Shaking table sorting and transferring device based on machine vision

Abstract

The invention relates to the technical field of shaking table sorting, and particularly discloses a shaking table sorting and transferring device based on machine vision, which comprises a device base, a shaking table body is mounted in the device base, and an identification assembly for performing visual identification on slag slurry is arranged above the device base. A collecting assembly used for collecting gold sand is arranged on one side of the device base, and a circulating assembly used for circulating screened unreceived middlings is arranged on one side of the collecting assembly. The recognition assembly comprises a mounting frame. A gold abrasive belt in an image is accurately recognized through an industrial camera, a servo motor drives a lead screw in a mounting shell to rotate, then a movable rod in threaded connection is driven to horizontally move, the horizontal position of a collecting box on a connecting base is adjusted, and an arranged guide rail and a sliding block can play a role in stable guiding, so that the collecting box is pushed to move to a designated position; accurate receiving of the gold sand is achieved, manual operation is not needed, and the labor intensity of workers is reduced.

Description

Technical Field

[0001] This invention relates to the field of shaking table sorting technology, and in particular to a shaking table sorting and transfer device based on machine vision. Background Technology

[0002] In the field of mineral resource processing, gravity separation technology has always held an important position due to its green and efficient characteristics. As the core equipment for gravity separation, the shaking table has undergone a century of technological iteration, and its individual equipment structure and separation process are now highly mature. Through differential motion of the bed surface, lateral water flow washing, and separation based on the density-particle size difference of mineral particles, the shaking table can effectively separate concentrate (gold sand belt), secondary concentrate, and tailings, possessing irreplaceable technological advantages, especially in the field of gold ore beneficiation. As a key separation device, the separation efficiency of the shaking table directly affects the resource utilization rate and economic benefits of the beneficiation plant.

[0003] In order to allow the sludge to flow smoothly from the feed trough to the bed surface, the existing equipment will add water to the feed trough. When the water volume is large on the side of the metal distribution zone, it will have a scouring effect on the gold sand, causing the metal to remix with the sludge, resulting in a decrease in mineral processing efficiency. An existing patent (publication number: CN209663471U) discloses a sorting shaking table, which includes a frame, a bed plate horizontally slidably connected to the frame, and a drive structure for differential reciprocating linear motion of the bed plate. The bed plate is inclined, and the inclination direction is perpendicular to its moving direction. A feeding trough is fixedly connected to the upper inclined end of the bed plate, and the feeding trough is parallel to the sliding direction of the bed plate. It uses an adjusting block to control the flow area of ​​the feeding port, thereby controlling the speed at which water flows out from the feeding port at the end of the feeding trough away from the guide channel, reducing the impact of the water flow on the separated metals, thus reducing the number of metal particles that remix with the sludge, and thereby improving mineral processing efficiency.

[0004] To address the aforementioned issues, while existing patents offer solutions that reduce the amount of metal particles remixed with sludge by controlling the flow area of ​​the feed inlet using adjusting blocks, thereby improving mineral processing efficiency, the operation requires the shaker operator to visually determine the position of the gold dust at the table and adjust the gold dust receiving tray and middlings receiving tray accordingly. Since the shaker table height is generally around 50cm, the operator must bend over to perform these tasks, requiring frequent handling of middlings, resulting in high labor intensity, occupational disease risks, and prolonged exposure to cold water, severely impacting the operator's health. Summary of the Invention

[0005] The purpose of this invention is to provide a machine vision-based shaking table sorting and transfer device. Through the precise identification of gold sand belts in an image using an industrial camera, a servo motor drives a lead screw inside the mounting housing to rotate, which in turn drives a threaded movable rod to move horizontally, adjusting the horizontal position of the collection box on the connecting seat. The provided guide rails and sliders provide stable guidance, thereby pushing the collection box to the designated position and achieving precise gold sand collection. This eliminates the need for manual operation, reducing the labor intensity of workers and solving the problems mentioned in the background art.

