An intelligent automatic detection and removal system for tobacco impurities

Abstract

The present invention provides an intelligent automatic detection and removal system for impurities in tobacco shreds, which is characterized by comprising: an impurity detection module for detecting impurities in tobacco raw materials, thereby detecting and obtaining impurity detection results, and the impurity detection results include impurity position information; and an impurity removal module for removing impurities from the tobacco raw materials according to the impurity position information to obtain clean tobacco shreds. Among them, the impurity detection module detects the tobacco raw materials according to the color characteristics, shape characteristics, texture characteristics and spectral characteristics of the impurities to obtain the impurity detection results. The intelligent automatic detection and removal system for impurities in tobacco shreds of the present invention can automatically remove impurities such as metals, stones, hemp ropes, paper pieces, cocoons, foam, stems, moldy tobacco leaves, and wet tobacco shreds in tobacco raw materials, thereby improving the efficiency of impurity removal, saving labor costs, and improving the efficiency of subsequent cigarette rolling processes and the quality of cigarettes.

Description

Technical Field

[0001] The invention belongs to the technical field of tobacco processing, and in particular relates to an intelligent automatic detection and removal system and device for tobacco impurities. Background Art

[0002] Tobacco raw materials are generally mixed with metals, stones, hemp ropes, paper scraps, insect cocoons, foam, moldy tobacco leaves and other impurities. Removing impurities from tobacco is one of the most important processes in the production of modern cigarette products. Its main functions are: preventing heavy impurities such as metals and stones from causing irreversible damage to subsequent automated equipment; preventing hemp ropes and insect cocoons from affecting the quality of cigarettes; preventing light impurities such as foam and moldy tobacco leaves from affecting the taste and endangering health.

[0003] At present, foreign tobacco foreign body removal systems mostly use laser scanning or CCD camera photography to obtain detection signals, while domestic research on the identification and removal of impurities in tobacco is relatively rare. The tobacco foreign body identification and removal automation equipment of various tobacco companies adopts a detection method based on color features, and the recognition accuracy is not high, which seriously restricts the development of the tobacco industry. Summary of the invention

[0004] In order to solve the above problems, a tobacco impurity identification and removal system with high identification efficiency is provided. The present invention adopts the following technical solutions:

[0005] The present invention provides an intelligent automatic detection and removal system for tobacco shredded impurities, characterized in that it comprises: an impurity detection module, which is used to detect impurities in tobacco raw materials, thereby obtaining impurity detection results, and the impurity detection results include impurity position information; and an impurity removal module, which removes impurities from the tobacco raw materials according to the impurity position information to obtain clean tobacco shreds, wherein the impurity detection module comprises a first detection unit, a second detection unit and a third detection unit, the first detection unit detects the tobacco raw materials according to the color characteristics and shape characteristics of the impurities to obtain a first detection result, the second detection unit detects the tobacco raw materials according to the texture characteristics of the impurities to obtain a second detection result, the third detection unit detects the tobacco raw materials according to the spectral characteristics of the impurities to obtain a third detection result, and the impurity removal module removes impurities according to the first detection result, the second detection result and the third detection result, respectively.

[0006] In an intelligent automatic detection and removal system for impurities in tobacco shreds provided by the present invention, it may also have such features, and further include: a tobacco preliminary screening module, used for preliminary screening of original tobacco containing different impurities to obtain tobacco raw materials; and a tobacco flattening and transportation module, used for flattening the tobacco raw materials and conveying them to the impurity detection module, wherein the tobacco preliminary screening module has a screening conveyor and a screening negative pressure device, and the tobacco flattening and transportation module has a feeding funnel, a dropper, a flattening transportation mechanism and a height limiting mechanism, the dropper is arranged at the outlet of the feeding funnel, and the height limiting mechanism is arranged Near the flattening transport mechanism, the screening conveyor is used to transport the raw tobacco to the bottom of the screening negative pressure device. The screening negative pressure device absorbs the raw materials with a smaller specific gravity in the raw tobacco as tobacco raw materials, and transports the tobacco raw materials to the feeding funnel. The dropper is used to control the falling flow rate of the tobacco raw materials, so that the tobacco raw materials are evenly flattened on the flattening transport mechanism. The flattening transport mechanism has multiple conveyors for transporting tobacco raw materials and flowing through the first detection part, the second detection part and the third detection part. The height limiting mechanism is used to spread out the overlapping tobacco raw materials, so as to ensure that all tobacco raw materials are completely flattened.

