Visual recognition and sorting material system based on convolutional neural network and deep learning

Abstract

The present invention discloses a visual recognition and sorting material system based on convolutional neural network and deep learning, including: a material grading screen, a material divider, a first conveyor belt, a visual recognition device, a GPU server, a control device and an action execution device, the material grading screen is located at the head end of the first conveyor belt, the material divider and the visual recognition device are sequentially arranged on the first conveyor belt along the direction of material transmission, and the action execution device is arranged at the tail end of the first conveyor belt; the visual recognition device is connected to the control device through a GPU, the visual recognition device quickly collects feature information and spatial position information of the material on the first conveyor belt, and transmits it to the GPU server, the control device is connected to the action execution device, and controls the action of the action execution device. The present invention uses convolutional neural network and deep learning plus visual recognition to sort materials, with high overall efficiency and high intelligence.

Description

Technical Field

[0001] The present invention belongs to the technical field of material sorting, and more specifically, relates to a visual recognition and sorting material system based on convolutional neural networks and deep learning. Background Art

[0002] In the fields of agriculture, industry, and recycling of environmentally friendly materials, the sorting process is involved. In agriculture, with the improvement of people's living standards, consumers' attention to fresh agricultural products has shifted from quantity to quality, and then there has been quality inspection and quality grading of fresh agricultural products. Fresh agricultural products are sorted according to their size, weight, color, degree of damage, shape and texture, surface defects, etc., and then classified and sold, which improves consumer welfare and also encourages agricultural producers to optimize resource allocation. In industry, the inspection of the quality of industrial raw materials and industrial finished products is a sorting process. According to specific standards, defective products are sorted out and removed, leaving qualified products to enter the next production link or directly enter the market for circulation and use; in the recycling of environmentally friendly materials, such as the current garbage classification, domestic garbage can be sorted, recycled and processed, which can turn waste into treasure, better enhance the value of recyclables, and better alleviate our environmental pressure.

[0003] At present, there are mainly manual sorting, conventional mechanical sorting, and ray recognition sorting. Manual sorting is only suitable for sorting materials with smaller particle sizes, and because of the instability of human subjective feelings, the uniformity of manual sorting is relatively poor, the labor intensity is high, and the work efficiency is low. Now the labor cost is also expensive, so the cost of manual sorting is also increasing; conventional mechanical sorting is to sort by mechanical equipment according to a single physical grading index such as material quality and density. Although it has the advantages of high efficiency and low labor intensity, it cannot effectively identify color, shape texture, surface characteristics, etc. The sorting index is too single, the application range is narrow, the overall sorting effect is not good, the missorting rate is high, and due to the size limitation of the mechanical equipment itself, it is impossible to sort large particle size materials; the sorting equipment in the ray recognition mode has the defects of needing to apply for a ray product use certificate and file with the public security bureau because of the radioactivity of the ray, needing to set up a dedicated person to be responsible for the management and maintenance of the ray product, having penetration limitations, limited scope of use, high cost, and unable to identify materials with larger particle sizes.

[0004] Therefore, how to provide a visual recognition and sorting material system based on convolutional neural networks and deep learning is an urgent problem that technical personnel in this field need to solve. Summary of the invention

[0005] In view of this, the present invention provides a visual recognition and sorting material system based on convolutional neural networks and deep learning. Materials are sorted through convolutional neural networks and deep learning plus visual recognition. Multi-dimensional feature information of the collected sorted materials is extracted through convolutional neural networks and deep learning algorithms. The materials passing on the conveyor belt are then identified and judged through a visual recognition device. After obtaining accurate material sorting results and information such as the size, shape, and orientation of the materials, sorting operations are performed, with high overall efficiency and a high degree of intelligence.

[0006] In order to achieve the above object, the present invention adopts the following technical solution:

[0007] A visual recognition and sorting material system based on convolutional neural network and deep learning, comprising: a material grading screen, a divider, a first conveyor belt, a visual recognition device, a GPU server, a control device and an action execution device, wherein the material grading screen is located at the head end of the first conveyor belt, the divider and the visual recognition device are sequentially arranged on the first conveyor belt along the direction of material conveying, and the action execution device is arranged at the tail end of the first conveyor belt; the visual recognition device is connected to the control device through the GPU, the visual recognition device quickly collects feature information and spatial position information of the material on the first conveyor belt, and transmits the information to the GPU server, the GPU server contains a convolutional neural network and a deep learning algorithm, calculates and analyzes the feature information and spatial position information of the material and obtains the result, and transmits the result to the control device; the control device is connected to the action execution device, controls the action of the action execution device, and separates materials with different particle sizes.