[0006] To achieve the above objectives, the present invention provides the following technical solution: a shaking table sorting and transfer device based on machine vision, comprising a device base, a shaking table body installed inside the device base, an identification component for visually identifying slag slurry provided above the device base, a collection component for collecting gold sand provided on one side of the device base, and a circulation component for circulating uncollected middlings on one side of the collection component. The identification component includes a mounting frame, which is welded and fixed above the device base. Two sets of supplementary lights are symmetrically fixed to the inner wall of the mounting frame by screws. A mounting base is fixed to the side wall of the mounting frame by bolts, and an industrial camera is fixed to the inner side of the mounting base by bolts.

[0007] Preferably, the collecting component includes a servo motor, which is fixed to one side of the device base by bolts, and one end of the servo motor is provided with a mounting shell that is fixedly connected to the device base.

[0008] Preferably, the power output shaft of the servo motor is fixed with a lead screw via a coupling, and the end of the lead screw is threadedly connected to a movable rod that is telescopically connected to the mounting housing.

[0009] Preferably, a connecting seat is welded and fixed to the end of the movable rod, and a collection box is detachably installed inside the connecting seat by screws.

[0010] Preferably, a connecting seat is welded and fixed to the end of the movable rod, and a collection box is detachably installed inside the connecting seat by screws.

[0011] Preferably, the circulation assembly includes a mixing tank connected to the end of the middlings chute, and a mixing motor is fixed to the upper surface of the mixing tank by bolts, with mixing blades fixed to the power output shaft of the mixing motor.

[0012] Preferably, the side wall of the mixing tank is connected to a slurry pipe, and the end of the slurry pipe is connected to a slurry pump.

[0013] Preferably, a flushing water pipe is connected to one side of the slurry pump, a conveying pipe is connected to the output end of the slurry pump, and a discharge component fixed on the device base is connected to the end of the conveying pipe.

[0014] Compared with the prior art, the beneficial effects of the present invention are: 1. By accurately identifying the gold sand belt in the image through an industrial camera, the servo motor drives the lead screw in the mounting housing to rotate, which in turn drives the threaded movable rod to move horizontally, adjusting the horizontal position of the collection box on the connecting seat. The set guide rail and slider can play a stable guiding role, thereby pushing the collection box to the designated position, realizing the accurate collection of gold sand without manual operation, reducing the labor intensity of workers. 2. A large amount of unsorted gold sand remains in the remaining uncollected middlings. It is collected in a mixing tank through a middlings chute. Clean water is then used to pre-flush the pipeline before the pump to prevent sand from settling in the pipeline and causing blockages. The slurry pump mixes the slurry with the flushing water and pumps it to prevent the pump from being blocked by a large amount of slurry when it starts up due to low frequency. The slurry is then transported to the discharge port through a conveying pipe, and then transported again to the shaking table for mineral particle sorting. Attached Figure Description

[0015] To more clearly illustrate the specific embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the specific embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0016] Figure 1 This is an overall structural view of the present invention; Figure 2 This is a three-dimensional structural diagram of the components used in this invention; Figure 3 This is a three-dimensional structural diagram of the present invention; Figure 4 This is a three-dimensional structural diagram of the circulation component of the present invention; Figure 5 This is a side view of the structure of the present invention.

[0017] Explanation of reference numerals in the attached figures: 1. Device base; 2. Shaking table body; 3. Mounting frame; 31. Supplementary light; 32. Fixed base; 33. Industrial camera; 4. Servo motor; 41. Mounting shell; 42. Lead screw; 43. Movable rod; 44. Connecting seat; 45. Collection box; 46. Guide rail; 47. Slider; 48. Mid-ore sluice; 5. Mixing tank; 51. Mixing motor; 52. Mixing blades; 53. Slurry pipe; 54. Slurry pump; 55. Flushing water pipe; 56. Discharge part; 57. Conveying pipe. Detailed Implementation

