. Binary hypotheses of different background distributions are established for both background-only and target-presence scenarios, and a target detection model is constructed based on the generalized likelihood ratio.

[0070] Step 14: According to whether there are foreign objects in the material of the air separation box, the air volume is adjusted by changing the frequency of the air separation fan to change the strength of removing foreign objects.

[0071] The purpose of the above-mentioned embodiments of the present disclosure is precisely to address the current state of the related art and to provide an apparatus and method for controlling the air separation and impurity removal process in tobacco processing using spectral imaging technology. The above-mentioned embodiments of the present disclosure use spectral imaging technology to detect impurities and wet clumps during the air separation process of cigarettes, thereby controlling their removal.

[0072] Figure 2Schematic diagrams of other embodiments of the tobacco material air separation and impurity removal process control method disclosed herein. Preferably, this embodiment can be performed by the tobacco material air separation and impurity removal process control device disclosed herein. The method includes at least one of steps 10 to 16, wherein:

[0073] Step 10: Data collection and processing.

[0074] In some embodiments of the present disclosure, step 10 may include: pre-establishing a component determination detection model, wherein the component determination detection model is an unsupervised model (which may include step 104), a semi-supervised model (which may include step 104) or a supervised model (which may include steps 101 to 104).

[0075] In some embodiments of the present disclosure, when the component determination detection model is a supervised model, step 10 may include at least one of steps 101 to 104, wherein:

[0076] Step 101: using hyperspectral measurement technology to collect data information on debris during the air separation and impurity removal process.

[0077] In some embodiments of the present disclosure, step 101 may include: using hyperspectral measurement technology to collect rejects and reject data information involved in the air separation and impurity removal process.

[0078] In some embodiments of the present disclosure, the rejected objects (ie, debris) may include at least one of tobacco stems (or stem sticks), wet clumps, fibers, plastics, metals, and the like.

[0079] In some embodiments of the present disclosure, hyperspectral images contain rich spectral information. Hyperspectral image target detection is the process of distinguishing different substances in the hyperspectral image using known target prior information, identifying target pixels that account for a relatively small proportion of the total number of background pixels. Based on the spectral information of the target of interest, the present disclosure can detect the possible presence of the target in the scene. It establishes binary hypotheses for different background distributions in the background-only and target-presence scenarios, and constructs a target detection model based on the generalized likelihood ratio.

[0080] In some embodiments of the present disclosure, hyperspectral technology can be roughly understood as hyperspectral camera photography. The difference between the present disclosure's hyperspectral technology and camera photography lies in that, while cameras only have three channels—red, green, and blue—hyperspectral cameras can capture data from hundreds of channels across the spectral range from 400 nm to 2500 nm. This provides richer data and target information, enabling identification not only of external morphology but also of internal substances on the surface. Camera data is formatted similarly to X*Y*3 (where X*Y represents resolution), while hyperspectral data is formatted similarly to X*Y*Z (where X*Y represents spatial resolution and Z represents spectral resolution).

[0081] In some embodiments of the present disclosure, step 101 may include: extracting data information on impurities, tobacco stems, stem sticks, and wet clumps in tobacco material.

[0082] Step 102 : Using spectral measurement and analysis technology, data information on the distribution state (distribution uniformity) of the materials in the air separation box during the air separation and impurity removal process is collected.

[0083] In some embodiments of the present disclosure, the material distribution status data information is presented in an image form, and an image of the material in the field of view can be obtained. Data collection is performed based on the distribution state of the material in the fluidized state (gas-solid two-phase, which can be understood as the wind blowing the tobacco particles up) in the field of view.

[0084] Step 103: perform data preprocessing and feature extraction on the debris data information and the material distribution status data information.

[0085] In some embodiments of the present disclosure, in step 103, a variety of preprocessing means and feature extraction methods are used, including but not limited to: correction, averaging, smoothing, differentiation, normalization, and dimensionality reduction, to perform data preprocessing and feature extraction.

[0086] In some embodiments of the present disclosure, calibration includes both spatial distortion correction and black-and-white calibration. Spatial distortion is typically integrated into the camera before shipment. Black-and-white calibration is performed before each measurement, similar to limiting the data dimension to a range of 0-1, where 0 represents a blackboard and 1 represents a whiteboard.

[0087] In some embodiments of the present disclosure, the effects of averaging, smoothing, differentiation, and normalization are similar to filtering properties. Since the surface of tobacco is uneven, these operations are mainly to correct the reflectivity so that its spectral curve is more consistent with its own reflectivity and less affected by external factors.

[0088] In some embodiments of the present disclosure, for dimensionality reduction, Z is the spectral resolution. A hyperspectral camera has hundreds of spectral channels, the data is too redundant, and it affects the processing speed, so the present disclosure needs to perform a dimensionality reduction operation.