[0008] Preferably, a vibration motor is installed on the material grading screen, and the material grading screen includes an upper large-aperture screen and a lower small-aperture screen, and the lower small-aperture screen is installed below the upper large-aperture screen. The upper large-aperture screen and the lower small-aperture screen can be used to preliminarily screen the material according to the particle size of the material.

[0009] Preferably, an iron remover is provided between the material distributor and the visual recognition device, and the iron substances contained in the material can be removed by the iron remover, thereby avoiding the influence on the material.

[0010] Preferably, the visual recognition device includes a shell, a high-definition industrial camera and a quadrilateral diffuse shadowless fill light source, the shell is installed above the first conveyor belt, and an opening is provided at the bottom of the shell, the high-definition industrial camera and the quadrilateral diffuse shadowless fill light source are both installed in the shell, thereby realizing the rapid collection of feature information and spatial position information of materials in a transport state randomly placed on the first conveyor belt.

[0011] Preferably, the action execution device includes an air compressor, an air tank and an air jet valve, wherein the air compressor is connected to the air tank through a pipeline, and the air jet valve is installed at the tail end of the first conveyor belt and is connected to the air tank through a pipeline. The air jet valve at the corresponding position is opened, and high-pressure gas of corresponding force is given according to the materials with different particle sizes, so that the specific materials with different particle sizes are blown out, and other materials that are not selected for firing are thrown down freely along the inertia of the conveyor belt.

[0012] Preferably, the control device includes a host computer and a programmable logic controller PLC, which are connected to the action execution device to control the air compressor to start supplying air to the air tank and control the opening and closing of the injection valve.

[0013] Preferably, a crushing device is provided between the material grading screen and the first conveyor belt, the material on the upper large-aperture screen enters the crushing device, and the material crushed by the crushing device enters the first conveyor belt. The crushing device can realize the crushing of the material, thereby improving the material sorting efficiency.

[0014] Preferably, the crushing device includes a crushing shell, a rotating motor, a rotating knife, a baffle and a spring mechanism. The rotating motor is installed in the middle position outside the crushing shell, and the output shaft of the rotating motor extends into the crushing shell and is connected to the rotating knife in a transmission manner; the baffle is installed on one side of the feeding port of the crushing shell, and the spring mechanism is installed on the crushing shell, and the output rod is against the baffle. The setting of the spring mechanism allows the baffle to move and adjust a certain distance along the direction of the output rod to avoid the problem of blockage when large pieces of material or materials that are difficult to crush enter; the setting of the baffle can limit large pieces of material to the area touched by the rotating knife, and then use the rotating knife to crush the large pieces of material, which is conducive to improving the crushing efficiency of the equipment.

[0015] Preferably, a screening machine is arranged below the tail end of the first conveyor belt, a second conveyor belt is arranged below the screening machine, a third conveyor belt is arranged at the head end of the screening machine, and the third conveyor belt conveys the material to the material grading screen. The crushed material can be screened by the screening machine to screen out small particles, and the large particles are transported back to the material grading screen via the third conveyor belt for re-screening.

[0016] Preferably, the screening machine comprises a screening housing, a screening motor and a screening mesh plate, wherein the screening mesh plate is installed at the top of the screening housing and is tilted with the head end lower and the tail end higher; the screening motor is fixed on one side of the screening housing. Under the vibration of the screening motor, small particles screened by the screening mesh plate fall into the second conveyor belt for transportation, and large particles fall onto the third conveyor belt along with the tilted screening mesh plate.

