Neural network principle-based disordered smoke detection and early warning equipment

By integrating a neural network detection system on the tobacco leaf conveying equipment, using AI judgment and indicator light positioning, the problem of detecting mold smoke and insect smoke in tobacco leaf storage is solved, and the detection efficiency and safety are improved.

CN222014001UActive Publication Date: 2024-11-15NANJING XINYOU INTELLIGENT TECH CO LTD
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202420489458.7
Authority / Receiving Office
CN · China
Patent Type
Utility models(China)
Current Assignee / Owner
Filing Date
2024-03-14
Publication Date
2024-11-15
Estimated Expiration
2034-03-14

AI Technical Summary

Technical Problem

During the storage of tobacco leaves, it is difficult to effectively prevent mold smoke and insect smoke, resulting in some cigarette packs being missed during manual inspection, affecting the quality and health and safety of cigarettes.

Method used

A random smoke detection and early warning device based on the principle of neural network was designed. The camera was used to take photos of tobacco and tobacco leaves in transport, and AI was used to determine whether there was insect smoke and mold smoke, and the indicator lights and alarms were controlled through the controller. The instrument carries out early warning and positioning to ensure that staff can detect and deal with it in a timely manner.

Benefits of technology

It improves the efficiency and accuracy of detection of moldy smoke and insect smoke, reduces the missed detection rate of manual inspection, and ensures the quality and health and safety of the cigarette pack.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN222014001U_ABST
    Figure CN222014001U_ABST
Patent Text Reader

Abstract

The utility model discloses disordered cigarette detection and early warning equipment based on a neural network principle, which comprises a base, a conveyor is arranged on the base, a portal frame is arranged on the outer side of the base, a camera is arranged on the portal frame, and an indicating mechanism with the same transmission speed as the conveyor is arranged on at least one side of the base. A plurality of indicator lamp structures are arranged on the indicating mechanism, a central processing unit is arranged at the output end of the camera, a controller is arranged at the output end of the central processing unit, the output end of the controller can control the indicator lamp structures to be turned on according to needs, and tobacco shreds, tobacco leaves, cigarettes and the like are conveyed through the conveyor. In the transportation process, the camera is used for photographing the conveyor and then uploading the pictures to the central processing unit for AI reasoning, whether sundries such as insect cigarettes and mildew cigarettes exist on the pictures or not is judged, and if yes, the controller controls the alarm to give an alarm.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The utility model belongs to the technical field of cigarette production and processing, and specifically relates to a smoke disturbance detection and early warning device based on the neural network principle. Background Art

[0002] When tobacco leaves are shipped from the redrying factory, they are packed and compressed, then put into cartons and put into the warehouse for aging, which usually takes more than two years. During such a long storage period, it is difficult to prevent the tobacco packages from getting moldy and infested with insects.

[0003] Since moldy and worm-eaten cigarettes have a significant impact on the taste of cigarettes and are harmful to human health, after the cigarette packs enter the silk-making workshop assembly line of the cigarette factory, the factory generally arranges people to check whether the cigarette packs contain moldy and worm-eaten cigarettes. Generally, the people who check for worm-eaten and moldy cigarettes are not full-time, and they have many other tasks to do. They often lose sight of one thing while focusing on another, and cannot ensure that all cigarette packs are checked. Utility Model Content

[0004] The purpose of the utility model is to provide a smoke detection and early warning device based on the neural network principle to solve the above-mentioned problems.

[0005] To achieve the above-mentioned purpose, the utility model provides the following technical solutions: a smoke detection and warning device based on the principle of neural network, comprising a base, a conveyor is arranged on the base, a gantry is arranged on the outside of the base, a camera is arranged on the gantry, at least one side of the base is provided with an indication mechanism with the same transmission speed as the conveyor, a plurality of indicator light structures are arranged on the indication mechanism, a central processing unit is arranged at the output end of the camera, a controller is arranged at the output end of the central processing unit, and the output end of the controller can control the plurality of indicator light structures to light up according to demand.

[0006] Preferably, the indicating mechanism further comprises two pulleys movably connected to the base and one of the pulleys is externally connected to a power device capable of driving the pulleys to rotate, a synchronous belt is sleeved on the outer side of the two pulleys, and the plurality of indicator light structures are evenly fixed on the synchronous belt.