[0089] Step 104: Pre-establish a component determination detection model.

[0090] In some embodiments of the present disclosure, the component determination detection model may adopt models including but not limited to: linear regression, least squares method, decision tree, random forest, vector machine, neural network, etc.

[0091] In some embodiments of the present disclosure, after the component determination detection model is established, the established model can be used to determine impurities and material distribution status (i.e., perform at least one of steps 11 to 16).

[0092] In some embodiments of the present disclosure, identification is based on the spectral difference between the foreign matter and the tobacco material curtain, which is a semi-supervised recognition model or an unsupervised recognition model, and does not require the establishment of a model in advance.

[0093] In other embodiments of the present disclosure, the supervised model needs to know the map information of the debris in advance, similar to first collecting the debris map data, then collecting the material map data, performing learning and modeling, solidifying the model after multiple iterations, and then applying it to the judgment.

[0094] The model described in step 104 is mostly a supervised model, which means that the map information of the debris needs to be known in advance. For example, the debris map data is first collected, and then the material map data is collected for learning and modeling. After the model is solidified after multiple iterations, it is applied to the judgment.

[0095] Steps 11 to 17 are for measurement and model determination.

[0096] Step 11-Step 12: Extracting data information of the material curtain in the air separation box during the air separation and impurity removal process. The details of Step 11 and Step 12 are given below.

[0097] Step 11: Using spectral measurement and analysis technology, collect the material spectrum image in the air separation box during the air separation and impurity removal process, wherein the material spectrum image includes material distribution status data information.

[0098] In some embodiments of the present disclosure, step 11 may include: in the production (or processing) process, during the air separation process, for the material curtain in the air separation box, a spectral detection system is used while an imaging system and a spectral system are used to collect material information.

[0099] Step 12: performing data preprocessing and feature extraction on the material spectrum image.

[0100] In some embodiments of the present disclosure, step 12 may include: preprocessing the material spectral image to obtain normalized two-dimensional image data.

[0101] In some embodiments of the present disclosure, step 12 may include: performing black-white frame correction on the collected spectral image; and filtering the reflectance data after the black-white frame correction.

[0102] Step 13: Identify and determine the debris in the air separation box.

[0103] In some embodiments of the present disclosure, step 13 may include: using a pre-established component determination detection model to determine whether there are foreign objects in the material in the air separation box.

[0104] In some embodiments of the present disclosure, step 13 may include: calculating the mean spectral vector based on the two-dimensional image data; determining the target spectral vector based on the two-dimensional image data and the mean spectral vector; determining the response graph of the entire image based on the target spectral vector and the covariance matrix; comparing the response graph of the entire image with a predetermined threshold to determine a binary graph of the detection result; and determining whether there are foreign objects in the material in the air separation box based on the binary graph of the detection result.

[0105] Step 14: Control the air volume in the air separation box.

[0106] In some embodiments of the present disclosure, step 14 may include: adjusting the air volume by changing the frequency of the air separation fan according to whether there are foreign objects in the material in the air separation box, thereby changing the strength of removing foreign objects.

[0107] In some embodiments of the present disclosure, in step 14, the step of adjusting the air volume by changing the frequency of the air separation fan to change the strength of debris removal may include: adjusting the air volume by the weight of the debris itself and changing the frequency of the air separation fan to change the strength of debris removal, wherein the step of adjusting the air volume by changing the frequency of the air separation fan may include: reducing the fan frequency or increasing the damper opening to adjust the air separation wind speed.

[0108] In some embodiments of the present disclosure, steps 13 and 14 may include: using the established model (steps 101, 103, 104) to compare whether there are debris, tobacco stems (or stem sticks) and wet clumps in the material curtain in the air separation box, and according to whether there are debris, tobacco stems (or stem sticks) and wet clumps, adjusting the air volume by changing the frequency of the air separation fan to change the intensity of debris removal.

[0109] Step 15: Determine the material distribution status in the air separation box.

[0110] In some embodiments of the present disclosure, step 15 may include: using a pre-established component determination detection model to identify the spectral image of the material in the air separation box, analyzing the material image distribution information, and determining whether the material distribution is uniform.

[0111] In some embodiments of the present disclosure, step 15 may include at least one of steps 151 to 153, wherein:

[0112] Step 151 : Analyze the spectral image of the material in the air separation box to analyze the material image distribution information and the material concentration per unit area.

[0113] Step 152: Identify the flow field distribution status in the air separation box based on the material concentration distribution information.

[0114] Step 153 , judging whether the material distribution is uniform by using the coefficient of variation of the material concentration ratio per unit area after binarization.

[0115] Step 16: Control the flow field distribution in the air separation box.