[0007] In an intelligent automatic detection and removal system for tobacco impurities provided by the present invention, it can also have such a feature, wherein the flattening transportation mechanism has at least 4 conveyors, namely the first conveyor, the second conveyor, the third conveyor and the fourth conveyor, the first conveyor is located below the dropper, the second conveyor is located below the first detection part, the third conveyor is located below the second detection part, and the fourth conveyor is located below the third detection part, and the corresponding heights of the first conveyor, the second conveyor, the third conveyor and the fourth conveyor are successively reduced, and the corresponding transmission speeds are successively accelerated.

[0008] In an intelligent tobacco impurity automatic detection and removal system provided by the present invention, it can also have such a feature, wherein the first detection part at least has a first image acquisition unit and a first detection model detection unit, the second detection part at least has a second image acquisition unit and a second detection model detection unit, and the third detection part at least has a third image acquisition unit and a third detection model detection unit, the first image acquisition unit acquires an RGB image of the tobacco raw material as a first image through a camera arranged above the tobacco raw material, and the first detection model detection unit detects the first image according to color and shape features through a pre-trained first impurity detection model to obtain a first detection result, and the first detection model detection unit The detection result includes the first impurity position information. The second image acquisition unit acquires multiple RGB images of the tobacco raw material from different angles as the second image through multiple cameras arranged above the tobacco raw material. The second detection model detection unit detects all the second images according to the texture features through a pre-trained second impurity detection model, so as to obtain a second detection result, which includes the second impurity position information. The third image acquisition unit acquires the spectral image of the tobacco raw material as the third image. The third detection model detection unit detects the third image according to the spectral features through a pre-trained third impurity detection model, so as to obtain a third detection result, which includes the third impurity position information.

[0009] In an intelligent automatic detection and removal system for tobacco impurities provided by the present invention, it can also have such a feature, wherein the first detection unit, the second detection unit and the third detection unit also have a light shield for shielding the tobacco raw materials from external natural light and a plurality of light sources arranged in the light shield for providing visible light for the tobacco raw materials.

[0010] In an intelligent tobacco impurity automatic detection and removal system provided by the present invention, it can also have such features, and also includes: a control module, wherein the impurity removal module has a negative pressure impurity removal device, a robotic arm and an impurity recovery container, the negative pressure impurity removal device has at least a negative pressure machine and a plurality of negative pressure suction heads, the negative pressure machine provides power for all negative pressure suction heads, the impurity recovery container is connected to the negative pressure suction head through a pipeline, and the negative pressure suction head is arranged at one end of the robotic arm, when the first detection model detection unit obtains the first impurity position information, the control module controls the movement of the robotic arm located above the second conveyor according to the first impurity position information, so that the negative pressure suction head moves to above the impurity corresponding to the first impurity position information, and sucks the impurity out And it is transported to the impurity recovery container through a pipeline. When the second detection model detection unit obtains the second impurity position information, the control module controls the movement of the mechanical arm located above the third conveyor according to the second impurity position information, so that the negative pressure suction head moves to the impurity corresponding to the second impurity position information, sucks out the impurity and transports it to the impurity recovery container through a pipeline. When the third detection model detection unit obtains the third impurity position information, the control module controls the movement of the mechanical arm located above the third conveyor according to the third impurity position information, so that the negative pressure suction head moves to the impurity corresponding to the third impurity position information, sucks out the impurity and transports it to the impurity recovery container through a pipeline, thereby completing the impurity removal.

[0011] In an intelligent tobacco impurity automatic detection and removal system provided by the present invention, it can also have such a feature, wherein the image corresponding to the tobacco raw material is the original image, and the detection process of the impurity detection module includes the following steps: step S1-1, performing mean suppression processing on the original image, so as to obtain a mean image in which the number of remaining non-zero pixels in the original image is less than a predetermined number threshold; step S1-2, segmenting the mean image to obtain a segmented image; step S1-3, performing connected domain detection on the segmented image, so as to obtain each connected domain, and calculating the minimum circumscribed circle of each connected domain; step S1-4, judging in turn whether the radius of the minimum circumscribed circle is less than a predetermined radius threshold; step S1-5, when step S1-4 judges to be no, the minimum circumscribed circle is the impurity circumscribed circle, the substance surrounded by the impurity circumscribed circle is the impurity, and the impurity position information is obtained according to the impurity circumscribed circle.

[0012] The intelligent tobacco impurity automatic detection and removal system provided by the present invention may also have such a feature that the segmented image is obtained by using the OSTU algorithm.