[0018] 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. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0019] This invention provides a technical solution: Please see Figures 1 to 5 A machine vision-based shaking table sorting and transfer device includes a device base 1, a shaking table body 2 installed inside the device base 1, an identification component for visual identification of slag slurry above the device base 1, a collection component for collecting gold sand on one side of the device base 1, and a circulation component for recycling uncollected middlings on one side of the collection component. The identification component includes a mounting frame 3, which is welded and fixed above the device base 1. Two sets of supplementary lights 31 are symmetrically fixed to the inner wall of the mounting frame 3 by screws. A fixing seat 32 is fixed to the side wall of the mounting frame 3 by bolts. An industrial camera 33 is fixed to the inner side of the fixing seat 32 by bolts. The collection component includes a servo motor 4, which is fixed to one side of the device base 1 by bolts. One end of the servo motor 4 is provided with a mounting shell 41 fixedly connected to the device base 1. The power output shaft of the servo motor 4 is fixed with a lead screw 42 by a coupling. The end of the lead screw 42 is threadedly connected to a movable rod 43 that is telescopically connected to the mounting shell 41. A connecting seat 44 is welded and fixed to the end of the movable rod 43. A collection box 45 is detachably installed inside the connecting seat 44 by screws. Two guide rails 46 that are welded to the device base 1 are connected to the lower end of the connecting seat 44. A slider 47 that is fixedly connected to the connecting seat 44 is slidably connected to the upper surface of the guide rails 46. A medium ore chute 48 that is fixedly connected to the device base 1 is provided on one side of the connecting seat 44.

[0020] By adopting the above technical solution, the mineral particles to be processed are poured onto the shaking table 2, where they are sorted. The supplementary lighting 31 on the mounting frame 3 provides illumination to the mineral particles on the shaking table 2. The industrial camera 33 mounted on the fixed base 32 can identify gold inclusions within the mineral particles. The industrial camera 33 can be connected to a computer image processing workstation via a network cable, transmitting real-time image data to the workstation. Image detection and processing are then performed within the workstation. This image detection and processing system is based on a deep learning algorithm and utilizes U2... The U2NET network model achieves accurate identification of gold sand bands in images. By constructing a multi-layered neural network, it simulates the workings of neurons in the human brain, automatically learning features and patterns from large amounts of data. (The image detection and processing system first needs to collect a large amount of image data containing gold sand bands and label these images to clarify the specific location and extent of the gold sand bands in the images, forming a training dataset. Then, this labeled image data is input into the U2NET network model for training. During training, the parameters in the network are continuously adjusted through the backpropagation algorithm to minimize the error between the model's prediction and the labeled true results. After training with a large amount of data, the U2NET network model can learn the feature patterns of gold sand bands in various scenarios, thus possessing the ability to accurately identify gold sand bands in new input images. When a new image is input into the system, the trained U2NET model will quickly process the image, and through the collaborative work of the encoder and decoder, output the identification result of the gold sand band in the image, providing reliable data support for subsequent positioning, control, etc.) In the process of gold sand band detection, to ensure the accuracy of the detection results, the system introduces traditional image processing algorithms in collaboration with the deep learning model, effectively avoiding errors. While deep learning models can extract complex features, they still pose a risk of misjudgment when faced with special circumstances such as changes in lighting and background interference, potentially resulting in multi-target detection outputs. In such cases, traditional image processing algorithms leverage their advantages by extracting basic features such as target contours and textures from the shaking table image through fundamental operations such as grayscale transformation, edge detection, and threshold segmentation. Based on the shaking table image features, the system can accurately calculate the size of the detected target. In the shaking table production scenario, the gold sand belt has a relatively fixed size range, and the system will set a threshold according to this standard to automatically filter out abnormal targets that exceed the reasonable size range.By leveraging the complementary advantages of deep learning algorithms and traditional image processing algorithms, this method utilizes deep learning to capture the complex features of gold sand belts while employing the robustness of traditional algorithms for target screening, ultimately achieving accurate detection of gold sand belts. After obtaining the detection results, a coordinate transformation is performed. Since the positions of the camera and the shaking table are relatively fixed, the actual width of the shaking table's tail and the number of pixels at the tail of the shaking table within the camera's view are measured (obtaining the transformation relationship between pixels and actual distance: A=B, R=C / B*A, where A: actual shaking table width; B: shaking table pixel width; C: detected pixel coordinates; R: actual distance; thus, the actual boundary coordinates between gold sand and middlings can be calculated). The servo motor 4 is controlled by a controller, which is connected to the servo driver via a network cable. The controller is connected to the computer image processing workstation of the recognition component via a network cable to realize the motion control of the servo drive. This enables real-time interaction of data such as target position, action commands, and parameter settings. After obtaining the actual coordinates of the boundary between gold sand and medium ore, the controller sends them to the controller via communication. The controller converts the actual coordinates into pulse signals that the servo drive can recognize through a conversion algorithm and sends the pulse signals to the servo motor 4. The servo motor 4 drives the lead screw 42 in the mounting housing 41 to rotate, which in turn drives the threaded movable rod 43 to move horizontally. This adjusts the horizontal position of the collection box 45 on the connecting seat 44. The guide rail 46 and slider 47 play a stabilizing guiding role, thereby pushing the collection box 45 to the designated position to achieve accurate collection of gold sand.