[0007] Preferably, the indicator light structure includes a shell fixedly connected to the synchronous belt, an indicator light body is arranged on the shell, a wireless transmission module for transmitting signals is also arranged inside the shell, a plurality of limit rods in a circular array are also arranged inside the shell, a moving block which is only limited to sliding inside the plurality of limit rods is arranged inside the plurality of limit rods, and a touch switch which can control the power supply of the indicator light body is fixedly arranged on the inner wall of the shell at the bottom of the moving block.

[0008] Preferably, the output end of the controller is also provided with an alarm, and the alarm is specifically arranged on the gantry and / or the base.

[0009] Preferably, an industrial display is also provided at the connection end of the central processing unit, and a storage battery capable of supplying power to the indicator light body is also provided inside the shell.

[0010] The technical effects and advantages of the utility model are as follows: shredded tobacco, tobacco leaves and cigarettes are transported by a conveyor. During the transportation, a camera is used to take pictures of the conveyor and then the pictures are uploaded to the central processing unit for AI reasoning to determine whether there are debris such as insect smoke and moldy smoke in the pictures. If so, the alarm is controlled by the controller to sound an alarm. At the same time, the part of the picture where the debris is located is calculated, and the indicator light structure at the corresponding position is lit in a targeted manner, which can determine the specific position, remind the staff to remove or remind them, and can have a good positioning effect on the search, removal and inspection of debris after removal. BRIEF DESCRIPTION OF THE DRAWINGS

[0011] Figure 1 It is a schematic diagram of the structure of the utility model;

[0012] Figure 2 For the utility model Figure 1 A partial enlarged view of the middle part;

[0013] Figure 3 This is a schematic diagram of the internal structure of the indicator light of the utility model;

[0014] Figure 4 This is a control system diagram of the utility model.

[0015] In the figure: 1. gantry; 2. camera; 3. conveyor; 4. base; 5. indicating mechanism; 501. pulley; 502. synchronous belt; 503. indicator light structure; 5031. shell; 5032. wireless transmission module; 5033. indicator light body; 5034. limit rod; 5035. moving block; 5036. touch switch; 5037. battery; 6. central processing unit; 7. industrial display; 8. alarm; 9. controller. DETAILED DESCRIPTION

[0016] The following will be combined with the drawings in the embodiments of the utility model to clearly and completely describe the technical solutions in the embodiments of the utility model. Obviously, the described embodiments are only part of the embodiments of the utility model, not all of the embodiments. Based on the embodiments in the utility model, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the utility model.

[0017] The utility model provides a smoke detection and early warning device based on the neural network principle as shown in the figure, including a base 4, a conveyor 3 is arranged on the base 4, a gantry 1 is arranged outside the base 4, a camera 2 is arranged on the gantry 1, at least one side of the base 4 is provided with an indicating mechanism 5 with the same transmission speed as the conveyor 3, a plurality of indicator light structures 503 are arranged on the indicating mechanism 5, a central processing unit 6 is arranged at the output end of the camera 2, a controller 9 is arranged at the output end of the central processing unit 6, and the output end of the controller 9 can control the plurality of indicator light structures 503 to light up according to demand.

[0018] Specifically, the indicating mechanism 5 also includes two pulleys 501 movably connected to the base 4, and one of the pulleys 501 is externally connected to a power device that can drive the pulley 501 to rotate. A synchronous belt 502 is sleeved on the outside of the two pulleys 501, and multiple indicator light structures 503 are evenly fixed on the synchronous belt 502.

[0019] Specifically, the indicator light structure 503 includes a shell 5031 fixedly connected to the synchronous belt 502, an indicator light body 5033 is arranged on the shell 5031, a wireless transmission module 5032 for transmitting signals is also arranged inside the shell 5031, a plurality of limit rods 5034 in a circular array are also arranged inside the shell 5031, a moving block 5035 which is only limited to sliding inside the plurality of limit rods 5034 is arranged inside the plurality of limit rods 5034, and a touch switch 5036 which can control the power supply of the indicator light body 5033 is fixedly arranged on the inner wall of the shell 5031 at the bottom of the moving block 5035.

[0020] Specifically, the output end of the controller 9 is further provided with an alarm 8 , and the alarm 8 is specifically arranged on the gantry 1 and / or the base 4 .