[0116] In some embodiments of the present disclosure, step 16 may include: adjusting the flow field using the air distribution mechanism in the air separation box according to the uniformity of the material distribution, changing the air flow distribution state in the air separation box, and then adjusting the material distribution in the air separation box.

[0117] In some embodiments of the present disclosure, in step 16, the step of adjusting the flow field by using the air distribution mechanism in the air separation box according to the uniformity of material distribution may include: when the coefficient of variation of the material concentration ratio is greater than a predetermined threshold, controlling the air distribution mechanism to adjust the flow field by adjusting the orifice plate opening in different areas of the air distribution plate.

[0118] In some embodiments of the present disclosure, steps 15 and 16 may include: using the established model (steps 102, 103, and 104) to identify the distribution state of the material curtain in the air separation box, and determine whether the material distribution is uniform, using the changes in the air distribution mechanism in the air separation box to adjust the flow field, change the airflow distribution state in the air separation box, and adjust the air separation effect.

[0119] In some embodiments of the present disclosure, the material curtain wind continuously blows up materials such as tobacco to form a material curtain.

[0120] In some embodiments of the present disclosure, the steps of identifying the distribution state of the material curtain within the air separation box and determining whether the material distribution is uniform may include: using imaging to determine where in a two-dimensional image there is more material and where there is less material, and applying angles to primarily determine when excessive material is being blown up, requiring adjustment of the air delivery volume, material flow rate, etc. The distribution position can be determined by an algorithm.

[0121] Step 17: Control and optimization of air separation and impurity removal process.

[0122] In some embodiments of the present disclosure, step 17 may include: utilizing a tobacco material air separation and impurity removal process control system and method, including, but not limited to, removing impurities during the tobacco processing process (step 14), optimizing the flow field distribution state (step 16), etc., to improve air separation efficiency and air separation quality.

[0123] In some embodiments of the present disclosure, the two airflows in steps 14 and 16 are unrelated. Step 14 involves detecting an anomaly during the fluidization of the material curtain and then applying negative or high pressure to the output position to remove foreign matter. The airflow in step 16 is intended to continuously lift the material. Large or small airflows can alter the distribution and density of the material curtain.

[0124] In some embodiments of the present disclosure, step 17 may include: process control, such as a single foreign body identification process, parameter selection.

[0125] Step 18: Control and optimize application.

[0126] In some embodiments of the present disclosure, step 18 may include: air separation and impurity removal control during the processing, parameter optimization, equipment development, etc.

[0127] In some embodiments of the present disclosure, step 18 may include: overall control, looking at the entire processing process, for example, some foreign objects are identified, but false alarms are frequent, affecting production, etc.

[0128] The above-mentioned embodiments of the present disclosure use spectral imaging technology to detect air-selected impurities and wet clumps in the cigarette processing process, thereby controlling their removal; at the same time, the above-mentioned embodiments of the present disclosure control the flow field distribution based on the material distribution state in the bellows, thereby improving the air selection efficiency and quality.

[0129] The tobacco material air separation and impurity removal process control method disclosed in the present invention is described below through specific examples.

[0130] First type of embodiment

[0131] The first type of embodiment is a tobacco material air separation and impurity removal process control method based on an unsupervised model, which may include:

[0132] Step 1: Collect data. Figure 3 The embodiment is a schematic diagram of images collected by data acquisition in some embodiments of the present disclosure.

[0133] Step 2: Data preprocessing. Figure 4 Schematic diagram of an image after filtering and denoising in some embodiments of the present disclosure. Figure 5 Schematic diagram of an image after background separation and morphological processing in some embodiments of the present disclosure.

[0134] In some embodiments of the present disclosure, step 2 may include: bilateral filtering to eliminate details and noise and retain edge information, such as Figure 4 As shown; background separation and morphological processing separate shadows and lower tobacco leaves, as shown Figure 5 shown.

[0135] Step 3: instance segmentation. Figure 6 Schematic diagram of an image after instance segmentation in some embodiments of the present disclosure.

[0136] In some embodiments of the present disclosure, step 3 may include: the first step of segmentation is to calculate the Euclidean distance of each foreground pixel to the nearest zero (i.e., background pixel, black pixel) by calculating the Euclidean distance transform (EDT), Figure 6 To visualize EDT; Figure 6 The method finds the valleys (i.e., local minima) in the image, ensuring that there is at least 50 pixels between each peak. The output provides labels, which are then fed into the segmentation algorithm to return a label matrix, an array with the same width and height as the input image. Each pixel value is treated as a unique label value. Pixels with the same label value belong to the same object.

[0137] Second type of embodiment

[0138] The second type of embodiment is a tobacco material air separation and impurity removal process control method using an unsupervised model, which may include:

[0139] The specific embodiments of the present invention are further described in detail below with reference to the accompanying drawings.