[0017] The beneficial effects of the present invention are:

[0018] The present invention sorts materials through convolutional neural networks, deep learning and visual recognition, extracts multi-dimensional feature information from the collected sorted materials through convolutional neural networks and deep learning algorithms, and then learns, models and stores the relevant feature information to master these information features. The material passing on the conveyor belt is identified and judged through a visual recognition device to obtain accurate material sorting results and information such as the size, shape, and orientation of the material. The control device receives the information and then sends an operating instruction to the action execution device to perform sorting operations. The overall efficiency is high, the degree of intelligence is high, and it has a good market development prospect. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on the provided drawings without paying creative work.

[0020] Figure 1 The accompanying drawing is a schematic structural diagram of an embodiment of the present invention.

[0021] Figure 2 The accompanying drawing is a block diagram of the overall architecture of the present invention.

[0022] Figure 3 The accompanying drawing is a structural block diagram of the control device of the present invention.

[0023] Figure 4 The accompanying drawing is a structural diagram of the grading screen of the present invention.

[0024] Figure 5 The accompanying drawing is a schematic structural diagram of another embodiment of the present invention.

[0025] Figure 6 The accompanying drawing is a schematic structural diagram of a crushing device of the present invention.

[0026] Figure 7 The accompanying drawing is a schematic structural diagram of yet another embodiment of the present invention.

[0027] Among them, in the figure,

[0028] 1 is a material grading screen; 1-1 is an upper large-aperture screen; 1-2 is a lower small-aperture screen; 1-3 is a vibration motor; 2 is a material distributor; 3 is an iron remover; 4 is a visual recognition device; 4-1 is a high-definition industrial camera; 4-2 is a quadrilateral diffuse shadowless fill light source; 5 is a GPU server; 6 is a control device; 7 is a recycling cover; 7-1 is an air injection valve, 7-2 is an air storage tank; 7-3 is an air compressor; 8 is the first conveyor belt; 8-1 is the fourth conveyor belt; 9 is a silo; 9-1 is a silo for defective materials; 9-2 is a silo for fine materials; 10 is a monitoring camera; 11 is a crushing device; 12 is a crushing shell; 14 is a rotating knife; 15 is a baffle; 16 is a spring mechanism; 17 is a screening machine; 18 is the second conveyor belt; 19 is the third conveyor belt; 20 is a screening shell; 21 is a screening motor; 22 is a screening mesh plate. DETAILED DESCRIPTION

[0029] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0030] Example 1

[0031] Please see attached Figure 1-4 The present invention provides a visual recognition and sorting material system based on convolutional neural network and deep learning, comprising: a material grading screen 1, a divider 2, a first conveyor belt 8, a visual recognition device 4, a GPU server 5, a control device 6 and an action execution device, wherein the material grading screen 1 is located at the head end of the first conveyor belt 8, and the divider 2 and the visual recognition device 4 are sequentially arranged on the first conveyor belt 8 along the direction of material conveying, and the action execution device is arranged at the tail end of the first conveyor belt 8; the visual recognition device 4 is connected to the control device 6 through a GPU, and the visual recognition device 4 quickly collects feature information and spatial position information of the material on the first conveyor belt 8, and transmits it to the GPU server 5. The GPU server 5 includes a convolutional neural network and a deep learning algorithm, which can calculate and analyze the feature information and spatial position information of the material under complex working conditions and obtain results, and transmit the obtained results to the control device 6; the control device 6 is connected to the action execution device, controls the action execution device to operate, and separates materials with different particle sizes.

[0032] The first conveyor belt 8 of the present invention runs at a certain speed, has a flat surface, runs stably and has a certain width, and plays the role of conveying materials; the divider 2 is a mechanical equipment for isolating and spreading materials, which is arranged at the starting section of the first conveyor belt 8, and receives materials of different sizes after screening by the material grading screen 1 in front, and classifies the materials of different sizes after screening and spreads them on the first conveyor belt 8 in their respective designated areas, and can spread the materials on the first conveyor belt 8 at a certain distance from each other, so that the materials are not stacked on each other, and are then transported forward by the first conveyor belt 8.

[0033] GPU server 5 includes convolutional neural networks and deep learning algorithms. It uses the convolution layer to extract image feature information from the collected material images as the input layer, analyzes the material feature information through RPN and Faster R-CNN algorithms, determines the shape, size, orientation, speed, etc. of the material, and realizes the identification and classification of the material. It learns, models, and stores a large amount of collected material feature information based on deep learning technology for subsequent classification and selection.

