Tobacco stem selection method, device, medium and equipment

Abstract

The invention discloses a tobacco stem selection method, device, medium and equipment. The method comprises the steps that tobacco stem materials are spread to be thin; identifying the target separated object of the tobacco stem material based on the characteristic difference of the target separated object in the double-sided tobacco stem image; qualified tobacco stems are separated through a high-speed air valve, and target separated objects are blown to a second conveying belt; and carrying out secondary identification and selection on qualified tobacco stems in the single-sided or double-sided material images of the separated materials based on preset features. Tobacco stem materials are spread to be in a dispersed and non-overlapping state through a combined material uniformizing process, so that the accuracy of visual recognition is guaranteed, and the length of a spreading path is reduced; meanwhile, double-sided visual inspection is carried out in the material throwing process, and an identification rule is established by utilizing the double-sided imaging characteristic difference of qualified tobacco stems and unqualified tobacco stems, so that the accurate identification effect on stem crutches, sundries with similar density, tobacco stems with different colors and the like in the tobacco stem raw materials is improved, and the tobacco shred quality is greatly improved.

Description

Technical Field

[0001] This invention relates to the field of tobacco manufacturing technology, and in particular to a method, apparatus, medium, and equipment for selecting tobacco stems. Background Technology

[0002] Tobacco stems are widely used as fillers in cigarette products due to their high filling capacity and good combustibility. However, existing tobacco stem raw materials often contain a certain amount of substandard products such as stem fragments and impurities. Stem fragments are hard, difficult to moisten, and easily crushed during pressing. After cutting, there are many stem pieces and chunks in the stem shreds, affecting the quality of the shreds. Furthermore, rubber, plastic pieces, paper pieces, and carbonized blackened tobacco stems in the stems can seriously affect the sensory quality of the product. Therefore, the quality of tobacco stem raw materials is the foundation for ensuring the quality of tobacco stem processing. However, as an auxiliary additive material in the cigarette industry, the quality of tobacco stems has not received widespread attention in the industry for many years. Only in recent years has the quality of tobacco stems gradually come into focus. According to existing publicly available patent documents, Henan China Tobacco Industry Co., Ltd.'s CN202510675415.7 describes an intelligent multi-stage sorting and impurity removal system for tobacco stems. This system uses three-stage air separation boxes connected in series and selects tobacco stems by adjusting the air volume and optimizing its adjustment parameters. However, this gravity-based screening method cannot remove discolored tobacco stems that are blackened, moldy, or carbonized, as well as other impurities such as plastic strips. The processed tobacco stems still contain some unqualified stems that enter subsequent processes, making it difficult to meet the quality requirements of tobacco stems, especially for high-end cigarettes. Summary of the Invention

[0003] The present invention aims to solve one of the above-mentioned technical problems. In a first aspect, it provides a method for selecting tobacco stems, comprising the following steps: Step 1: Spread the tobacco stem material thinly using a combined homogenization process; Step 2: The thinned tobacco stem material is thrown to the first image recognition area via the first conveyor belt. During the material throwing process, double-sided tobacco stem images of the tobacco stem material are acquired, and the target separator is identified based on the feature differences of the target separator in the double-sided tobacco stem images. Step 3: Separate the qualified tobacco stems from the identification results using a high-speed air valve, and then blow the target separated material to the next station.

[0004] Furthermore, the combined uniform material process includes a vibratory trough uniform material process, a sliding plate uniform material process, and a differential speed uniform material process arranged sequentially. Step 1 specifically includes: Adjust the corresponding material mixing parameters according to the preset material mixing mode; Control the tobacco stem material to fall into the uniform vibrating trough, and use the guide strip of the vibrating trough to disperse the tobacco stem material to the left and right sides along the herringbone angle; The dispersed tobacco stem material is controlled to fall into the slide plate and slide downward under the action of vibration and gravity, so that the overlapping tobacco stem material is dispersed a second time by speed difference and falls into the first conveyor belt; The current horizontal speed of the tobacco stem material during the falling process and the current spreading state of the tobacco stem material on the first conveyor belt are collected. The current spreading state is compared with the target spreading state, and the current running speed of the first conveyor belt is adjusted according to the comparison result so that the tobacco stem material is dispersed three times through speed difference and the spreading density of the tobacco stem material on the first conveyor belt reaches the target spreading state.