[0013] Function and Effect of the Invention

[0014] According to an intelligent automatic detection and removal system for tobacco impurities of the present invention, since the first detection unit detects the tobacco raw material according to the color characteristics and shape characteristics of the impurities to obtain a first detection result, the impurity removal module can remove impurities (such as foam, plastic, metal, stone, paper, moldy tobacco, tobacco stems, etc.) whose colors or contours are obviously different from those of tobacco according to the first detection result. Also, since the second detection unit detects the tobacco raw material according to the texture characteristics of the impurities to obtain a second detection result, the impurity removal module can remove impurities (such as hemp rope, cocoon, stem stick, etc.) whose textures are obviously different from those of tobacco according to the second detection result. In addition, since the third detection unit detects the tobacco raw material according to the spectral characteristics of the impurities to obtain a third detection result, the impurity removal module can remove impurities (such as leaves, grass leaves, etc.) whose spectral characteristics are obviously different from those of tobacco according to the third detection result to obtain purer tobacco and improve tobacco quality.

[0015] The intelligent automatic detection and removal system for tobacco impurities of the present invention can automatically remove impurities such as metals, stones, hemp ropes, paper pieces, insect cocoons, foam, stems, moldy tobacco leaves, and wet tobacco from tobacco raw materials, thereby improving the efficiency of impurity removal, saving labor costs, and improving the efficiency of subsequent cigarette rolling processes and the quality of cigarettes. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] Figure 1 It is a structural block diagram of an intelligent tobacco impurity automatic detection and removal system according to an embodiment of the present invention;

[0017] Figure 2 It is a structural schematic diagram of an intelligent automatic detection and removal system for tobacco impurities according to an embodiment of the present invention;

[0018] Figure 3 A schematic diagram of the connection between the control module and other modules of an embodiment of the present invention;

[0019] Figure 4 is a schematic structural diagram of a first detection unit according to an embodiment of the present invention;

[0020] Figure 5 is a flow chart of a detection process according to an embodiment of the present invention;

[0021] Figure 6 A schematic diagram of a spectral image processing process according to an embodiment of the present invention; and

[0022] Figure 7 The present invention is a flowchart of the working process of the intelligent tobacco impurity automatic detection and removal system according to the embodiment of the present invention. DETAILED DESCRIPTION

[0023] In order to make the technical means, creative features, objectives and effects achieved by the present invention easy to understand, the following takes raw tobacco containing impurities such as metals, stones, hemp ropes, paper pieces, cocoons, foam, stems, moldy tobacco leaves, wet tobacco, etc. as an example, and combines the embodiments and drawings to specifically explain an intelligent automatic detection and removal system for tobacco impurities of the present invention.

[0024] <Example>

[0025] Figure 1 It is a structural block diagram of an intelligent tobacco impurity automatic detection and removal system according to an embodiment of the present invention; Figure 2 Schematic diagram of the structure of the intelligent tobacco impurity automatic detection and removal system according to an embodiment of the present invention. Figure 2 The portion framed by the dotted line frame b is a top view of the portion framed by the dotted line frame a.

[0026] like Figure 1 as well as Figure 2 As shown, an intelligent automatic detection and removal system 1 for tobacco impurities includes a tobacco preliminary screening module 11, a tobacco flattening and transportation module 12, an impurity detection module 13, an impurity removal module 14 and a control module 15.

[0027] The tobacco preliminary screening module 11 can perform preliminary screening on raw tobacco containing different impurities to obtain tobacco raw materials, and has a screening conveyor 111 and a screening negative pressure device 112 .

[0028] The tobacco flattening and transporting module 12 can flatten the tobacco raw material and transport it to the impurity detection module 13 , and comprises a loading hopper 121 , a dropper 122 , a flattening and transporting mechanism 123 and a height limiting mechanism 124 .

[0029] The impurity detection module 13 can detect and identify impurities in the tobacco raw material to obtain an impurity detection result, which includes impurity position information. The impurity detection module 13 has a first detection unit 131, a second detection unit 132, and a third detection unit 133. The first detection unit 131, the second detection unit 132, and the third detection unit 133 all have a light shield 1311 and a plurality of light sources 1312.

[0030] The first detection unit 131 also has a first image acquisition unit 1313 and a first detection model detection unit 1314, the second detection unit 132 also has a second image acquisition unit and a second detection model detection unit, and the third detection unit 133 also has a third image acquisition unit and a third detection model detection unit.

[0031] The impurity removal module 14 removes impurities from the tobacco raw material according to the impurity position information to obtain clean tobacco shreds, and comprises a negative pressure impurity removal device 141 , a mechanical arm 142 and an impurity recovery container 143 .

[0032] The negative pressure impurity removal device 141 at least comprises a negative pressure machine and a plurality of negative pressure suction heads.