[0034] A vibration motor 1-3 is installed on the material grading screen 1, and the material grading screen 1 includes an upper large-aperture screen 1-1 and a lower small-aperture screen 1-2, and the lower small-aperture screen 1-2 is installed below the upper large-aperture screen 1-1. The upper large-aperture screen 1-1 and the lower small-aperture screen 1-2 can be used to preliminarily screen the material according to the particle size of the material. The material falls onto the material grading screen 1. Through the gravity of the material and the vibration of the vibration motor 1-3, the material with a larger aperture than the upper large-aperture screen 1-1 is effectively retained on the upper large-aperture screen 1-1, and then flows along the guide groove to the first conveyor belt 8; the material with a smaller aperture than the upper large-aperture screen 1-1 but larger than the aperture of the lower small-aperture screen 1-2 is retained on the lower small-aperture screen 1-2, and then flows along the guide groove to the other side of the first conveyor belt 8; the finely crushed material with a smaller aperture than the lower small-aperture screen 1-2 falls onto the fourth conveyor belt 8-1, and is processed in other ways. In addition, the material grading screen 1 can also be designed with multiple layers (more than two layers) of screens with different apertures according to different needs.

[0035] In order to further optimize the above technical solution, an iron remover 3 is arranged between the distributor 2 and the visual recognition device 4, and the iron substances contained in the material can be removed by the iron remover 3, thereby avoiding the influence on the material.

[0036] The visual recognition device 4 includes a shell, a high-definition industrial camera 4-1 and a quadrilateral diffuse shadowless fill light source 4-2. The shell is installed above the first conveyor belt 8, and an opening is provided at the bottom of the shell. The high-definition industrial camera 4-1 and the quadrilateral diffuse shadowless fill light source 4-2 are both installed in the shell, thereby realizing the rapid collection of feature information and spatial position information of materials in a transport state randomly placed on the first conveyor belt 8. The outer shell is a hollow cube that is closed at the top, front, back, left and right except for the bottom. Combined with the quadrilateral diffuse shadowless fill light source 4-2, it creates a stable light background for the high-definition industrial camera 4-1 to avoid being affected by extremely strong and weak light from the outside. The outer shell is fixed across the conveyor belt. The high-definition industrial camera 4-1 is fixed in the center of the upper part of the inner shell through a bracket or directly, and the lens is facing the target area of ​​the conveyor belt below. The quadrilateral diffuse shadowless fill light source 4-2 is fixed in a quadrilateral shape around the high-definition industrial camera 4-1 at a certain height inside the outer shell and faces the belt surface of the conveyor belt. The fixed position of the quadrilateral diffuse shadowless fill light source 4-2 is lower than the lens of the high-definition industrial camera 4-1, providing a fill light source for the high-definition industrial camera 4-1. The light is evenly and stably irradiated on the target area on the belt surface, which can effectively enhance the surface characteristics of the material passing through the target area.

[0037] The action execution device includes an air compressor 7-3, an air tank 7-2 and an air jet valve 7-1. The air compressor 7-3 is connected to the air tank 7-2 through a pipeline. The air jet valve 7-1 includes a large air jet valve 7-1 and a small air jet valve 7-1, which are arranged in a row at a certain angle at a specific position at the end of the first conveyor belt 8 and are connected to the air tank 7-2 through a pipeline. The air jet valve 7-1 at the corresponding position is opened, and high-pressure gas with corresponding force is given according to the materials with different particle sizes, so that the specific materials with different particle sizes are blown out, and other materials that are not selected for firing are thrown down freely along the inertia of the conveyor belt.

[0038] The control device 6 includes a host computer and a programmable logic controller PLC, which receives analysis data such as material feature information and spatial position information transmitted by the GPU server 5; the host computer and the programmable logic controller PLC are connected to the action execution device, and the control device 6 sends instruction information to the action execution device, such as whether to open the jet valve 7-1, whether to open the corresponding multiple large jet valves 7-1 and small jet valves 7-1 in the corresponding area to work together, and the duration of opening and closing the jet valve 7-1, to control the air compressor 7-3 to start supplying air to the air tank 7-2 and to control the simultaneous opening and closing of several jet valves 7-1 at the corresponding positions.