[0005] Furthermore, the material leveling parameters include the setting angle of the guide strip in the material leveling vibrating trough, the motor vibration frequency of the material leveling vibrating trough, the tilt angle of the slide plate, and the initial running speed of the first conveyor belt.

[0006] Further, in step 2, the target separator is identified based on the feature differences of the target separator in the double-sided tobacco stem image, specifically as follows: A first camera and a second camera are respectively set at symmetrical positions on both sides of the tobacco stem in the first image recognition area; During the material throwing process, the first camera and the second camera simultaneously capture images of the front and back sides of the tobacco stem material, respectively, and mark them. A preset image classification model is used to initially identify the target separators in the front and back tobacco stem images based on preset features, generating the type, front feature parameter value and back feature parameter value of each target separator, and calculating the feature difference between the front feature parameter value and the back feature parameter value; Determine whether the feature differences meet the preset conditions corresponding to the target separation type, and optimize the initial identification result based on the determination result, so that the qualified tobacco stems in the optimized identification result are blown to the next process through the high-speed air valve; otherwise, they are blown to the next station as the target separation.

[0007] Furthermore, it also includes: Step 4, the next station is a second conveyor belt, the second conveyor belt throws the separated material to the second image recognition area, during the material throwing process, single-sided or double-sided material images are collected, and qualified tobacco stems in the single-sided or double-sided material images are re-identified and selected based on preset features; The first conveyor belt and the second conveyor belt are arranged vertically. The target object blown off from the first conveyor belt falls onto the second conveyor belt and is transported in the opposite direction to the second image recognition area.

[0008] Furthermore, in step 4, the qualified tobacco stems in the single-sided or double-sided material image are further identified and selected based on preset features, specifically as follows: A third camera and a fourth camera are respectively set at symmetrical positions on both sides of the tobacco stem in the second image recognition area; During the material throwing process, the third camera and the fourth camera simultaneously acquire and mark the front and back images of the separated material. A preset image classification model is used, and qualified tobacco stems in the front and back material images are identified based on the preset features; If an object at the same location is identified as a qualified tobacco stem in both the front and back material images, it is blown to the next process through a high-speed air valve; otherwise, it is classified and recycled as a target separated material.

[0009] A second aspect of the present invention provides a tobacco stem refining device, comprising a uniform feeding module, a first identification module, a first refining module, a second identification module, and a second refining module. The material spreading module is used to spread the tobacco stem material thinly using a combined material spreading process; The first identification module is used to throw the thinned tobacco stem material to the first image recognition area via the first conveyor belt, acquire double-sided tobacco stem images of the tobacco stem material during the material throwing process, and identify the target separator based on the feature differences of the target separator in the double-sided tobacco stem images; The first selection module is used to separate qualified tobacco stems from the identification results through a high-speed air valve and blow the target separated material to the second conveyor belt; The second identification module is used to throw the separated material to the second image recognition area via the second conveyor belt, collect single-sided or double-sided material images during the material throwing process, and perform secondary identification on qualified tobacco stems in the single-sided or double-sided material images based on preset features, so as to control the second selection module to perform secondary selection.

[0010] A third aspect of the present invention provides a computer-readable storage medium storing a computer program, which, when executed by a processor, implements the tobacco stem selection method.

[0011] A fourth aspect of the present invention provides a tobacco stem selection device, including a computer-readable storage medium and a processor, wherein the processor executes a computer program on the computer-readable storage medium to implement the steps of the tobacco stem selection method described above.

[0012] The beneficial effects of this invention are as follows: This invention provides a method, apparatus, medium, and device for selecting tobacco stems, which have the following beneficial effects: (1) In order to ensure the accuracy of visual recognition, the conveying state of tobacco stem material under normal production conditions is analyzed, and the material homogenization process is improved so that the tobacco stem material is spread thinly to a relatively dispersed and non-overlapping state before detection, while reducing the length of the thinning path to meet the on-site location requirements.

[0013] (2) Machine vision imaging technology is used to perform double-sided visual inspection during the material throwing process, and recognition rules are established by using the difference in double-sided imaging features between tobacco stems and unqualified products, which improves the accuracy of identifying target separation objects such as stem tangles, similar density impurities, and discolored tobacco stems in tobacco stem raw materials.