[0140] Step (1) Experimental instruments and parameter requirements.

[0141] Spectral imager, with a wavelength range of 1000-2500nm, a spectral resolution of 12nm, an image resolution of 384*288 pixels, and a spectrometer frame rate of 400.

[0142] Step (2) Collection and preparation of tobacco wet mass and debris samples.

[0143] In order to make the established tobacco wet mass material model more widely applicable, the tobacco wet mass material samples in this embodiment selected tobacco wet mass between 14% and 30%, tobacco cuts exported from a tobacco drying machine, stems provided by a cigarette company, metal, plastic and other debris.

[0144] Step (3) spectral imaging of wet tobacco pellets and debris samples.

[0145] Wet tobacco clumps with different moisture contents and debris were spread onto a black backing plate and marked. A spectrometer was used to scan and image the wet tobacco clumps with different moisture contents and different types of debris. The images were collected and corrected for black and white, and stored as raw spectral data.

[0146] Step (4) Preprocessing of spectral images of wet tobacco balls and debris samples and acquisition of characteristic images.

[0147] In order to eliminate the influence of light source, system background, etc. on the sample, the collected spectral image is subjected to black and white frame correction. Black and white frame correction means eliminating external influences, as shown in formula (1):

[0148]

[0149] In formula (1): R is the corrected spectral image; I is the original spectral image; B is the all-black image acquired by turning off the camera lens; W is the all-white image obtained by scanning the white calibration plate.

[0150] To eliminate the effects of noise, this disclosure filters (e.g., smoothing) the reflectance data of the black and white frames after correction. Filtering can improve the smoothness of the spectral curve and reduce noise interference. The key to convolution smoothing lies in solving the matrix operator.

[0151] In some embodiments of the present disclosure, target detection algorithms typically assume a uniform background distribution when modeling background information. However, in real images acquired by hyperspectral sensors, background pixels and target pixels are often intermingled, making it difficult to separate target pixels when calculating background information. Consequently, the only way to obtain an approximate autocorrelation matrix is ​​to directly calculate the information of the entire image. To address this, the present disclosure proposes a target detection algorithm that establishes binary hypotheses for different background distributions, for both background-only and target-presence scenarios, and constructs a target detection model based on the generalized likelihood ratio.

[0152] Assume that the hyperspectral image is represented as a set of all pixels, namely {r1, r2, ..., r N}, where r i =(r i1 , r i2 ,...,r iL ) T , L represents r i The dimension of the pixel, N represents the total number of pixels in the hyperspectral image to be tested. Furthermore, we assume that the desired target feature element vector to be detected is d = (d1, d2, ..., d L ) T The target detection algorithm disclosed in this paper first establishes binary hypotheses for the two cases of target absence and target presence. In the background hypothesis, there is no target signal, so there are only background and additive noise n, n obeys a multivariate normal distribution with a mean of μ. To further standardize the data, the current mean μ can be subtracted from the original image. At this time, the covariance matrix of the background is ∑. In the target hypothesis, there are both background and target, so the binary hypothesis and Gaussian distribution are expressed as shown in formulas (2) and (3):

[0153]

[0154]

[0155] In formulas (2) and (3), H0 and H1 are the hypotheses that the target does not exist and the target exists, respectively. In the hypothesis H1 that the target exists, it also obeys the multivariate normal distribution, with a mean of ad and a covariance matrix of c 2∑, a represents the proportion of target endmembers d in a pixel, and c represents the proportion of background noise n. Under the two assumptions H0 and H1, the calculation formula designed by generalized likelihood ratio is shown in formula (4):

[0156]

[0157] Since r and d are both vectors, the covariance matrix ∑ -1 Is a symmetric matrix, formula (4) can be further simplified to formula (5):

[0158]

[0159] The disclosed target detection algorithm can detect the similarity between a pixel r and a target sample d. The larger the detection value, the more similar the two are, and the greater the probability of being a target. By setting a threshold η, it can automatically determine whether the pixel is a target, as shown in formula (6).

[0160]

[0161] Step (5) spectral image information preprocessing and feature extraction.

[0162] Figure 7 Schematic diagram of some embodiments of the target detection algorithm disclosed in this disclosure. Figure 7 As shown, the basic steps of the target detection algorithm disclosed in this disclosure are as follows:

[0163] Step 5-1: receiving an input hyperspectral image.

[0164] Step 5-2, determine the target spectrum vector d.

[0165] In some embodiments of the present disclosure, step 5-2 may include: spectral scanning of fluidized tobacco material (including wet masses and impurities, etc.) during the air selection process; for the air selection process, in the fluidized state of the tobacco material, a binocular spectrometer is used through a window to perform a spectral scan of the fluidized tobacco material during the air selection process, collect material spectrum information of the air selection process, and determine the target spectral vector d.