[0033] The control module 15 can control the tobacco preliminary screening module 11, the tobacco flattening and transporting module 12, the impurity detection module 13 and the impurity removal module 14, and has a PLC, a variable frequency motor, a power supply, an overload protector, a relay, a switch, a stepper motor, etc. The PLC controls the running speed of the belt conveyor, the power of the negative pressure machine, and the running state of the mechanical arm 142 by controlling the speed of the variable frequency motor and the stepper motor. The switch controls the start and stop of the entire system, the power supply provides power for the variable frequency motor, the stepper motor, etc., and the overload protector protects the circuit safety to avoid the danger of circuit short circuit or overload.

[0034] Figure 3 This is a schematic diagram of the connection between the control module and other modules in an embodiment of the present invention.

[0035] like Figure 3 As shown, the PLC is respectively connected to the variable frequency motor of the belt conveyor of the flattening transport mechanism 123, the negative pressure machine of the negative pressure impurity removal device 141 and the mechanical arm 142. The change of the variable frequency motor directly affects the speed of the conveyor belt of the belt conveyor, and the rotation of the mechanical arm 142 directly drives the rotation of the suction head.

[0036] The screening negative pressure device 112 is arranged above the screening conveyor 111 (specifically at the tail end of the screening conveyor 111 in the transport direction), and the thickness of the original tobacco spread on the screening conveyor 111 is close to the distance between the screening negative pressure device 112 and the screening conveyor 111. The screening negative pressure device 112 can be connected to a plurality of tobacco flattening and transporting modules 12.

[0037] In this embodiment, the screening negative pressure device 112 at least comprises a negative pressure machine and a negative pressure suction head, the negative pressure machine provides power, the negative pressure suction head is aimed at the original tobacco for adsorption, and the screening negative pressure device 112 is connected to the two tobacco flattening and transporting modules 12. The screening conveyor 111 is a vibrating conveyor.

[0038] When the screening conveyor 111 conveys the raw tobacco to the bottom of the screening negative pressure device 112, the screening negative pressure device 112 adsorbs the raw tobacco, leaving impurities with higher specific gravity (such as heavy metals, stones, stems, wet tobacco and other impurities) on the screening conveyor 111, and adsorbs the remaining raw tobacco (i.e., tobacco raw materials) to the tobacco flattening and transportation module, thereby completing the preliminary screening.

[0039] The feeding funnel 121 is connected to the negative pressure device 112 for screening. The negative pressure device 112 for screening transports the tobacco raw materials into the feeding funnel 121. The feeding funnel 121 can temporarily store the tobacco raw materials.

[0040] The dropper 122 is arranged at the outlet of the loading hopper 121, so that the tobacco raw materials maintain a certain thickness (the thickness at which all the tobacco raw materials are spread out as completely as possible) when they fall to the flattening transportation mechanism 123.

[0041] The control module 15 adjusts the dropping speed of the dropper 122 so that the tobacco raw material can be evenly spread on the flattening transportation mechanism 123 and ensures that the thickness of the tobacco raw material meets the needs of subsequent detection and identification.

[0042] The flattening transport mechanism 123 has a plurality of belt conveyors for transporting tobacco raw materials and flowing through the impurity detection module 13 .

[0043] The above-mentioned belt conveyors are composed of a frame, a conveying belt, active and passive belt rollers, a tensioning device, a transmission device, a reduction motor, etc. The reduction motor is used for driving and can realize frequency conversion speed regulation. The conveying belt can also be equipped with accessories such as a height limiting mechanism 124 and a skirt. The height limiting mechanism 124 is used to adjust the thickness of the material, and the skirt is used to protect the material to prevent leakage and waste of the material.

[0044] In this embodiment, the flattening transport mechanism 123 has a total of 5 belt conveyors, including a first conveyor 1231 arranged below the loading hopper 121, a second conveyor 1232 arranged below the first detection section 131, a third conveyor 1233 arranged below the second detection section 132, a fourth conveyor 1234 arranged below the third detection section 133, and a fifth conveyor 1235 arranged behind the fourth conveyor 1234, which are used to transport clean tobacco.

[0045] The corresponding heights of the first conveyor 1231, the second conveyor 1232, the third conveyor 1233 and the fourth conveyor 1234 are successively reduced. When the tobacco raw materials roll from the previous conveyor to the next conveyor, the materials on the lower surface of the previous conveyor can roll to the upper surface of the next conveyor.

[0046] In addition, the transmission speeds of the first conveyor 1231, the second conveyor 1232, the third conveyor 1233 and the fourth conveyor 1234 are successively accelerated, so that adjacent tobacco raw materials can be further separated to avoid stacking and blocking, thereby improving the accuracy of subsequent detection. The above transmission speeds can be adjusted.

[0047] The height limiting mechanism 124 is disposed above the flattening transport mechanism 123 and can flatten tobacco raw materials whose height is higher than a predetermined height threshold (ie, overlapping tobacco raw materials).