[0014] To make the above-mentioned objects, features and advantages of the invention more apparent and understandable, preferred embodiments of the invention are described below in detail with reference to the accompanying drawings. Attached Figure Description

[0015] To more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings used in the embodiments will be briefly introduced below. It should be understood that the following drawings only show some embodiments of the present invention and should not be regarded as a limitation on the scope. For those skilled in the art, other related drawings can be obtained based on these drawings without creative effort.

[0016] Figure 1 This is a flowchart illustrating the tobacco stem selection method provided in Example 1; Figure 2 This is a schematic diagram of the tobacco stem selection device provided in Embodiment 2; Figure 3 This is a schematic diagram of the circuit structure of the tobacco stem refining device provided in Embodiment 3; Figure 4 yes Figure 2 The corresponding structural design diagram. Detailed Implementation

[0017] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention. All other embodiments obtained by those skilled in the art based on the embodiments of this invention without inventive effort are within the scope of protection of this invention.

[0018] It should be noted that, unless otherwise specified, the various features in the embodiments of this invention can be combined with each other, all of which are within the protection scope of this invention. Furthermore, although functional modules are divided in the device schematic diagram and a logical order is shown in the flowchart, in some cases, the steps shown or described can be executed in a different order than the module division in the device or the order in the flowchart. Moreover, the terms "first," "second," and "third" used in this invention do not limit the data or execution order, but only distinguish identical or similar items with essentially the same function and effect.

[0019] Please see Figure 1 This is a flowchart illustrating a method for selecting tobacco stems according to Embodiment 1 of the present invention. Figure 1 As shown, the method includes the following steps: Step 1: Spread the tobacco stem material thinly using a combined homogenization process; Step 2: The thinned tobacco stem material is thrown to the first image recognition area via the first conveyor belt. During the material throwing process, double-sided tobacco stem images of the tobacco stem material are acquired, and the target separator is identified based on the feature differences of the target separator in the double-sided tobacco stem images. Step 3: Separate the qualified tobacco stems from the identification results using a high-speed air valve, and blow the target separated material to the second conveyor belt; Step 4: The separated material is thrown to the second image recognition area via the second conveyor belt. During the material throwing process, single-sided or double-sided material images are collected, and qualified tobacco stems in the single-sided or double-sided material images are re-identified and selected based on preset features.

[0020] The above embodiments provide a method for selecting tobacco stems. By combining a homogenization process, the tobacco stem material is spread thinly to a dispersed and non-overlapping state, which not only ensures the accuracy of visual recognition but also reduces the length of the spreading path. At the same time, double-sided visual inspection is performed during the material throwing process, and recognition rules are established by using the differences in double-sided imaging features between qualified and unqualified tobacco stems. This improves the accuracy of identifying stem crooks, impurities with similar density, and tobacco stems with different colors in the tobacco stem raw material, and greatly improves the quality of tobacco shreds.

[0021] The above steps will be explained in detail below through specific embodiments.

[0022] Preliminary research revealed that, constrained by factors such as space, production capacity, and imaging effects, a single material homogenization method could not achieve the objectives of this invention. Therefore, this invention employs a combination of multiple material homogenization methods to achieve uniform material spreading before imaging. In a preferred embodiment, the combined material homogenization process includes a vibratory groove material homogenization process, a sliding plate material homogenization process, and a differential speed material homogenization process arranged sequentially. Step 1 specifically includes: Step 101: Adjust the corresponding material leveling parameters according to the preset material leveling mode. For example, the preset material leveling mode can be preset according to different selection accuracies required. The material leveling parameters include the setting angle of the guide strip in the material leveling vibrating trough, the motor vibration frequency of the material leveling vibrating trough, the tilt angle of the slide plate, and the initial running speed of the first conveyor belt.

[0023] Step 102: Control the tobacco stem material to fall into the uniform material vibrating trough. The guide strips in the trough guide the tobacco stem material to disperse to the left and right sides along a herringbone angle. The basic principle of the vibrating trough uniform material distribution is to use a motor vibrating at a suitable frequency and to lay a certain number of herringbone dividing strips with a certain vertex angle in the direction of material movement, so that when the material falls, the relatively concentrated tobacco stems are dispersed to the left and right sides along a herringbone angle.

