A multi-station parallel type grading device and grading method for cured tobacco leaves

Abstract

The application discloses a kind of post-curing tobacco multi-station parallel type grading device and grading method, including bearing substrate, camera, controller and at least one grading station are installed on bearing substrate, grading station includes tobacco loading plate, weighing sensor, transmission light source, reflection light source and man-machine interaction module, tobacco loading plate is used to carry single tobacco leaf, camera is fixed above tobacco loading plate, through the cooperation of grading station, camera and controller and the like structure, the multi-parameter fusion of single tobacco leaf is completed, and post-curing tobacco multi-station intelligent sorting guide is provided through man-machine interaction module;Post-curing tobacco grader can mark tobacco grade through man-machine interaction module on site, complete local training of grading model.

Description

Technical Field

[0001] This invention belongs to the field of grading technology for flue-cured tobacco leaves, specifically relating to a multi-station parallel grading device and grading method for flue-cured tobacco leaves. Background Technology

[0002] Currently, the grading of cured tobacco leaves is mainly done manually, which suffers from problems such as significant subjective influence and low efficiency. In recent years, automated sorting systems for cured tobacco leaves based on conveyor belts have emerged. However, these systems primarily rely on limited parameters such as weight and reflectance images, lacking the ability to fuse and detect multi-dimensional parameters that better reflect the internal quality of tobacco leaves, such as transmission images. Updating existing system models requires technicians to train the model on a dedicated server and then deploy it back to the equipment, resulting in a long cycle and high barriers to entry. Therefore, this paper proposes a grading device and method for cured tobacco leaves that features multi-station intelligent sorting guidance, multi-parameter fusion, and local training of the grading model. Summary of the Invention

[0003] (a) Technical problems to be solved

[0004] To address the problems mentioned in the background section, this invention provides a multi-station parallel grading device and method for cured tobacco leaves. Through the coordinated operation of grading stations, cameras, and controllers, it achieves multi-parameter fusion of individual tobacco leaves and provides intelligent sorting guidance for multi-station cured tobacco leaf processing via a human-computer interaction module. Cured tobacco leaf graders can use the human-computer interaction module to label tobacco leaf grades on-site and complete local training of the grading model. This invention offers advantages such as intelligent multi-station sorting guidance, multi-parameter fusion, and local training of the grading model.

[0005] (II) Technical Solution

[0006] To achieve the above objectives, the present invention provides the following technical solution:

[0007] A multi-station parallel grading device for cured tobacco leaves includes a support base plate. A camera, a controller, and at least one grading station are mounted on the support base plate. Each grading station includes a tobacco-carrying plate, a weighing sensor, a reflective light source, and a human-machine interface module. The tobacco-carrying plate is used to place single tobacco leaves. The tobacco-carrying plate is connected to the support base plate through the weighing sensor. The camera and the reflective light source are fixed above the tobacco-carrying plate. The camera, the weighing sensor, the reflective light source, and the human-machine interface module are all electrically connected to the controller.

[0008] In the above technical solution, preferably, the smoke carrier plate is made of a uniform light transmission material; the grading station further includes a transmission light source, which is installed on the support substrate below the smoke carrier plate, and the transmission light source is electrically connected to the controller.

[0009] The present invention also provides a grading method using the above-mentioned device. First, the single tobacco leaf is placed on the tobacco carrier plate. Then, the camera and the grading station cooperate to collect the weight and reflection images of the single tobacco leaf and send them to the controller. Next, the tobacco leaf grader performs on-site grading of the single tobacco leaf through the human-computer interaction module. Finally, the controller associates and stores the grading results with the corresponding weight and reflection images, and completes the construction or update of the grading model locally.

[0010] The present invention also provides a grading method using the above-mentioned device. First, the single tobacco leaf is placed on the tobacco carrier plate. Then, the camera and the grading station cooperate to collect the weight and reflection images of the single tobacco leaf and send them to the controller. Next, the controller determines the grade of the tobacco leaf according to the local grading model and provides intelligent sorting guidance for the cured tobacco leaf through the human-computer interaction module.

[0011] In the above technical solution, preferably, the bottom four corners of the supporting substrate are provided with adjustable feet, and the supporting substrate is also provided with a universal level. The adjustable feet and the universal level work together to achieve leveling of the supporting substrate.

