Garbage sorting method, equipment and device based on large model and readable storage medium
By combining large-scale model technology with sorting robots, efficient and precise automatic sorting of waste has been achieved, solving the problems of labor intensity, low efficiency and poor accuracy in the existing waste sorting process, and realizing unmanned processing.
Patent Information
- Application Number
- CN202511342000.4
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-09-19
- Publication Date
- 2025-11-14
AI Technical Summary
The existing waste sorting process suffers from problems such as labor intensity, low efficiency, and poor accuracy, and there is an urgent need for an efficient and accurate waste sorting solution.
By combining large-scale model technology with sorting robots, the system identifies waste types and coordinates using cameras and utilizes a dexterous hand gripping mechanism to achieve automatic waste sorting. This includes training large-scale models and using large-scale models that support object detection, such as YOLOv5, YOLOv8, YOLOv9, Detectron2, RT-DETR, and SAM.
It has improved the efficiency and accuracy of waste disposal, achieved unmanned processing, and greatly saved labor costs.
Smart Images
Figure CN120940266A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of intelligent waste sorting, and in particular to a waste sorting method, equipment, apparatus, and readable storage medium based on a large model. Background Technology
[0002] With the acceleration of urbanization, the amount of waste generated is increasing exponentially, leading to an increasingly heavy workload in waste management.
[0003] Currently, waste disposal relies on physical methods such as manual labor, magnetic separation / air separation / screening, which results in labor-intensive, inefficient, and inaccurate waste sorting processes.
[0004] Therefore, there is an urgent need for a waste sorting solution based on a large model to solve the above-mentioned technical problems. Summary of the Invention
[0005] This specification provides a waste sorting method, equipment, device, and readable storage medium based on a large model. In the waste sorting process, the large model technology is combined with a sorting robot to improve the efficiency and accuracy of waste treatment and realize unmanned processing, which greatly saves labor costs.
[0006] This specification provides a waste sorting method based on a large model, including:
[0007] Based on images of waste to be sorted on the conveyor system acquired by the camera, a trained large model is used for identification, and the type and coordinates of the waste to be sorted are output; wherein, the large model is a large model that supports target detection.
[0008] Based on the type of waste to be sorted and the coordinates of the waste to be sorted, multiple gripping mechanisms of the sorting robot are controlled to grip the waste to be sorted and place it in a storage device corresponding to the type of waste to be sorted.
[0009] Optionally, the grasping mechanism is a dexterous hand; wherein, the types of dexterous hands include H-type dexterous hands and Y-type dexterous hands.
[0010] Optionally, based on the type of waste to be sorted, a corresponding dexterous hand is determined to grasp the waste to be sorted.
[0011] Optionally, based on the identified type of waste to be sorted on the conveyor, if the type of dexterous hand corresponding to the identified type is different from the type of dexterous hand currently performing the grasping task, then switch to the type of dexterous hand corresponding to the identified type.
[0012] Optionally, the large model supporting object detection includes at least one of the following: YOLOv5, YOLOv8, YOLOv9, Detectron2, RT-DETR, and SAM.
[0013] Optionally, different types of garbage images are collected and labeled to form a training dataset, and the large model is trained until a preset accuracy is achieved.
[0014] Optionally, before acquiring images of the waste to be sorted on the conveying device based on the camera, the method further includes: conveying the waste to be sorted from the feeding structure to the tension screen; and, through the shaking of the tension screen, keeping the waste in a flat state as it moves toward the sorting robot.
[0015] This specification also provides a waste sorting device based on a large model, including: a large model recognition module and a sorting module;
[0016] in,
[0017] The large model recognition module is used to identify the waste to be sorted on the conveyor based on the image of the waste to be sorted on the camera, and output the type of waste to be sorted and the coordinates of the waste to be sorted; wherein, the large model is a large model that supports target detection;
[0018] The sorting module is used to control multiple gripping mechanisms of the sorting robot to grip the waste to be sorted based on the type of waste to be sorted and the coordinates of the waste to be sorted, and to place the waste to be sorted in a storage device corresponding to the type of waste to be sorted.
[0019] This specification also provides a waste sorting device based on a large model, including a processor and a memory, wherein the memory stores a program, and the processor executes the program to implement the waste sorting method based on a large model as described in any of the above-described technical solutions.
[0020] This specification also provides a readable storage medium storing a program that, when executed, implements the large-model-based waste sorting method described in any of the above-described technical solutions.
[0021] This specification provides a large-scale model-based waste sorting method, comprising: identifying images of waste to be sorted on a conveyor system acquired by a camera using a trained large-scale model, and outputting the type and coordinates of the waste to be sorted; wherein the large-scale model is a large-scale model supporting object detection; and controlling multiple gripping mechanisms of a sorting robot based on the type and coordinates of the waste to be sorted, grasping the waste, and placing it in a storage device corresponding to the type of waste. This solution combines large-scale model technology with a sorting robot in the waste sorting process, improving the efficiency and accuracy of waste processing, and achieving unmanned processing, greatly saving labor costs. Attached Figure Description
[0022] To more clearly illustrate the technical solutions in the embodiments of this application, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.