[0024] Step 103: Control the dispersed tobacco stem material to fall into the sliding plate under the action of inertia and gravity and slide downwards, so that the overlapping tobacco stem material is dispersed again by speed difference and falls into the first conveyor belt. The basic principle of sliding plate material uniformization is that the tobacco stems slide downwards under the action of inertia and gravity, and the overlapping materials are naturally separated by speed difference. This method is very simple to implement and can be used as an auxiliary method when a high uniformity effect is required. In this invention, the tobacco stem material is already in a higher position before entering the first conveyor belt, which is suitable for the direct use of this uniformization method.

[0025] Step 104: Collect the current horizontal velocity of the tobacco stem material during its descent and its current spreading state on the first conveyor belt. Compare the current spreading state with the target spreading state, and adjust the current operating speed of the first conveyor belt based on the comparison result. This allows the tobacco stem material to be dispersed three times through speed difference, achieving the target spreading density on the first conveyor belt. The basic principle of differential speed material uniformization is to utilize the speed difference between the material on two conveyor belts with different speeds to create a mass difference in a certain planar area. For example, based on the inclination angle set by the slide plate, the horizontal velocity of the tobacco stem material falling along the slide plate is approximately 1.4 m / s. Therefore, the initial operating speed of the first conveyor belt can be set to 2.7 m / s. By collecting the current horizontal velocity and current spreading state of the tobacco stems during the descent, the operating speed of the first conveyor belt can be adjusted accordingly. This differential speed conveying reduces the spreading density of the tobacco stem material per unit area to half, which is beneficial for subsequent imaging and recognition.

[0026] In a preferred embodiment, the first conveyor belt and the second conveyor belt are arranged vertically. The target material blown off from the first conveyor belt falls onto the second conveyor belt and is transported in the opposite direction to the second image recognition area, thereby reducing the spreading path length and meeting the on-site location requirements. Specifically, the tobacco stem material falls into the uniform vibrating trough. During the conveying process, it is gradually dispersed by the guide strips of the vibrating trough and evenly distributed throughout the trough. Due to the short length of the vibrating trough, the tobacco stems are not evenly distributed when they exit the trough. Then, the tobacco stems that have filled the vibrating trough fall onto the sliding plate, where they are accelerated to spread evenly. The evenly distributed tobacco stems fall into the first conveyor belt, where the high-speed conveyor belt further disperses the falling tobacco stems to facilitate subsequent selection. The entire vibration mode adopts a high-frequency, low-amplitude vibration mode to spread the material as evenly as possible.

[0027] In another preferred embodiment, as exemplified, step 2 involves identifying the target separator based on feature differences in the double-sided tobacco stem image, specifically as follows: Step 201: A first camera and a second camera are respectively set at symmetrical positions on both sides of the tobacco stem in the first image recognition area. For example, a light source is simultaneously set at each camera. Since different materials have different absorption rates for different spectra, different materials will exhibit different levels of brightness under the same light source illumination. Therefore, by selecting the correct light source spectrum, they can be clearly distinguished. This invention uses visible light imaging mainly to better acquire the image feature differences of tobacco stems, stem stalks, impurities of similar density, and carbonized or dark-colored tobacco stems.

[0028] Step 202: During the material throwing process, the first camera and the second camera simultaneously acquire images of the front and back sides of the tobacco stem material, respectively, and mark them.

[0029] Step 203: Using a preset image classification model and based on preset features, the target separators in the front and back tobacco stem images are initially identified, generating the type, front feature parameter value and back feature parameter value of each target separator, and calculating the feature difference between the front feature parameter value and the back feature parameter value.

[0030] For example, deep learning methods can be used to build a neural network model for identifying target separations, including stems, similar density debris, and discolored tobacco stems, such as carbonized or dark-colored tobacco stems, to perform image recognition and object classification. Model training methods are extensively described in the prior art and will not be elaborated here.

[0031] For example, the preset features used by the present invention to identify tobacco stems and target separated objects include shape features, color features, or texture features. For instance, shape features are used to identify stem tangles, color features are used to identify tobacco stems with different colors, texture features are used to identify impurities with similar density, and a combination of shape features, color features, and texture features is used to identify qualified tobacco stems, etc.

[0032] Specifically, color features are among the most intuitive and widely used visual features in images. Commonly used color feature extraction methods include: color space transformation, which converts the RGB color space to other color spaces (such as HSV and Lab) to better describe color characteristics; color histograms, which statistically analyze the frequency of each color in an image; and color statistical features, which use low-order moments of color (such as mean, variance, and frequency of color intensity values) to describe color distribution. For example, by analyzing the intensity distribution of qualified tobacco stems, dark-colored tobacco stems, and debris samples across different channels, it can be found that qualified tobacco stems have better distinguishability than dark-colored tobacco stems and debris in the G and B channels, while qualified tobacco stems and dark-colored tobacco stems have better distinguishability than debris in the S channel.