[0012] In the above technical solution, preferably, a trigger switch is also included, which is electrically connected to the controller. When the trigger switch is triggered, the multi-parameter acquisition of the single tobacco leaf is initiated.

[0013] (III) Beneficial Effects

[0014] Compared with the prior art, the beneficial effects of the present invention are as follows:

[0015] 1. This invention integrates multi-parameter fusion acquisition and judgment of tobacco leaf weight, transmission image and reflection image at the same grading station, which overcomes the shortcomings of existing automatic sorting systems that only collect reflection images of cured tobacco leaves and have a single information dimension, and improves the grading accuracy.

[0016] 2. This invention integrates multiple sorting stations side by side on the same carrier substrate to achieve parallel operation of multiple stations. At the same time, it provides intelligent sorting guidance to operators through a human-machine interaction module, which solves the problems of low efficiency and large subjective influence of manual sorting.

[0017] 3. This invention allows graders of cured tobacco leaves to annotate the grade of the tobacco leaves on-site via a human-computer interaction module. The controller then associates and stores the annotation results with multiple parameters, enabling the grading model to be trained directly on-site without the need for offline retraining. This solves the problems of existing systems where model updates rely on offline retraining, have long cycles, and high barriers to entry, and enables continuous optimization of the model during on-site use. Attached Figure Description

[0018] Figure 1 This is a schematic diagram of the structure of the present invention.

[0019] In the diagram: 1. Supporting substrate; 11. Adjustable feet; 12. Universal level; 13. Support frame; 2. Grading station; 21. Smoke carrier plate; 22. Weighing sensor; 23. Transmitted light source; 24. Reflected light source; 25. Human-machine interaction module; 3. Camera; 4. Controller; 5. Trigger switch. Detailed Implementation

[0020] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without creative effort are within the scope of protection of the present invention.

[0021] like Figure 1 As shown, this invention provides a multi-station parallel grading device for cured tobacco leaves, including a support base plate 1. The support base plate 1 is a rectangular flat plate structure made of aluminum alloy with a thickness of not less than 5mm. Adjustable feet 11 are provided at the four bottom corners of the support base plate 1. The adjustable feet 11 are threaded lifting mechanisms, which can adjust the height of the support base plate 1 by rotation. A universal level 12 is also provided on the support base plate 1. The adjustable feet 11 and the universal level 12 work together to achieve leveling of the support base plate 1. A support frame 13 is provided below the support base plate 1 for supporting the support base plate 1.

[0022] A camera 3, a controller 4, and at least one grading station 2 are mounted on the carrier substrate 1. In this embodiment, there are two grading stations 2, which are arranged at equal intervals along the length of the carrier substrate 1, with gaps between adjacent grading stations 2. Each grading station 2 includes a smoke-carrying plate 21, a weighing sensor 22, a transmitted light source 23, a reflected light source 24, and a human-machine interaction module 25.

[0023] The tobacco carrier plate 21 is made of a light-transmitting material (such as frosted acrylic sheet or PC sheet) and is used to hold single tobacco leaves. The tobacco carrier plate 21 is connected to the support substrate 1 via a weighing sensor 22, but there is no direct contact between the tobacco carrier plate 21 and the support substrate 1. A transmission light source 23 is installed on the support substrate 1 below the tobacco carrier plate 21, and a camera 3 and a reflection light source 24 are fixed on the top of the tobacco carrier plate 21.

[0024] The weighing sensor 22 is a single-point weighing sensor, with its bottom fixed to the supporting base plate 1 and its top supporting the tobacco carrier plate 21. During weighing, the tobacco leaves are placed on the tobacco carrier plate 21, and gravity is transmitted to the weighing sensor 22 through the tobacco carrier plate 21. The elastic body inside the sensor undergoes a slight deformation, and the strain gauge attached to the surface of the elastic body converts the deformation into an electrical signal output. After processing by the analog-to-digital converter circuit, the signal is transmitted to the controller 4. The overall level of the supporting base plate 1 is achieved by using four adjustable feet 11 in conjunction with a universal level 12.