[0048] Figure 4 Schematic diagram of the structure of the first detection unit according to an embodiment of the present invention.

[0049] like Figure 4 As shown, the shading hood 1311 is suspended on the flattening transport mechanism 123 and maintains a predetermined distance from the belt conveyor to shield the tobacco raw materials from external natural light. The light source 1312 is arranged in the shading hood 1311 and provides visible light for the tobacco raw materials. The first image acquisition unit 1313 is arranged above the second conveyor 1232. The acquired first image is transmitted to the first detection model detection unit 1314 via a network cable. The first detection model detection unit 1314 is located on the PC end.

[0050] The negative pressure machine provides power for all negative pressure suction heads by means of negative pressure extraction. The impurity recovery container 143 is connected to all negative pressure suction heads through pipes, and the negative pressure suction heads are arranged at one end of the mechanical arm 142.

[0051] The first image acquisition unit 1313 acquires an RGB image of the tobacco raw material as a first image through a camera arranged above the tobacco raw material.

[0052] The first detection model detection unit 1314 detects the first image according to the color and shape features through a pre-trained first impurity detection model, thereby obtaining a first detection result, which includes first impurity position information.

[0053] The control module 15 controls the movement of the robot arm 142 located above the second conveyor 1232 according to the first impurity position information, so that the negative pressure suction head moves to the top of the impurity corresponding to the first impurity position information, sucks out the impurity and transports it to the impurity recovery container 143 through a pipeline, and the remaining tobacco raw material is the raw material after the initial removal.

[0054] The second image acquisition unit acquires multiple RGB images of the raw material after the initial rejection from different angles as the second image through multiple cameras arranged above the tobacco raw material (ie, the raw material after the initial rejection).

[0055] In this embodiment, the second image acquisition unit acquires multiple RGB images of the raw material after the initial rejection from different angles through three cameras as the second image.

[0056] The second detection model detection unit detects all second images according to the texture features through a pre-trained second impurity detection model, thereby obtaining a plurality of second detection results, which include second impurity position information.

[0057] The control module 15 controls the movement of the robot arm 142 located above the third conveyor 1233 according to the second impurity position information, so that the negative pressure suction head moves to the top of the impurity corresponding to the second impurity position information, sucks out the impurity and transports it to the impurity recovery container 143 through a pipeline, and the remaining raw materials after the first removal are the raw materials after the second removal.

[0058] The third image acquisition unit acquires a spectral image of the tobacco raw material (ie, the raw material after secondary elimination) as the third image.

[0059] The third detection model detection unit detects the third image through a pre-trained third impurity detection model according to the spectral features, thereby obtaining a third detection result, which includes third impurity position information.

[0060] The control module 15 controls the movement of the robot arm 142 located above the third conveyor 1233 according to the third impurity position information, so that the negative pressure suction head moves to the top of the impurity corresponding to the third impurity position information, sucks out the impurity and transports it to the impurity recovery container 143 through a pipeline, thereby completing the impurity removal, and the remaining raw material after the secondary removal is clean tobacco.

[0061] In addition, the clean tobacco in the impurity recovery container 143 can be further manually sorted for impurities and the mistakenly removed tobacco can be mixed back into the finished product, or automatically sent to the vibrating conveyor 111 for further impurity identification and removal to reduce the mistaken rejection rate.

[0062] Figure 5 is a flow chart of a detection process according to an embodiment of the present invention; Figure 6 Schematic diagram of the spectral image processing process according to an embodiment of the present invention.

[0063] The image corresponding to the tobacco raw material in each of the first detection model detection unit 1314, the second detection model detection unit and the third detection model is called the original image. At the same time, the detection process of each detection unit is the same, as shown in the following example. Figure 5 As shown, the following steps are included:

[0064] Step S1-1, the original image (taking the spectral image in the detection process of the third detection model detection unit as an example, such as Figure 6 (a)) is subjected to mean suppression processing, thereby obtaining a mean-valued image (such as Figure 6 (as shown in (b)).

[0065] Step S1-2, segment the averaged image to obtain a segmented image (such as Figure 6 (c)).

[0066] Among them, the segmented image is obtained through the OSTU algorithm.

[0067] Step S1-3, performing connected domain detection on the segmented image to obtain each connected domain, and calculating the minimum circumscribed circle of each connected domain.

[0068] Step S1-4, determine in turn whether the radius of the minimum circumscribed circle is less than a predetermined radius threshold, if it is determined to be no, enter step S1-5, if it is determined to be yes, re-enter step S1-4 to determine the next minimum circumscribed circle.