[0166] Step 5-3: pre-process the hyperspectral image (collected spectral image) to obtain normalized two-dimensional data r (L×N).

[0167] In some embodiments of the present disclosure, the two-dimensional data is hyperspectral image data of tobacco material.

[0168] Step 5-4, calculate the covariance matrix of the image according to formula (7).

[0169]

[0170] In some embodiments of the present disclosure, step 5-4 may include: calculating the mean spectral vector μ based on the image data r(L×N); calculating the covariance matrix of the current image according to formula (7) based on the image data r(L×N) and μ, wherein the current image contains background and material, and the current image may contain foreign matter.

[0171] Step 5-5, for each pixel spectrum r, calculate the output response δ ACE (r).

[0172] In some embodiments of the present disclosure, step 5-5 may include: calculating the response map of the entire image according to formula (5), such as Figure 8 The second column is shown. Figure 8 Schematic diagram of detection results and binarized images in some embodiments of the present disclosure.

[0173] Steps 5-6, setting thresholds).

[0174] Steps 5-7, perform binarization processing, such as Figure 8 shown.

[0175] Steps 5-8, get the target detection results, such as Figure 8 shown.

[0176] In some embodiments of the present disclosure, steps 5-6 to 5-8 may include: setting a threshold value), and obtaining a binary image of the detection result according to formula (6).

[0177] Step (6) removing wet tobacco clumps and debris from the air separation box.

[0178] Based on the results of the identification of tobacco clumps and debris in the air separation box, the air volume is adjusted by the weight of the tobacco clumps and debris, and by changing the frequency of the air separation fan to change the force of debris removal. In other words, the fan frequency is reduced or the damper opening is increased to adjust the air separation speed and achieve the removal of tobacco clumps and debris.

[0179] In some embodiments of the present disclosure, weight is measured on a conveyor belt scale on the production line, and the weight and flow rate are displayed in real time during the material flow process. The weight of the debris must be identified in advance, similar to empirical properties, and is related to both the material flow rate and wind speed. Foreign matter is heavier than tobacco and will fall down if it cannot be blown up.

[0180] In some embodiments of the present disclosure, the fan frequency or damper opening is adjusted according to the weight through model control such as PID.

[0181] In some embodiments of the present disclosure, the relationship between velocity, fluidization wind speed and material weight can be determined according to the following formula: suspension velocity = (square of particle diameter * gravitational acceleration * (particle density - fluid density)) / (18 * fluid viscosity).

[0182] Step (7) Analyze and process the image information of the material in the air separation box.

[0183] Analyze the material image in the air separation box, analyze the material image distribution information and the material concentration per unit area, identify the flow field distribution status in the air separation box based on the material concentration distribution information, evaluate and determine whether the material distribution is uniform, and use this as the basis for adjusting the air volume and distribution of the flow field in the air separation box.

[0184] Step (8) adjusting the material distribution state in the air separation box.

[0185] Based on the identification of the distribution state of the material curtain in the air separation box in the above step (7), it is determined whether the material distribution is uniform. The uniformity of the distribution is determined by the coefficient of variation of the material concentration ratio after binarization per unit area. If the coefficient of variation is greater than 10%, the flow field is adjusted by changing the air distribution mechanism in the air separation box, changing the air flow distribution state in the air separation box, and adjusting the air separation effect. The air distribution mechanism is achieved by adjusting the opening of the orifice plate in different areas of the air distribution plate.

[0186] Whether the distribution of the present invention is uniform can be determined by calculating the coefficient of variation of the material concentration after binarization per unit area: calculate the particle concentration at each moment, and adjust the fan if the change is too large.

[0187] The present invention utilizes the change of the air distribution mechanism in the air separation box to adjust the flow field, including: if there is too little material in the upper left corner, it is possible that the wind in the upper left corner area is too weak and the material is not suspended, then adjust the orifice plate opening in the upper left corner of the air distribution mechanism to increase the air volume in the upper left corner.

[0188] The present disclosure provides a method for detecting tobacco wet mass substances based on shortwave imaging spectroscopy technology, which has the following significant improvements compared to related technologies:

[0189] (1) The above embodiments of the present disclosure utilize spectral imaging to collect material characteristic data and establish a model, and can perform an online determination of debris and wet clumps in a winnowing box, which is simple and quick.

[0190] (2) The method of the above embodiment of the present disclosure can measure the material distribution state in the air separation box, determine the uniformity of material distribution, and provide a method and basis for process distribution control and optimization and improvement.

[0191] (3) The above embodiments of the present disclosure use a device and method for air separation and impurity removal during the cigarette processing process to detect, determine, and remove impurities and wet clumps from the cigarette processing process; at the same time, the flow field distribution is controlled according to the material distribution state in the bellows, thereby improving the air separation efficiency and quality, and providing technical means for the development of air separation equipment.