[0025] The transmitted light source 23 is an LED surface light source, using an industrial-grade backlit LED panel with a color temperature of 5000K, a color rendering index Ra>90, a length of 750mm, and a power of 10W. Its luminous area matches the size of the tobacco carrier plate 21, ensuring uniform light coverage of the entire tobacco leaf. The transmitted light source 23 is fixed to the supporting substrate 1, shining upwards. The light passes through the tobacco carrier plate 21 and the tobacco leaf before entering the camera 3.

[0026] As an alternative, the transmission light source 23 can use LED light sources of various wavelengths, and the controller 4 can automatically switch the light source mode according to different tobacco varieties or grades to obtain more comprehensive transmission image information.

[0027] The reflective light source 24 is a full-spectrum 4000K warm white LED tube, 800mm in length and 12W in power, used to provide illumination for the camera 3 when acquiring images. The reflective light source 24 is fixed above the grading station 2, its direction parallel to the long side of the tobacco carrier plate 21, illuminating the surface of the tobacco leaves downwards. The light is reflected from the tobacco leaf surface and then enters the camera 3. The controller 4 is configured to independently control the on / off state of the transmitted light source 23 and the reflective light source 24.

[0028] Camera 3 uses an OV5640 camera with a 120° wide-angle lens. The distance from the lens to the tobacco leaf surface is 300mm, with a horizontal coverage range of approximately 1039mm and a vertical coverage range of approximately 600mm, ensuring complete coverage of the tobacco leaf surface and keeping it entirely within the field of view. Camera 3 is fixed above grading station 2 with the lens facing downwards to capture images of the tobacco leaf surface.

[0029] The human-machine interaction module 25 uses a 7-inch capacitive touch screen (model: TAIDU TD-070C, resolution 1024×600, supports I2C interface), which is electrically connected to the controller 4. It includes a display screen and a touch button input unit, which is used to display the tobacco leaf grade judgment results and receive grade information input by the operator.

[0030] Controller 4 employs an ESP32-S3 microcontroller equipped with an Xtensa® 32-bit LX7 dual-core processor with a clock speed of 240MHz. It has 512KB of SRAM and 16MB of Flash storage, supports Wi-Fi and Bluetooth communication, and features a rich set of peripheral interfaces including PWM, I2C, SPI, and UART. Controller 4 is electrically connected to the weighing sensor 22 (via an analog-to-digital converter), the transmitted light source 23, the reflected light source 24, the camera 3, the human-machine interface module 25, and the trigger switch 5. Controller 4 has a built-in storage unit (such as an SD card or EEPROM) for storing local grading models and sample data.

[0031] It also includes a trigger switch 5, which is a push-button switch mounted on the side of the support base plate 1, located within the operator's reach, and electrically connected to the controller 4. When the trigger switch 5 is triggered, the controller 4 initiates the multi-parameter acquisition process for a single tobacco leaf.

[0032] This device is powered by an external DC power supply with an input voltage of DC 12V. The power is supplied to controller 4 via a power interface. Controller 4 has a built-in power conversion module that converts the 12V voltage to the operating voltage required by each component: 5V powers camera 3, weighing sensor 22, transmitted light source 23, reflected light source 24, and human-machine interface module 25; 3.3V powers controller 4. A fuse and filter circuit are provided at the power interface to protect the components from power fluctuations.

[0033] This invention has two modes, which together constitute the grading method of this invention: first, a local grading model is built through the training mode, and then the model is used to intelligently grade the cured tobacco leaves through the working mode.

[0034] 1. Training Mode

[0035] (1) Adjust the support plate 1 to a horizontal state by rotating the adjustable foot 11 and observing the universal level 12;

[0036] (2) The grader of the flue-cured tobacco leaves logs into the system with a username and password and switches to training mode;

[0037] (3) Place the four tobacco leaves on the tobacco carrier plates 21 of the four grading stations 2 respectively;

[0038] (4) After curing, the tobacco leaf graders use the human-computer interaction module 25 at each workstation to mark the grade of each of the four tobacco leaves;

[0039] (5) After the tobacco leaves are cured, the grader presses the trigger switch 5 to confirm the labeling results and starts the multi-parameter acquisition process.

[0040] (6) The controller 4 controls the weighing sensors 22 at each workstation to collect the weight data of each tobacco leaf and send it to the controller;

[0041] (7) The controller 4 controls the reflection light source 24 of each station to turn on, and the camera 3 collects the reflection image of each tobacco leaf and sends it to the controller. Then the reflection light source 24 is turned off.