[0039] The present invention also includes a material receiving device, which includes a recycling cover 7, a monitoring camera 10, a fine material bin 9-2 and a defective material bin 9-1, wherein the recycling cover 7 is arranged at the tail end of the first conveyor belt 8, and the monitoring camera 10 is fixed in the recycling cover 7 to monitor the material receiving process. The setting of the recycling cover 7 avoids the diffusion of dust during the material receiving process and causes environmental pollution. The material blown away by the jet valve 7-1 falls into the defective material bin 9-1, and the material that is not selected for firing is freely thrown down along the inertia of the conveyor belt into the fine material bin 9-2.

[0040] Example 2

[0041] See attached Figure 5 and 6 On the basis of Example 1, a crushing device 11 is provided between the material grading screen 1 and the first conveyor belt 8, and the material on the upper large-aperture screen 1-1 enters the crushing device 11, and the material crushed by the crushing device 11 enters the first conveyor belt 8. The crushing device 11 can realize the crushing of the material, thereby improving the sorting efficiency of the material.

[0042] The crushing device 11 includes a crushing shell 12, a rotating motor, a rotating knife 14, a baffle 15 and a spring mechanism 16. The rotating motor is installed in the middle position outside the crushing shell 12. The output shaft of the rotating motor extends into the crushing shell 12 and is in transmission connection with the rotating knife 14. The baffle 15 is installed on one side of the feeding port of the crushing shell 12. The spring mechanism 16 is installed on the crushing shell 12, and the output rod is against the baffle 15. The setting of the spring mechanism 16 allows the baffle 15 to move and adjust a certain distance along the direction of the output rod to avoid the problem of blockage when large pieces of material or materials that are difficult to crush enter; the setting of the baffle 15 can limit the large pieces of material to the area touched by the rotating knife 14, and then the large pieces of material are crushed by the rotating knife 14, which is conducive to improving the crushing efficiency of the equipment.

[0043] Example 3

[0044] See attached Figure 7 On the basis of Example 2, a screening machine 17 is arranged below the tail end of the first conveyor belt 8, a second conveyor belt 18 is arranged below the screening machine 17, and a third conveyor belt 19 is arranged at the head end of the screening machine 17, and the third conveyor belt 19 conveys the material to the material grading screen 1. The crushed material can be screened by the screening machine 17 to screen out small particle materials, and the large particle materials can be transported back to the material grading screen 1 through the third conveyor belt 19 for re-screening.

[0045] The screening machine 17 includes a screening housing 20, a screening motor 21 and a screening mesh plate 22. The screening mesh plate 22 is installed at the top of the screening housing 20 and is tilted with the head end lower and the tail end higher. The screening motor 21 is fixed to one side of the screening housing 20. Under the vibration of the screening motor 21, small particles screened by the screening mesh plate 22 fall into the second conveyor belt 18 for transportation, and large particles fall onto the third conveyor belt 19 along with the tilted screening mesh plate 22.

[0046] The present invention uses convolutional neural network and deep learning plus visual recognition to sort materials, extracts multi-dimensional feature information from the collected sorted materials through convolutional neural network and deep learning algorithm, and then learns, models and stores relevant feature information to master these information features, and uses visual recognition device 4 to identify and judge the materials passing on the conveyor belt to obtain accurate material sorting results and information such as the size, shape, and orientation of the materials. The control device 6 receives the information and then issues an operation instruction to the action execution device to perform sorting operations. The overall efficiency is high, the degree of intelligence is high, and it has a good market development prospect. In addition, the present invention is a method that cannot be effectively identified by other methods such as rays, and visual recognition does not produce harmful substances. Materials with different particle sizes are classified in advance. The material recognition efficiency based on convolutional network and deep learning algorithm is high, the jet valve 7-1 responds quickly, large particle size materials and small particle size materials are separately identified and knocked out, the particle size range of the sorted materials is larger, the sorting effect is better, the system architecture is stable and reliable, and the overall hardware cost and operating cost are low.

[0047] In this specification, each embodiment is described in a progressive manner, and each embodiment focuses on the differences from other embodiments. The same or similar parts between the embodiments can be referred to each other. For the device disclosed in the embodiment, since it corresponds to the method disclosed in the embodiment, the description is relatively simple, and the relevant parts can be referred to the method part.