[0021] Specifically, such as Figure 1 , Figure 3 , Figure 4 and Figure 5 As shown, the circulation assembly includes a mixing tank 5, which is connected to the end of the ore chute 48. A mixing motor 51 is fixed to the upper surface of the mixing tank 5 by bolts. A mixing blade 52 is fixed to the power output shaft of the mixing motor 51. A slurry pipe 53 is connected to the side wall of the mixing tank 5. A slurry pump 54 is connected to the end of the slurry pipe 53. A flushing water pipe 55 is connected to one side of the slurry pump 54. A conveying pipe 57 is connected to the output end of the slurry pump 54. A discharge component 56 fixed on the device base 1 is connected to the end of the conveying pipe 57.

[0022] By adopting the above technical solution, a large amount of gold sand that has not been effectively sorted still exists in the remaining uncollected middlings. It is collected uniformly into the mixing tank 5 through the middlings chute 48. Clean water is transported through the flushing water pipe 55 to pre-flush the pipeline before the pump to prevent sand from settling in the pipeline and causing blockage of the pipe and pump. The mixing motor 51 drives the mixing blades 52 to rotate and stir the slurry in the mixing tank 5. The slurry is mixed with the flushing water and pumped by the slurry pump 54 through the slurry pipe 53 to prevent the pump from sucking in a large amount of slurry at a low frequency when it is first started, which would cause blockage of the pump. The slurry is transported to the discharge device 56 through the conveying pipe 57, and then transported to the shaking table body 2 again for mineral particle sorting.

[0023] Working Principle: Before using the machine vision-based shaking table sorting and transfer device, first place the device base 1 in a suitable position, then pour the mineral particles to be processed onto the shaking table body 2, where the shaking table body 2 sorts the mineral particles. The supplementary lighting 31 on the mounting frame 3 illuminates the mineral particles on the shaking table body 2, while the industrial camera 33 on the fixed base 32 identifies gold inclusions within the mineral particles. The industrial camera 33 can be connected to a computer image processing workstation via a network cable, transmitting real-time image data to the workstation for image detection and processing. The image detection and processing system is based on deep learning algorithms, utilizing the U2NET network model to accurately identify gold inclusions in images. By constructing a multi-layer neural network, it simulates the working method of neurons in the human brain, automatically learning features and patterns from large amounts of data. The servo motor 4 is controlled by a controller, which is connected to the servo driver via a network cable to achieve motion control of the servo driver. The controller is connected to the computer image processing workstation of the recognition component via a network cable, enabling real-time interaction of data such as target position, action commands, and parameter settings. After obtaining the actual coordinates of the boundary between gold sand and middlings, the data is sent to the controller via communication. The controller uses a conversion algorithm to convert the actual coordinates into pulse signals that the servo driver can recognize, and then sends the pulse signals to the servo motor 4. The servo motor 4 drives the lead screw 42 inside the mounting housing 41 to rotate, which in turn drives the threaded movable rod 43 to move horizontally, adjusting the horizontal position of the collection box 45 on the connecting seat 44. The guide rail 46 and slider 47 provide stable guidance, thereby pushing the collection box 45 to the designated position to achieve precise collection of gold sand. A large amount of unsorted gold sand remains in the remaining middlings, which is collected uniformly into the mixing tank 5 through the middlings chute 48. Clean water is delivered through the flushing water pipe 55 to pre-flush the pipeline before the pump to prevent sand from settling in the pipeline and causing blockages in the pipes and pump. The mixing motor 51 drives the mixing blades 52 to rotate, stirring the slurry in the mixing tank 5. The slurry pump 54 mixes the slurry with the flushing water and pumps it through the slurry pipe 53 to prevent the pump from clogging due to the low frequency of slurry intake at startup. The slurry is then transported to the discharge port 56 through the conveying pipe 57, and then transported again to the shaking table 2 for mineral particle separation.