[0069] Step S1-5, the minimum circumscribed circle is the impurity circumscribed circle, the substance surrounded by the impurity circumscribed circle is the impurity, and the impurity position information is obtained according to the impurity circumscribed circle (such as Figure 6 As shown in (d), the Figure 6 The white circle in (d) is the impurity circumscribed circle), and then enters the end state.

[0070] Figure 7 The present invention is a flowchart of the working process of the intelligent tobacco impurity automatic detection and removal system according to the embodiment of the present invention.

[0071] like Figure 7 As shown, the working process of the intelligent tobacco impurity automatic detection and removal system includes the following steps:

[0072] Step S2-1, the control module 15 controls the screening conveyor 111 to move, so as to transport the raw tobacco to the bottom of the screening negative pressure device 112, and the screening negative pressure device 112 absorbs the lighter raw materials in the raw tobacco into the feeding funnel 121 as tobacco raw materials, and then enters step S2-2;

[0073] Step S2-2, the control module 15 controls the operating speed of the dropper 122, so that the tobacco raw materials in the loading hopper 121 fall onto the first conveyor 1231 according to a preset falling flow rate, and then the height limiting mechanism 124 further flattens the tobacco raw materials with an excessively high thickness flattened on the first conveyor 1231, and then enters step S2-3;

[0074] Step S2-3, the control module 15 controls the first image acquisition unit 1313 to acquire the RGB image of the tobacco raw material as the first image through the camera, and then proceeds to step S2-4;

[0075] Step S2-4, the control module 15 controls the first detection model detection unit 1314 to detect the first image through the first impurity detection model according to the color and shape features, so as to obtain a first detection result including the first impurity position information, and then enters step S2-5;

[0076] Step S2-5, the control module 15 controls the movement of the mechanical arm 142 located above the second conveyor 1232 according to the first impurity position information, so that the corresponding negative pressure suction head moves above the impurities, thereby sucking out the impurities and transporting them to the impurity recovery container 143 through the pipeline, and then enters step S2-6;

[0077] Step S2-6, the control module 15 controls the second image acquisition unit to acquire multiple RGB images of different angles of the tobacco raw material transported by the first conveyor 1231 as the second image through the camera, and then enters step S2-7;

[0078] Step S2-7, the control module 15 controls the second detection model detection unit to detect all second images through the second impurity detection model according to the texture feature, so as to obtain a second detection result including the second impurity position information, and then enters step S2-8;

[0079] Step S2-8, the control module 15 controls the movement of the mechanical arm 142 located above the third conveyor 1233 according to the second impurity position information, so that the corresponding negative pressure suction head moves above the impurity, thereby sucking out the impurity and transporting it to the impurity recovery container 143 through the pipeline, and then enters step S2-9;

[0080] Step S2-9, the control module 15 controls the third image acquisition unit to acquire the spectral image of the tobacco raw material transported from the fourth conveyor 1234 as the third image, and then proceeds to step S2-10;

[0081] Step S2-10, the control module 15 controls the third detection model detection unit to detect the third image through the pre-trained third impurity detection model according to the spectral feature, so as to obtain a third detection result including the third impurity position information, and then enters step S2-11;

[0082] Step S2-11, the control module 15 controls the movement of the mechanical arm 142 located above the third conveyor 1233 according to the third impurity position information, so that the negative pressure suction head moves above the impurities, sucks out the impurities and transports them to the impurity recovery container 143 through the pipeline, and then enters step S2-12;

[0083] In step S2-12, the control module 15 controls the fourth conveyor 1234 to convey the tobacco raw material as shredded tobacco to the fifth conveyor 1235, and then enters the end state.

[0084] Example Function and Effect

[0085] According to the intelligent automatic detection and removal system for tobacco impurities provided by the present embodiment, since the first detection unit detects the tobacco raw material according to the color characteristics and shape characteristics of the impurities to obtain a first detection result, the impurity removal module can remove impurities (such as foam, plastic, metal, stone, paper, moldy tobacco, tobacco stems, etc.) whose colors or contours are obviously different from those of tobacco according to the first detection result. Also, since the second detection unit detects the tobacco raw material according to the texture characteristics of the impurities to obtain a second detection result, the impurity removal module can remove impurities (such as hemp rope, cocoon, stem sticks, etc.) whose textures are obviously different from those of tobacco according to the second detection result. In addition, since the third detection unit detects the tobacco raw material according to the spectral characteristics of the impurities to obtain a third detection result, the impurity removal module can remove impurities (such as leaves, grass leaves, etc.) whose spectral characteristics are obviously different from those of tobacco according to the third detection result to obtain purer tobacco and improve tobacco quality.