[0042] (8) The controller 4 controls the transmission light source 23 of each station to turn on, and cooperates with the camera 3 to collect the transmission image of each tobacco leaf and send it to the controller, and then turns off the transmission light source 23.

[0043] (9) Controller 4 associates and stores the received data with the annotation results;

[0044] (10) If there is no hierarchical model at present, the controller 4 completes the construction of the hierarchical model; if there is a hierarchical model, the controller 4 completes the update of the hierarchical model.

[0045] 2. Work Mode

[0046] (1) Adjust the support plate 1 to a horizontal state by rotating the adjustable foot 11 and observing the universal level 12;

[0047] (2) Place the single tobacco leaf on the tobacco carrier plate 21 of the grading station 2;

[0048] (3) The operator presses the trigger switch 5;

[0049] (4) The controller 4 controls the weighing sensor 22 to collect tobacco weight data and send it to the controller;

[0050] (5) The controller 4 controls the reflection light source 24 to turn on, and the camera 3 collects the reflection image of the tobacco leaves and sends it to the controller. Then the reflection light source 24 is turned off.

[0051] (6) The controller 4 controls the transmission light source 23 to turn on, and the camera 3 collects the transmission image of the tobacco leaves and sends it to the controller. Then the transmission light source 23 is turned off.

[0052] (7) The controller 4 inputs the weight data, reflection image and transmission image into the local grading model to comprehensively judge the grade of tobacco leaves;

[0053] (8) The controller 4 displays the judgment result through the human-machine interaction module 25, providing guidance for intelligent sorting of the cured tobacco leaves.

[0054] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions, and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A multi-station parallel classification device for post-cured tobacco leaves, comprising a support base plate (1), characterized in that: A camera (3), a controller (4), and at least one grading station (2) are mounted on the carrier substrate (1). The grading station (2) includes a tobacco-carrying plate (21), a weighing sensor (22), a reflective light source (24), and a human-machine interaction module (25). The tobacco-carrying plate (21) is used to place single tobacco leaves. The tobacco-carrying plate (21) is connected to the carrier substrate (1) through the weighing sensor (22). The camera (3) and the reflective light source (24) are fixed above the tobacco-carrying plate (21). The camera (3), the weighing sensor (22), the reflective light source (24), and the human-machine interaction module (25) are all electrically connected to the controller (4).

2. A multi-station parallel post-cured tobacco leaf grading device according to claim 1, wherein: The smoke carrier plate (21) is made of a uniform light transmission material; the grading station (2) also includes a transmission light source (23), which is installed on the support substrate (1) below the smoke carrier plate, and the transmission light source (23) is electrically connected to the controller (4).

3. A grading method using the apparatus of claim 1, characterized in that: (1) Place the single tobacco leaf on the tobacco carrier plate (21); (2) The camera (3) and the grading station (2) work together to collect the weight and reflection image of a single tobacco leaf and send it to the controller (4); (3) The grader of the flue-cured tobacco leaves marks the grade of the single tobacco leaf on site through the human-computer interaction module (25); (4) The controller (4) associates and stores the annotation results with the corresponding weight and reflection image, and completes the construction or update of the hierarchical model locally.

4. A grading method using the apparatus of claim 1, characterized in that: (1) Place the single tobacco leaf on the tobacco carrier plate (21); (2) The camera (3) and the grading station (2) work together to collect the weight and reflection image of a single tobacco leaf and send it to the controller (4); (3) The controller (4) determines the grade of tobacco leaves according to the local grading model and provides intelligent sorting guidance for the cured tobacco leaves through the human-computer interaction module (25).

5. A multi-station parallel post-cured tobacco leaf grading device according to claim 1, wherein: The bottom four corners of the support substrate (1) are provided with adjustable feet (11), and the support substrate (1) is also provided with a universal level (12). The adjustable feet (11) and the universal level (12) work together to achieve leveling of the support substrate (1).

6. The multi-station parallel grading device for cured tobacco leaves according to claim 1, characterized in that: It also includes a trigger switch (5), which is electrically connected to the controller (4). When the trigger switch (5) is triggered, the multi-parameter acquisition of the single tobacco leaf is started.

Images