[0033] Shape features primarily focus on the outline and shape of an object. Outline features mainly describe the outer boundary of an object, typically found in a binary image. The outline is a continuous curve representing the basic shape of the object. Distinguishing features like stubble, stubble, and miscellaneous objects is mainly based on outline and region features. Outline features can be described using area, perimeter, centroid, convex hull, and minimum bounding rectangle. Region features focus on the entire shape region and are typically described using the following methods: Geometric parametric methods, which describe the shape by calculating quantitative measurement parameters (such as moments, area, perimeter, etc.); Shape-invariant moment methods, which utilize the moment invariance of the image to describe the shape, and are invariant to translation, scaling, and rotation; and Hu moments, a commonly used type of shape-invariant moment, which describes the shape by calculating the zeroth-order moment, first-order moment, etc., of the image.

[0034] Finally, step 204 is executed to determine whether the characteristic differences meet the preset conditions corresponding to the target separation type. Based on the determination result, the initial identification result is optimized so that qualified tobacco stems in the optimized identification result are blown to the next process through a high-speed air valve. Otherwise, they are blown to the second conveyor belt as target separations. For example, the preset conditions here can be established based on historical data of different target separations. At the same time, considering the disordered orientation of tobacco stems during the conveying process and the long and thin physical characteristics of tobacco stems, this embodiment uses a high-speed air valve to regulate the conveying posture of tobacco stems, that is, to maintain a consistent relative direction during the throwing process, thereby effectively reducing the cross-directional situation of some tobacco stems during the rejection action, reducing the rejection of tobacco stems by mistake, and further realizing the accurate rejection of unqualified tobacco stems and impurities.

[0035] In a preferred embodiment, multiple continuously changing images of the front or back of the tobacco stem can be acquired to obtain the motion trajectory of the same target separation object during the throwing process. The motion trajectory of the target separation object can be compared with the reference motion trajectory of the corresponding type. The comparison result can be used to help determine whether the initial identification result or the optimized identification result of the target separation object is correct, thereby further improving the tobacco stem selection effect.

[0036] As those skilled in the art will know, in order to achieve better technical results, for example, in step 4, qualified tobacco stems in the single-sided or double-sided material images can be re-identified and selected based on preset features. In specific implementation, either single-sided or double-sided material images can be selected, but joint identification using double-sided material images yields better accuracy. The following uses a double-sided material image as an example, specifically including the following steps: Step 401: Set up a third camera and a fourth camera at symmetrical positions on both sides of the tobacco stem in the second image recognition area; Step 402: During the material throwing process, the third camera and the fourth camera simultaneously acquire images of the front and back of the separated material, respectively, and mark them. Step 403: Using a preset image classification model and based on the preset features, identify qualified tobacco stems in the front material image and the back material image respectively; Step 404: Obtain the classification results. If the object at the same location is identified as a qualified tobacco stem in both the front material image and the back material image, it is blown to the next process through a high-speed air valve. Otherwise, it is classified and recycled as a target separation material.

[0037] It should be noted that in the above embodiments, there is no necessarily a certain order between the above steps. Those skilled in the art can understand from the description of the embodiments of the present invention that the above steps may have different execution orders in different embodiments, that is, they may be executed in parallel or in interchange, etc.

[0038] As another aspect of this invention, this embodiment also provides a tobacco stem selection device. The tobacco stem selection device can be a software module, which includes several instructions stored in a memory. A processor can access the memory and execute the instructions to complete the tobacco stem selection method described in the above embodiments.

[0039] In some embodiments, the tobacco stem selection device can also be constructed from hardware components. For example, the tobacco stem selection device can be constructed from one or more chips, which can work in coordination to complete the tobacco stem selection method described in the various embodiments above. As another example, the tobacco stem selection device can also be constructed from various logic devices, such as general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), microcontrollers, ARM (AcornRISC) or other programmable logic devices, discrete gate or transistor logic, discrete hardware components, or any combination of these components.