[0192] (4) The above embodiments of the present disclosure have the advantages of low cost, high efficiency, fast, accurate and simple operation.

[0193] (5) The experimental process of the above embodiments of the present disclosure is simple and quick, non-destructive to the samples and non-polluting to the environment.

[0194] Figure 9 Schematic diagram of some embodiments of the tobacco material air separation and impurity removal process control device disclosed in the present invention. Figure 9 As shown, the tobacco material air separation and impurity removal process control device disclosed herein may include an image acquisition module 91, a pre-processing module 92, a debris recognition module 93 and a rejection control module 94, wherein:

[0195] The image acquisition module 91 is configured to use spectral measurement and analysis technology to acquire a spectral image of the material in the air separation box during the air separation and impurity removal process, wherein the spectral image of the material includes material distribution status data information.

[0196] The preprocessing module 92 is configured to perform data preprocessing and feature extraction on the material spectrum image.

[0197] The debris identification module 93 is configured to use a pre-established component determination detection model to determine whether there are debris in the material in the air separation box.

[0198] The rejection control module 94 is configured to adjust the air volume by changing the frequency of the air separation fan according to whether there are foreign objects in the material in the air separation box, thereby changing the strength of the foreign objects rejection.

[0199] In some embodiments of the present disclosure, the tobacco material air separation and impurity removal process control device of the present disclosure can be configured to execute any of the above embodiments of the present disclosure (for example Figures 1 to 8 The operation of the tobacco material air separation and impurity removal process control method described in any embodiment).

[0200] The above embodiments of the present disclosure provide a device and method for controlling air separation and impurity removal in a cigarette processing process. Specifically, the above embodiments of the present disclosure relate to a device and method for controlling air separation and impurity removal of tobacco shreds (or tobacco sheets) using spectral imaging technology.

[0201] The above-mentioned embodiments of the present disclosure use spectral imaging technology to control the air separation of wet clumps, debris and tobacco stems (or stem sticks) during the cigarette processing process, and at the same time control the flow field according to the material distribution state in the bellows to improve the air separation efficiency and quality.

[0202] Figure 10 Schematic diagram of the structure of some other embodiments of the tobacco material air separation and impurity removal process control device disclosed in the present invention. Figure 10 As shown, the tobacco material air separation and impurity removal process control device includes a memory 101 and a processor 102.

[0203] The memory 101 is used to store instructions. The processor 102 is coupled to the memory 101. The processor 102 is configured to execute the above embodiments (for example, Figures 1-8 The tobacco material air separation and impurity removal process control method described in any embodiment).

[0204] like Figure 10 As shown, the tobacco material air separation and impurity removal process control device also includes a communication interface 103 for exchanging information with other devices. At the same time, the tobacco material air separation and impurity removal process control device also includes a bus 104, through which the processor 102, the communication interface 103, and the memory 101 communicate with each other.

[0205] Memory 101 may include high-speed RAM memory or non-volatile memory, such as at least one disk storage device. Memory 101 may also be a memory array. Memory 101 may also be divided into blocks, and the blocks may be combined into virtual volumes according to certain rules.

[0206] Furthermore, the processor 102 may be a central processing unit (CPU), or may be an application-specific integrated circuit (ASIC), or may be configured to implement one or more integrated circuits of the embodiments of the present disclosure.

[0207] The above-mentioned embodiments of the present disclosure provide a tobacco material air separation and impurity removal process control device and method based on spectral imaging technology.

[0208] The above embodiments of the present disclosure provide a method for detecting foreign matter, tobacco stems (or stem sticks), and the number of wet clumps in tobacco material.

[0209] The above-mentioned embodiments of the present disclosure provide a method for identifying and determining debris, tobacco stems (or stem sticks) and wet clumps in a winnowing box based on spectral imaging technology.

[0210] The above-mentioned embodiments of the present disclosure provide a method for identifying and determining the uniformity of material distribution in an air separation box based on spectral imaging technology.

[0211] The above-mentioned embodiments of the present disclosure provide a tobacco material air separation and impurity removal process control device and method based on spectral imaging technology. Through high-infrared spectral imaging technology, the air-separated impurities, tobacco stems (or stem sticks) and wet clumps in the cigarette processing process are detected and determined, thereby performing removal control. At the same time, the flow field distribution is controlled according to the material distribution state in the bellows, thereby improving the air separation efficiency and air separation quality, and also providing technical means for the research and development of air separation equipment.

[0212] According to another aspect of the present disclosure, a computer-readable storage medium is provided, wherein the computer-readable storage medium stores computer instructions, and when the instructions are executed by a processor, the computer-readable storage medium implements any of the above embodiments (e.g. Figures 1-8 The tobacco material air separation and impurity removal process control method described in any embodiment).