[0086] In the embodiment, since the screening negative pressure device 112 uses the raw materials with a lighter specific gravity in the original tobacco as tobacco raw materials and conveys the tobacco raw materials to the feeding funnel 121, impurities with a larger specific gravity (such as stones, metals, stems, wet tobacco shreds, etc.) can be removed to complete the preliminary screening. Also, since the dropper 122 can control the falling flow rate of the tobacco raw materials, the height limiting mechanism 124 can limit the thickness of the tobacco raw materials spread on the belt conveyor, therefore, the tobacco raw materials can be completely spread on the belt conveyor, so that the impurity detection module 13 can detect all impurities in the tobacco raw materials.

[0087] In the embodiment, since the corresponding heights of the first conveyor 1231, the second conveyor 1232, the third conveyor 1233 and the fourth conveyor 1234 are successively reduced and the corresponding transmission speeds are successively accelerated, adjacent tobacco shreds can be separated to achieve tumbling of the upper and lower layers of the material, avoiding stacking and obstruction of the tobacco shreds and facilitating detection by the impurity detection module 13.

[0088] In the embodiment, since the second image acquisition unit acquires RGB images of the tobacco raw material from different angles through three cameras, it can acquire sufficient texture features of the tobacco raw material, thereby ensuring that the second detection model detection unit can detect impurities whose texture is obviously different from that of tobacco.

[0089] In the embodiment, since the first detection unit 131, the second detection unit 132 and the third detection unit 133 also have a light shield 1311 for blocking external natural light and a light source 1312 for providing visible light, the first image acquisition unit 1313, the second image acquisition unit and the third image acquisition unit can obtain images with higher quality.

[0090] In the embodiment, since the control module 15 moves the corresponding robot arm 142 according to the first impurity position information, the second impurity position information and the third impurity position information, the negative pressure suction head moves to the top of the corresponding impurity and adsorbs the impurity into the impurity recovery container 143. Therefore, it can be accurately positioned above the impurity to complete the removal and recovery of the impurities.

[0091] In the embodiment, since the negative pressure device 112 for screening is connected to the two tobacco flattening and transporting modules 12, a large amount of tobacco raw materials can be detected and removed by diversion, which greatly improves the efficiency of impurity removal.

[0092] In the above embodiment, the negative pressure device 112 for screening is connected to two tobacco flattening and transporting modules 12. In other schemes of the present invention, the negative pressure device 112 for screening can also be connected to one or more tobacco flattening and transporting modules 12 to meet the impurity identification and removal requirements of production lines with different material flow rates. For production lines with large material flow rates, multiple tobacco flattening and transporting modules 12 can be configured to improve the detection and removal efficiency of tobacco raw materials.

[0093] In the above embodiment, detection is performed by the first detection unit 131, the second detection unit 132 and the third detection unit 133. In other schemes of the present invention, only one or two detection units can be configured to automatically detect and remove tobacco impurities according to the need to detect the purity of the material.

Claims

1. An intelligent automatic detection and removal system for tobacco impurities, characterized in that: include: An impurity detection module, used to detect impurities in the tobacco raw material, thereby obtaining an impurity detection result, the impurity detection result including the impurity position information; as well as The impurity removal module removes the impurities from the tobacco raw material according to the impurity position information to obtain clean tobacco, and comprises a negative pressure impurity removal device, a mechanical arm and an impurity recovery container. The negative pressure impurity removal device comprises at least a negative pressure machine and a plurality of negative pressure suction heads. The impurity detection module comprises a first detection unit, a second detection unit and a third detection unit. The first detection unit detects the tobacco raw material according to the color characteristics and shape characteristics of the impurities to obtain a first detection result. The second detection unit detects the tobacco raw material according to the texture characteristics of the impurities to obtain a second detection result. The third detection unit detects the tobacco raw material according to the spectral characteristics of the impurities to obtain a third detection result. The impurity removal module removes the impurities according to the first detection result, the second detection result and the third detection result respectively. The first detection unit at least comprises a first image acquisition unit and a first detection model detection unit. The second detection unit at least comprises a second image acquisition unit and a second detection model detection unit. The third detection unit at least comprises a third image acquisition unit and a third detection model detection unit. The first image acquisition unit acquires an RGB image of the tobacco raw material as a first image through a camera arranged above the tobacco raw material. The first detection model detection unit detects the first image using a pre-trained first impurity detection model according to the color and shape features, thereby obtaining the first detection result, which includes first impurity position information. The second image acquisition unit acquires a plurality of RGB images of the tobacco raw material from different angles as the second image through a plurality of cameras arranged above the tobacco raw material. The second detection model detection unit detects all the second images using a pre-trained second impurity detection model according to the texture feature, thereby obtaining the second detection result, which includes second impurity position information. The third image acquisition unit acquires the spectral image of the tobacco raw material as the third image, The third detection model detection unit detects the third image through a pre-trained third impurity detection model according to the spectral feature, thereby obtaining the third detection result, which includes third impurity position information. The intelligent tobacco impurity automatic detection and removal system also includes: A tobacco preliminary screening module, used for preliminary screening of raw tobacco containing different impurities to obtain the tobacco raw material; and The tobacco flattening and transporting module is used to flatten the tobacco raw material and transport it to the impurity detection module. The tobacco preliminary screening module comprises a screening conveyor and a screening negative pressure device. The tobacco flattening and transporting module comprises a feeding hopper, a material dropper, a flattening and transporting mechanism and a height limiting mechanism. The dropper is arranged at the outlet of the feeding hopper, and the height limiting mechanism is arranged near the flattening transport mechanism. The screening conveyor is used to convey the raw tobacco to the bottom of the screening negative pressure device. The negative pressure device for screening absorbs the raw material with a smaller specific gravity in the original tobacco as the tobacco raw material, and transports the tobacco raw material to the feeding funnel. The dropper is used to control the falling flow rate of the tobacco raw material, so that the tobacco raw material is evenly spread on the flattening transport mechanism. The flattening transport mechanism has a plurality of transporters for transporting the tobacco raw material and passing through the first detection section, the second detection section and the third detection section. The height limiting mechanism is used to spread out the overlapping tobacco raw materials, thereby ensuring that all the tobacco raw materials are completely flattened.