[0040] Figure 2 , Figure 4 This is a schematic diagram of the structure of a tobacco stem selection device provided in Embodiment 2 of the present invention. The tobacco stem selection device includes a uniform material module 100, a first identification module 200, a first selection module 300, a second identification module 400 and a second selection module 500, and a second conveyor belt 700.

[0041] The material leveling module 100 is used to spread the tobacco stem material thinly using a combined material leveling process, including a vibrating trough 101 and a herringbone guide strip 102, a sliding plate 103, and a first conveyor belt 104 disposed in the vibrating trough. The first identification module 200 is used to throw the thinned tobacco stem material to the first image recognition area via the first conveyor belt, acquire double-sided tobacco stem images of the tobacco stem material during the material throwing process, and identify the target separator based on the feature differences of the target separator in the double-sided tobacco stem images; The first selection module 300 is used to separate qualified tobacco stems from the identification results through a high-speed air valve and blow the target separated material to the second conveyor belt; The second identification module 400 is used to throw the separated material to the second image recognition area via the second conveyor belt, collect single-sided or double-sided material images during the material throwing process, and perform secondary identification on qualified tobacco stems in the single-sided or double-sided material images based on preset features, so as to control the second selection module 500 to perform secondary selection.

[0042] The above embodiments provide a tobacco stem selection device that uses a combined homogenization process to spread tobacco stem material into a dispersed and non-overlapping state, which not only ensures the accuracy of visual recognition but also reduces the length of the spreading path. At the same time, double-sided visual inspection is performed during the material throwing process, and recognition rules are established by using the differences in double-sided imaging features between qualified and unqualified tobacco stems. This improves the accuracy of identifying stem crooks, similar density impurities, and discolored tobacco stems in the tobacco stem raw material, and greatly improves the quality of tobacco shreds.

[0043] For example, in a preferred embodiment, the uniform material module 100 specifically includes: The setting unit is used to adjust the corresponding material uniformization parameters according to the preset material uniformization mode. The first control unit is used to control the tobacco stem material to fall into the uniform vibrating trough, and to disperse the tobacco stem material to the left and right sides along the herringbone angle through the guide bar of the vibrating trough. The second control unit is used to control the dispersed tobacco stem material to fall into the slide plate and slide downward under the action of inertia and gravity, so that the overlapping tobacco stem material is dispersed a second time by speed difference and falls into the first conveyor belt. The third control unit is used to collect the current horizontal speed of the tobacco stem material during the falling process and the current spreading state of the tobacco stem material on the first conveyor belt, compare the current spreading state with the target spreading state, and adjust the current running speed of the first conveyor belt according to the comparison result, so as to disperse the tobacco stem material three times through speed difference and make the spreading density of the tobacco stem material on the first conveyor belt reach the target spreading state.

[0044] For example, in a preferred embodiment, the first identification module 200 specifically includes: The first image acquisition unit is used to acquire, and mark, the front and back images of the tobacco stem material at the same time through the first camera and the second camera during the material throwing process; The first image recognition unit is used to perform initial recognition of target separators in the front tobacco stem image and the back tobacco stem image respectively using a preset image classification model and based on preset features, generate the type, front feature parameter value and back feature parameter value of each target separator, and calculate the feature difference between the front feature parameter value and the back feature parameter value; The fourth control unit is used to determine whether the feature difference meets the preset conditions corresponding to the target separation type, and optimize the initial identification result according to the judgment result, so as to blow the qualified tobacco stems in the optimized identification result to the next process through the high-speed air valve; otherwise, it is blown to the second conveyor belt as the target separation.

[0045] For example, in a preferred embodiment, the first identification module 400 specifically includes: The second image acquisition unit is used to acquire, and mark, the front and back images of the separated material at the same time using the third and fourth cameras during the material throwing process. The second image recognition unit is used to identify qualified tobacco stems in the front material image and the back material image respectively using a preset image classification model and based on the preset features; The fifth control unit is used to obtain the classification results. If an object at the same location is identified as a qualified tobacco stem in both the front material image and the back material image, it is blown to the next process through a high-speed air valve. Otherwise, it is classified and recycled as a target separated material.

[0046] It should be noted that the above-mentioned tobacco stem selection device can execute the tobacco stem selection method provided in the embodiments of the present invention, and has the corresponding functional modules and beneficial effects of the method. Technical details not described in detail in the embodiments of the tobacco stem selection device can be found in the tobacco stem selection method provided in the embodiments of the present invention.