[0213] In some embodiments of the present disclosure, the computer-readable storage medium may be a non-transitory computer-readable storage medium.

[0214] The above description of the present disclosure in conjunction with the embodiments is in no way intended to limit the embodiments of the present disclosure. The tobacco material air separation and impurity removal process control device and method include, but are not limited to, removing impurities during tobacco processing, optimizing flow field distribution, and improving air separation efficiency and quality. Various changes, modifications, substitutions, and variations can be made within the scope of knowledge possessed by those skilled in the art without departing from the principles and purpose of the present disclosure. The scope of the present disclosure is defined by the claims and their equivalents.

[0215] Those skilled in the art will appreciate that embodiments of the present disclosure may be provided as methods, apparatus, or computer program products. Therefore, the present disclosure may take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present disclosure may take the form of a computer program product implemented on one or more computer-usable non-transient storage media (including but not limited to magnetic disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0216] The present disclosure is described with reference to flowcharts and / or block diagrams of methods, devices (systems) and computer program products according to embodiments of the present disclosure. It should be understood that each process and / or block in the flowchart and / or block diagram and the combination of processes and / or blocks in the flowchart and / or block diagram can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 a process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0217] These computer program instructions may also be stored in a computer readable memory that can direct a computer or other programmable data processing device to work in a specific manner, so that the instructions stored in the computer readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 a process or multiple processes and / or boxes Figure 1The function specified in one or more boxes.

[0218] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operational steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing the instructions executed on the computer or other programmable device for implementing the process. Figure 1 a process or multiple processes and / or boxes Figure 1 A step that specifies a function in one or more boxes.

[0219] The tobacco material air separation and impurity removal process control device, image acquisition module, preprocessing module, impurity identification module and rejection control module described above can be implemented as a general-purpose processor, programmable logic controller (PLC), digital signal processor (DSP), application-specific integrated circuit (ASIC), field programmable gate array (FPGA) or other programmable logic device, discrete gate or transistor logic device, discrete hardware component or any appropriate combination thereof for performing the functions described in this application.

[0220] The present disclosure has been described in detail so far. To avoid obscuring the concept of the present disclosure, some details known in the art have not been described. Based on the above description, those skilled in the art can fully understand how to implement the technical solutions disclosed herein.

[0221] Those skilled in the art will understand that all or part of the steps to implement the above embodiments may be accomplished by hardware, or may be accomplished by instructing the relevant hardware through a program, and the program may be stored in a non-transitory computer-readable storage medium, and the above-mentioned storage medium may be a read-only memory, a disk or an optical disk, etc.

[0222] The description of the present disclosure is provided for purposes of illustration and description and is not intended to be exhaustive or to limit the disclosure to the disclosed form. Many modifications and variations will be apparent to those skilled in the art. The embodiments are selected and described in order to better illustrate the principles and practical applications of the present disclosure and to enable those skilled in the art to understand the present disclosure and design various embodiments with various modifications suitable for specific applications.

Claims

1. A tobacco material air separation and impurity removal process control method, comprising: Using spectral measurement and analysis technology, collect the material spectrum image in the air separation box during the air separation and impurity removal process, wherein the material spectrum image includes material distribution status data information; Performing data preprocessing and feature extraction on the material spectral image, wherein the performing data preprocessing and feature extraction on the material spectral image comprises: preprocessing the material spectral image to obtain normalized two-dimensional image data; Using a pre-established component determination detection model, determining whether there are foreign objects in the material in the winnowing box, wherein using the pre-established component determination detection model to determine whether there are foreign objects in the material in the winnowing box includes: calculating a mean spectral vector based on the two-dimensional image data; determining a target spectral vector based on the two-dimensional image data and the mean spectral vector; determining a response map of the entire image based on the target spectral vector and the covariance matrix; comparing the response map of the entire image with a predetermined threshold to determine a binary map of the detection result; and determining whether there are foreign objects in the material in the winnowing box based on the binary map of the detection result; According to whether there are foreign objects in the air separation box, the air volume is adjusted by changing the frequency of the air separation fan to change the strength of removing foreign objects; Using a pre-established component determination detection model, the spectral image of the material in the air separation box is identified, the material image distribution information is analyzed, and whether the material distribution is uniform is determined. The identifying the spectral image of the material in the air separation box, analyzing the material image distribution information, and determining whether the material distribution is uniform includes: analyzing the spectral image of the material in the air separation box, analyzing the material image distribution information and the material concentration per unit area, identifying the flow field distribution condition in the air separation box based on the material concentration distribution information, and determining whether the material distribution is uniform through the material concentration ratio variation coefficient after binarization per unit area; According to the uniformity of material distribution, the air distribution mechanism in the air separation box is used to adjust the flow field, change the air flow distribution state in the air separation box, and then adjust the material distribution in the air separation box. Among them, according to the uniformity of material distribution, the air distribution mechanism in the air separation box is used to adjust the flow field, including: when the coefficient of variation of the material concentration ratio is greater than a predetermined threshold, the air distribution mechanism is controlled to adjust the flow field by adjusting the orifice plate opening in different areas of the air distribution plate.