2. The intelligent tobacco impurity automatic detection and removal system according to claim 1 is characterized by: in, The flattening transport mechanism has at least four transport planes, namely a first transport plane, a second transport plane, a third transport plane and a fourth transport plane. The first conveyor is located below the dropper, the second conveyor is located below the first detection unit, the third conveyor is located below the second detection unit, and the fourth conveyor is located below the third detection unit. The corresponding heights of the first conveyor, the second conveyor, the third conveyor and the fourth conveyor are successively reduced, and the corresponding conveying speeds are successively accelerated.

3. The intelligent automatic detection and removal system for tobacco impurities according to claim 2 is characterized in that: in, The first detection unit, the second detection unit and the third detection unit also have a light shield for shielding the tobacco raw material from external natural light and a plurality of light sources arranged in the light shield for providing visible light for the tobacco raw material.

4. The intelligent tobacco impurity automatic detection and removal system according to claim 2 is characterized in that: Also includes: Control module, The negative pressure machine provides power for all the negative pressure suction heads, the impurity recovery container is connected to the negative pressure suction heads through a pipeline, and the negative pressure suction heads are arranged at one end of the mechanical arm. When the first detection model detection unit obtains the first impurity position information, the control module controls the movement of the mechanical arm located above the second conveyor according to the first impurity position information, so that the negative pressure suction head moves to the impurity corresponding to the first impurity position information, sucks out the impurity and transports it to the impurity recovery container through the pipeline. When the second detection model detection unit obtains the second impurity position information, the control module controls the movement of the mechanical arm located above the third conveyor according to the second impurity position information, so that the negative pressure suction head moves to the impurity corresponding to the second impurity position information, sucks out the impurity and transports it to the impurity recovery container through the pipeline. When the third detection model detection unit obtains the third impurity position information, the control module controls the movement of the robotic arm located above the third conveyor according to the third impurity position information, so that the negative pressure suction head moves to above the impurity corresponding to the third impurity position information, sucks out the impurity and transports it to the impurity recovery container through the pipeline, thereby completing the impurity removal.

5. The intelligent tobacco impurity automatic detection and removal system according to claim 1, Features: Wherein, the image corresponding to the tobacco raw material is the original image, The detection process of the impurity detection module includes the following steps: Step S1-1, performing mean suppression processing on the original image, so as to obtain a mean-valued image in which the number of remaining non-zero pixels in the original image is less than a predetermined number threshold; Step S1-2, segmenting the averaged image to obtain a segmented image; Step S1-3, performing connected domain detection on the segmented image to obtain each connected domain, and calculating the minimum circumscribed circle of each connected domain; Step S1-4, determining in sequence whether the radius of the minimum circumscribed circle is less than a predetermined radius threshold; Step S1-5, when the judgment in step S1-4 is no, the minimum circumscribed circle is the impurity circumscribed circle, the substance surrounded by the impurity circumscribed circle is the impurity, and the impurity position information is obtained according to the impurity circumscribed circle.

6. The intelligent tobacco impurity automatic detection and removal system according to claim 5 is characterized in that: in, The segmented image is obtained by using the OSTU algorithm.

Images