[0048] The above description of the disclosed embodiments enables one skilled in the art to implement or use the present invention. Various modifications to these embodiments will be apparent to one skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present invention. Therefore, the present invention will not be limited to the embodiments shown herein, but rather to the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A visual recognition and sorting material system based on convolutional neural network and deep learning, characterized in that: include: A material grading screen, a distributor, a first conveyor belt, a visual recognition device, a GPU server, a control device and an action execution device, wherein the material grading screen is installed with, and the material grading screen comprises an upper large-aperture screen and a lower small-aperture screen, and the lower small-aperture screen is installed below the upper large-aperture screen; a crushing device is arranged between the material grading screen and the first conveyor belt, the material on the upper large-aperture screen enters the crushing device, and the material crushed by the crushing device enters the first conveyor belt, the material grading screen is located at the head end of the first conveyor belt, and the distributors are arranged on the first conveyor belt in sequence along the direction of material transmission. The material container, the visual recognition device, and the action execution device are arranged at the tail end of the first conveyor belt; the visual recognition device is connected to the control device through the GPU, and the visual recognition device quickly collects the feature information and spatial position information of the material on the first conveyor belt, and transmits it to the GPU server. The GPU server includes a convolutional neural network and a deep learning algorithm, calculates and analyzes the feature information and spatial position information of the material and obtains the result, and transmits the result to the control device; the control device is connected to the action execution device, controls the action of the action execution device, and separates materials with different particle sizes; An iron remover is provided between the material distributor and the visual recognition device; The visual recognition device comprises a shell, a high-definition industrial camera and a quadrilateral diffuse shadowless fill light source, wherein the shell is installed above the first conveyor belt, and an opening is provided at the bottom of the shell, and the high-definition industrial camera and the quadrilateral diffuse shadowless fill light source are both installed in the shell; wherein the shell is a hollow cube style which is closed above and in front, back, left and right except for the bottom, and the shell is fixed across the conveyor belt; the high-definition industrial camera is fixed at the central position above the inside of the shell, and the lens is facing the target area of ​​the conveyor belt below; the quadrilateral diffuse shadowless fill light source is in the shape of a quadrilateral surrounding the high-definition industrial camera fixed at a certain height inside the shell and facing the belt surface of the conveyor belt; the fixed position of the quadrilateral diffuse shadowless fill light source is lower than the lens of the high-definition industrial camera, and provides a fill light source for the high-definition industrial camera; The action execution device includes an air compressor, an air tank and an injection valve. The air compressor is connected to the air tank through a pipeline. The injection valve is installed at the tail end of the first conveyor belt and is connected to the air tank through a pipeline.

2. According to claim 1, a visual recognition and sorting material system based on convolutional neural network and deep learning is characterized in that: The control device includes a host computer and a programmable logic controller PLC, which are connected to the action execution device to control the air compressor to start supplying air to the air storage tank and control the opening and closing of the injection valve.

3. According to claim 1, a visual recognition and sorting material system based on convolutional neural network and deep learning is characterized in that: The crushing device includes a crushing shell, a rotating motor, a rotating knife, a baffle and a spring mechanism. The rotating motor is installed at the middle position outside the crushing shell, and the output shaft of the rotating motor extends into the crushing shell and is transmission-connected with the rotating knife; the baffle is installed on one side of the feeding port of the crushing shell, the spring mechanism is installed on the crushing shell, and the output rod is against the baffle.

4. According to claim 1, a visual recognition and sorting material system based on convolutional neural network and deep learning is characterized in that: A screening machine is arranged below the tail end of the first conveyor belt, a second conveyor belt is arranged below the screening machine, a third conveyor belt is arranged at the head end of the screening machine, and the third conveyor belt conveys the material to the material grading screen.

5. According to claim 4, a visual recognition and sorting material system based on convolutional neural network and deep learning is characterized in that: The screening machine comprises a screening housing, a screening motor and a screening mesh plate. The screening mesh plate is installed on the top of the screening housing and is inclined with the head end lower and the tail end higher. The screening motor is fixed on one side of the screening housing.

Images