[0021] Specifically, the connection end of the central processor 6 is also provided with an industrial display 7 , and the housing 5031 is also provided with a storage battery 5037 for supplying power to the indicator light body 5033 .

[0022] Working principle: When using the utility model, tobacco shreds, tobacco leaves and cigarettes are transported by the conveyor 3. During the transportation, the camera 2 is used to take pictures on the conveyor 3 and then uploaded to the central processor for AI reasoning to determine whether there are debris such as insect tobacco and moldy tobacco in the picture. If there are, the alarm 8 is controlled by the controller to alarm. At the same time, the position of the debris in the photo is calculated, and the indicator light structure 503 at the corresponding position is lit in a targeted manner, which can determine the specific position, remind the staff to remove or remind, and can have a good positioning effect on the search, removal and inspection after removal of debris;

[0023] Then, the setting of the moving block 5035 can physically cut off the power at the indicator light structure 503 at the bottom of the synchronous belt 502, which is more intelligent and convenient, and can perfectly solve the problem of position error and waste of electricity caused by the moving blocks 5035 at the upper and lower positions being lit at the same time.

[0024] AI reasoning determines whether there are debris such as insect smoke and moldy tobacco in the picture. Since insect smoke and moldy tobacco are different from normal tobacco leaves in color and texture, we plan to use FasterCNN as the AI ​​model for target recognition, and use resnet50 as the backbone network. The resnet50 backbone network has a deeper network layer, so it can have a rich variety of feature extraction operators (different scales and different convolution kernels). It can automatically extract the color and texture features of insect smoke and moldy tobacco through a large amount of training, which is an ideal algorithm for this project.

[0025] Finally, it should be noted that the above is only a preferred embodiment of the present utility model and is not intended to limit the present utility model. Although the present utility model has been described in detail with reference to the aforementioned embodiments, it is still possible for those skilled in the art to modify the technical solutions described in the aforementioned embodiments, or to make equivalent substitutions for some of the technical features therein. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present utility model shall be included in the protection scope of the present utility model.

Claims

1. A smoke detection and early warning device based on the principle of a neural network, comprising a base (4), a conveyor (3) being arranged on the base (4), a gantry (1) being arranged outside the base (4), a camera (2) being arranged on the gantry (1), characterized in that: At least one side of the base (4) is provided with an indication mechanism (5) having the same transmission speed as the conveyor (3), and a plurality of indicator light structures (503) are provided on the indication mechanism (5). The output end of the camera (2) is provided with a central processing unit (6), and the output end of the central processing unit (6) is provided with a controller (9), and the output end of the controller (9) can control the lighting of the plurality of indicator light structures (503) according to demand.

2. According to claim 1, a smoke detection and warning device based on the neural network principle is characterized by: The indicating mechanism (5) further comprises two pulleys (501) movably connected to the base (4), and one of the pulleys (501) is externally connected to a power device capable of driving the pulley (501) to rotate. A synchronous belt (502) is sleeved on the outer side of the two pulleys (501), and a plurality of indicator light structures (503) are evenly fixed on the synchronous belt (502).

3. According to claim 1, a smoke detection and early warning device based on the principle of neural network is characterized by: The indicator light structure (503) comprises a shell (5031) fixedly connected to the synchronous belt (502), an indicator light body (5033) being arranged on the shell (5031), a wireless transmission module (5032) for transmitting signals being also arranged inside the shell (5031), a plurality of limiting rods (5034) arranged in a circular array being also arranged inside the shell (5031), a moving block (5035) being only limited to sliding inside the plurality of limiting rods (5034) being arranged inside the plurality of limiting rods (5034), and a touch switch (5036) capable of controlling the power supply of the indicator light body (5033) being fixedly arranged on the inner wall of the shell (5031) at the bottom of the moving block (5035).

4. According to claim 1, a smoke detection and warning device based on the principle of neural network is characterized by: The output end of the controller (9) is also provided with an alarm (8), and the alarm (8) is specifically arranged on the gantry (1) and / or the base (4).

5. According to claim 3, a smoke detection and warning device based on the principle of neural network is characterized by: The connection end of the central processing unit (6) is also provided with an industrial display (7), and the housing (5031) is also provided with a storage battery (5037) capable of supplying power to the indicator light body (5033).