[0047] Figure 3 This is a schematic diagram of the circuit structure of the tobacco stem selection device provided in Embodiment 3 of the present invention. Figure 3 As shown, the controller 600 includes one or more processors 61 and a memory 62. Wherein, Figure 3 Take a processor 61 as an example.

[0048] Processor 61 and memory 62 can be connected via a bus or other means. Figure 3 Taking the example of a connection between China and Israel via a bus.

[0049] The memory 62, as a non-volatile computer-readable storage medium, can be used to store non-volatile software programs, non-volatile computer-executable programs, and modules, such as the program instructions / modules corresponding to the tobacco stem selection method in the embodiments of the present invention. The processor 61 executes various functional applications and data processing of the tobacco stem selection device by running the non-volatile software programs, instructions, and modules stored in the memory 62, thereby realizing the functions of the tobacco stem selection method provided in the above method embodiments and the various modules or units in the above device embodiments.

[0050] Memory 62 may include high-speed random access memory, and may also include non-volatile memory, such as at least one disk storage device, flash memory device, or other non-volatile solid-state storage device. In some embodiments, memory 62 may optionally include memory remotely located relative to processor 61, which can be connected to processor 61 via a network. Examples of such networks include, but are not limited to, the Internet, intranets, local area networks, mobile communication networks, and combinations thereof.

[0051] The program instructions / modules are stored in the memory 62 and, when executed by one or more processors 61, execute the tobacco stem selection method in any of the above method embodiments.

[0052] This invention also provides a non-volatile computer storage medium storing computer-executable instructions that are executed by one or more processors, for example... Figure 3One of the processors 61 can enable the one or more processors to execute the tobacco stem selection method in any of the above method embodiments.

[0053] This invention also provides a tobacco stem selection device, including the aforementioned computer-readable storage medium and a processor, wherein the processor executes a computer program on the computer-readable storage medium to implement the steps of the tobacco stem selection method. This invention also provides a computer program product, which includes a computer program stored on a non-volatile computer-readable storage medium. The computer program includes program instructions that, when executed by an electronic device, cause the electronic device to perform any of the tobacco stem selection methods described above.

[0054] The apparatus or device embodiments described above are merely illustrative. The unit modules described as separate components may or may not be physically separate, and the components shown as module units may or may not be physical units; that is, they may be located in one place or distributed across multiple network module units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs.

[0055] Through the above description of the embodiments, those skilled in the art can clearly understand that each embodiment can be implemented using software plus a general-purpose hardware platform, or of course, using hardware. Based on this understanding, the above technical solutions, in essence or the parts that contribute to the related technology, can be embodied in the form of a software product. This computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, magnetic disk, optical disk, etc., and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute the methods described in the various embodiments or some parts of the embodiments.

[0056] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; under the concept of the present invention, the technical features of the above embodiments or different embodiments can also be combined, the steps can be implemented in any order, and there are many other variations of different aspects of the present invention as described above, which are not provided in detail for the sake of brevity; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method of tobacco stem culling, characterized by, The method comprises the following steps: Step 1: thinning the tobacco stem material by using a combined uniform material process; Step 2: throwing the thinned tobacco stem material to a first image recognition area by a first conveying belt, collecting double-sided tobacco stem images of the tobacco stem material during the throwing process, and identifying target separated objects based on feature differences of the target separated objects in the double-sided tobacco stem images; Step 3: separating and identifying qualified tobacco stems in the identification results by a high-speed air valve, and blowing the target separated objects to a next station.

2. The tobacco stem culling method according to claim 1, wherein, The combined uniform material process comprises a vibrating trough uniform material process, a sliding plate uniform material process and a differential speed uniform material process arranged in sequence, and the step 1 specifically comprises: adjusting corresponding uniform material parameters according to a preset uniform material mode; controlling the tobacco stem material to fall into the uniform material vibrating trough, and making the tobacco stem material disperse to left and right sides along a herringbone angle by a vibrating trough guide strip; controlling the dispersed tobacco stem material to fall into a sliding plate and slide downward under the action of vibration and gravity, so that the tobacco stem material overlapped together is dispersed for the second time by speed difference and falls into the first conveying belt; collecting a current horizontal speed of the tobacco stem material during the falling process and a current thinning state of the tobacco stem material on the first conveying belt, comparing the current thinning state with a target thinning state, and adjusting a current running speed of the first conveying belt according to a comparison result, so that the tobacco stem material is dispersed for the third time by speed difference, and the flat laying density of the tobacco stem material on the first conveying belt reaches the target thinning state.