2. The tobacco material air separation and impurity removal process control method according to claim 1, wherein: The foreign matter includes at least one of tobacco stems, stem sticks, wet lumps, fibers, plastics and metals.

3. The tobacco material air separation and impurity removal process control method according to claim 1 or 2, wherein: The method of adjusting the air volume by changing the frequency of the air separation fan to change the debris removal strength includes: The force of removing debris is changed by adjusting the air volume by the weight of the debris itself and by changing the frequency of the air separation fan. The air volume adjustment by changing the frequency of the air separation fan includes: reducing the fan frequency or increasing the damper opening to adjust the air separation wind speed.

4. The tobacco material air separation and impurity removal process control method according to claim 1 or 2, wherein: The data preprocessing and feature extraction of the material spectral image includes: Perform black and white frame correction on the collected spectral images; Filter the corrected reflectance data of the black and white frames.

5. The tobacco material air separation and impurity removal process control method according to claim 1 or 2, further comprising: A component determination detection model is established in advance, wherein the component determination detection model is an unsupervised model, a semi-supervised model or a supervised model.

6. The tobacco material air separation and impurity removal process control method according to claim 5, wherein: In the case where the component determination detection model is a supervised model, the pre-established component determination detection model includes: Hyperspectral measurement technology is used to collect data on debris during the air separation and impurity removal process; Spectral measurement and analysis technology is used to collect data on the material distribution status in the air separation box during the air separation and impurity removal process; Perform data preprocessing and feature extraction on debris data information and material distribution status data information; Establish a component determination and detection model in advance.

7. A tobacco material air separation and impurity removal process control device, comprising: An image acquisition module is configured to use spectral measurement and analysis technology to acquire a spectral image of the material in the air separation box during the air separation and impurity removal process, wherein the spectral image of the material includes data information on the material distribution state; a preprocessing module configured to perform data preprocessing and feature extraction on the material spectral image, wherein the data preprocessing and feature extraction on the material spectral image comprises: preprocessing the material spectral image to obtain normalized two-dimensional image data; The debris identification module is configured to use a pre-established component determination detection model to determine whether debris exists in the material in the winnowing box, wherein the use of the pre-established component determination detection model to determine whether debris exists in the material in the winnowing box includes: calculating a mean spectral vector based on the two-dimensional image data; determining a target spectral vector based on the two-dimensional image data and the mean spectral vector; determining a response map of the entire image based on the target spectral vector and the covariance matrix; comparing the response map of the entire image with a predetermined threshold to determine a binary map of the detection result; and determining whether debris exists in the material in the winnowing box based on the binary map of the detection result; The rejection control module is configured to adjust the air volume by changing the frequency of the air separation fan according to whether there are foreign objects in the material in the air separation box, thereby changing the strength of the foreign objects rejection; The tobacco material air separation and impurity removal process control device is further configured to use a pre-established component determination detection model to identify the spectral image of the material in the air separation box, analyze the material image distribution information, and determine whether the material distribution is uniform; based on the uniformity of the material distribution, the air distribution mechanism in the air separation box is used to adjust the flow field, change the airflow distribution state in the air separation box, and thus adjust the material distribution in the air separation box; Among them, the tobacco material air separation and impurity removal process control device is configured to analyze the material spectral image in the air separation box, analyze the material image distribution information, and determine whether the material distribution is uniform, and identify the flow field distribution status in the air separation box based on the material concentration distribution information, and determine whether the material distribution is uniform through the material concentration ratio variation coefficient after binarization per unit area. The tobacco material air separation and impurity removal process control device uses the air distribution mechanism in the air separation box to adjust the flow field according to the uniformity of material distribution. When the coefficient of variation of the material concentration ratio is greater than a predetermined threshold, the air distribution mechanism is controlled to adjust the flow field by adjusting the orifice plate opening in different areas of the air distribution plate.

8. A tobacco material air separation and impurity removal process control device, comprising: a memory configured to store instructions; The processor is configured to execute the instructions so that the tobacco material air separation and impurity removal process control device performs the operation of the tobacco material air separation and impurity removal process control method according to any one of claims 2 to 6.

9. A computer-readable storage medium, wherein: The computer-readable storage medium stores computer instructions, and when the instructions are executed by a processor, the tobacco material air separation and impurity removal process control method according to any one of claims 1 to 6 is implemented.

Images