[0024] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some or all of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention.

Claims

1. A machine vision-based shaking table sorting and transfer device, comprising a device base (1), characterized in that: The device base (1) is equipped with a shaking table body (2) inside. The device base (1) is provided with an identification component for visual identification of slag slurry above it. The device base (1) is provided with a collection component for collecting gold sand on one side. The collection component is provided with a circulation component for circulating the unreceived middlings. The identification component includes a mounting bracket (3), which is welded and fixed above the device base (1). Two sets of supplementary lights (31) are symmetrically fixed to the inner wall of the mounting bracket (3) by screws. A fixing seat (32) is fixed to the side wall of the mounting bracket (3) by bolts. An industrial camera (33) is fixed to the inner side of the fixing seat (32) by bolts.

2. The shaking table sorting and transfer device based on machine vision according to claim 1, characterized in that: The collection component includes a servo motor (4), which is fixed to one side of the device base (1) by bolts. One end of the servo motor (4) is provided with a mounting shell (41) that is fixedly connected to the device base (1).

3. The shaking table sorting and transfer device based on machine vision according to claim 2, characterized in that: The power output shaft of the servo motor (4) is fixed with a lead screw (42) via a coupling, and the end of the lead screw (42) is threadedly connected to a movable rod (43) that is telescopically connected to the mounting shell (41).

4. The shaking table sorting and transfer device based on machine vision according to claim 3, characterized in that: The end of the movable rod (43) is welded and fixed with a connecting seat (44), and a collection box (45) is detachably installed inside the connecting seat (44) by screws.

5. The shaking table sorting and transfer device based on machine vision according to claim 4, characterized in that: The lower end of the connecting seat (44) is connected to two guide rails (46) that are welded to the device base (1). The upper surface of the guide rails (46) is slidably connected to a slider (47) that is fixedly connected to the connecting seat (44). A medium ore chute (48) that is fixedly connected to the device base (1) is provided on one side of the connecting seat (44).

6. The shaking table sorting and transfer device based on machine vision according to claim 5, characterized in that: The circulation assembly includes a mixing tank (5), which is connected to the end of the ore chute (48). A mixing motor (51) is fixed to the upper surface of the mixing tank (5) by bolts, and a mixing blade (52) is fixed to the power output shaft of the mixing motor (51).

7. The shaking table sorting and transfer device based on machine vision according to claim 6, characterized in that: The side wall of the mixing tank (5) is connected to a slurry pipe (53), and the end of the slurry pipe (53) is connected to a slurry pump (54).

8. A shaking table sorting and transfer device based on machine vision according to claim 7, characterized in that: One side of the slurry pump (54) is connected to a flushing water pipe (55), the output end of the slurry pump (54) is connected to a conveying pipe (57), and the end of the conveying pipe (57) is connected to a discharge component (56) fixed on the device base (1).

Images