3. The tobacco stem culling method according to claim 2, wherein, The uniform material parameters comprise a setting angle of the vibrating trough guide strip in the uniform material vibrating trough, a motor vibration frequency of the uniform material vibrating trough, an inclination angle of the sliding plate and an initial running speed of the first conveying belt.

4. The tobacco stem sorting method according to claim 1, wherein, In the step 2, the target separated objects are identified based on feature differences of the target separated objects in the double-sided tobacco stem images, specifically: first and second cameras are arranged at symmetrical positions on both sides of the tobacco stem in the first image recognition area; during the throwing process, the first and second cameras collect a front tobacco stem image and a back tobacco stem image of the tobacco stem material at the same time and mark the images; a preset image classification model is used to initially identify the target separated objects in the front and back tobacco stem images based on preset features, type, front feature parameter value and back feature parameter value of each target separated object are generated, and feature differences of the front and back feature parameter values are calculated; it is judged whether the feature differences meet preset conditions corresponding to the type of the target separated object, and the initial identification result is optimized according to a judgment result, so that qualified tobacco stems in the optimized identification result are blown to a next process by the high-speed air valve, otherwise, the target separated objects are blown to the next station.

5. The tobacco stem culling method according to any one of claims 1 to 4, characterized in that, Further comprising: Step 4: the next station is a second conveying belt, the second conveying belt throws the separated material to a second image recognition area, collects single-sided or double-sided material images during the throwing process, and performs secondary identification and selection on qualified tobacco stems in the single-sided or double-sided material images based on preset features. The first conveying belt and the second conveying belt are arranged in an up-down manner, and the target separated objects falling from the first conveying belt fall onto the second conveying belt and are transported in a reverse direction to the second image recognition area.

6. The tobacco stem culling method according to claim 4 or 5, wherein, The step 4 comprises a second identification and selection of qualified tobacco stems in the single-sided or double-sided material image based on preset features, specifically: A third camera and a fourth camera are arranged at symmetrical positions on both sides of the tobacco stems in the second image recognition area; During the material throwing process, the third camera and the fourth camera are used to capture a front material image and a back material image of the separated material at the same time, and the images are labeled; A preset image classification model is used to identify the qualified tobacco stems in the front material image and the back material image based on the preset features; Classification results are obtained, and if the objects at the same position are identified as qualified tobacco stems in both the front material image and the back material image, the objects are blown to the next process by a high-speed air valve, otherwise, the objects are classified and recycled as target separated objects.

7. The tobacco stem culling method according to claim 6, wherein, The target separated objects include stem hooks, density similar impurities, and color different tobacco stems, the preset features include shape features, color features, or texture features, the stem hooks are identified by the shape features, the color different tobacco stems are identified by the color features, the density similar impurities are identified by the texture features, and the qualified tobacco stems are identified by a combination of the shape features, the color features, and the texture features.

8. A tobacco stem sorting device based on the method of any one of claims 1-7, wherein, The method comprises a uniform material module, a first identification module, a first selection module, a second identification module, and a second selection module, The uniform material module is used to thin the tobacco stem material by a combined uniform material process; The first identification module is used to throw the thinned tobacco stem material to a first image recognition area by a first conveying belt, capture double-sided tobacco stem images of the tobacco stem material during the material throwing process, and identify target separated objects based on feature differences of the target separated objects in the double-sided tobacco stem images; The first selection module is used to separate qualified tobacco stems in the identification results by a high-speed air valve, and blow the target separated objects to a second conveying belt; The second identification module is used to throw the separated material to a second image recognition area by a second conveying belt, capture single-sided or double-sided material images during the material throwing process, and perform a second identification of qualified tobacco stems in the single-sided or double-sided material images based on preset features, so as to control the second selection module to perform a second selection.

9. A computer readable storage medium storing a computer program, characterized in that, The computer program is executed by the processor to implement the tobacco stem selection method of any one of claims 1-7.

10. A tobacco stem culling apparatus characterized by, The computer readable storage medium and the processor of claim 9 are used to implement the steps of the tobacco stem selection method of any one of claims 1-7 when the processor executes the computer program on the computer readable storage medium.

Images