[0023] Figure 1 A schematic diagram of the waste sorting method based on a large model provided in the embodiments of this specification;
[0024] Figure 2 A schematic diagram of a waste sorting device based on a large model provided for an embodiment of this specification;
[0025] Figure 3 A schematic diagram of a waste sorting device based on a large model provided for embodiments of this specification;
[0026] Figure 4 This is a schematic diagram of the structure of an electronic device provided in an embodiment of this specification;
[0027] Figure 5 This is a schematic diagram of a computer-readable medium provided for embodiments of this specification.
[0028] Terminology Explanation:
[0029] 300. Electronic device; 310. Processing unit; 320. Storage unit; 330. Bus; 340. Display unit; 3201. Random access memory (RAM); 3202. Cache memory; 3203. Read-only memory (ROM); 3204. Program / utility; 3205. Program module; 350. Input / output (I / O) interface; 360. Network adapter; 400. External device. Detailed Implementation
[0030] The following description is intended to disclose the present invention and enable those skilled in the art to implement it. The preferred embodiments described below are merely examples, and other obvious variations will occur to those skilled in the art. The basic principles of the invention defined in the following description can be applied to other embodiments, modifications, improvements, equivalents, and other technical solutions that do not depart from the spirit and scope of the invention.
[0031] The following is in conjunction with the appendix Figures 1-5 Exemplary embodiments of the invention will be described more fully here. However, exemplary embodiments can be implemented in many forms and should not be construed as limiting the invention to the embodiments set forth herein. Rather, these exemplary embodiments are provided to make the invention more comprehensive and complete, and to facilitate a full communication of the inventive concept to those skilled in the art. The same reference numerals in the figures denote the same or similar elements, components, or parts, and therefore repeated descriptions of them are omitted.
[0032] Subject to the technical concept of this invention, the features, structures, characteristics or other details described in a particular embodiment may be combined in one or more other embodiments in a suitable manner.
[0033] In the description of specific embodiments, the features, structures, characteristics, or other details described in this invention are intended to enable those skilled in the art to fully understand the embodiments. However, it is not excluded that those skilled in the art can practice the technical solutions of this invention without one or more of the specific features, structures, characteristics, or other details.
[0034] The flowcharts shown in the accompanying drawings are merely illustrative and do not necessarily include all content and operations / steps, nor do they necessarily have to be performed in the described order. For example, some operations / steps can be broken down, while others can be combined or partially combined; therefore, the actual execution order may change depending on the specific circumstances.
[0035] The block diagrams shown in the accompanying drawings are merely functional entities and do not necessarily correspond to physically independent entities. That is, these functional entities can be implemented in software, in one or more hardware modules or integrated circuits, or in different network and / or processor devices and / or microcontroller devices.
[0036] The terms “and / or” or “and / or” include all combinations of any one or more of the listed items.
[0037] Figure 1 The schematic diagram of the waste sorting method based on a large model provided in the embodiments of this specification includes the following steps:
[0038] Step S101: Based on the image of the waste to be sorted on the conveyor obtained by the camera, the waste is identified by the trained large model, and the type and coordinates of the waste to be sorted are output.
[0039] Optionally, the types of waste to be sorted include, but are not limited to: metals, batteries, glass, porcelain, bricks, small appliances, and paper products.
[0040] Optionally, the large model is a large model that supports object detection.
[0041] Optionally, the large model supporting object detection includes at least one of the following: YOLOv5, YOLOv8, YOLOv9, Detectron2, RT-DETR, and SAM.
[0042] The following explanation uses YOLOv5 as an example:
[0043] The Yolov5 network mainly consists of the following parts: Input: Mosaic data augmentation, adaptive anchor box calculation, and adaptive image scaling; Backbone: Focus structure and CSP structure; Neck: FPN (top-down) + PAN (bottom-up) structure; Prediction: CIOU_Loss (the loss function used for bounding box regression, which replaces the traditional MSE or IoU Loss, improving target localization accuracy and convergence speed).
[0044] Input end:
[0045] (1) Mosaic data augmentation
[0046] The authors of Mosaic data augmentation, also members of the YOLOv5 team, stitched together four images by randomly scaling, cropping, and arranging them.
[0047] (2) Adaptive anchor frame calculation
[0048] In the YOLO algorithm, clustering is performed on different datasets to generate initial anchor boxes of varying widths and heights. During network training, the grid outputs predicted boxes based on these initial anchor boxes, which are then compared with the ground truth boxes. The difference between the two is calculated, and the network parameters are updated iteratively. In YOLOv3 and YOLOv4, the calculation of initial anchor box values for different datasets was done through a separate program. In YOLOv5, this functionality is embedded in the code, adaptively calculating the optimal anchor box values for different training sets during each training iteration. The control for this is in train.py; setting it to False prevents automatic calculation.
[0049] The adaptive anchor frame calculation process is as follows:
[0050] Load the dataset and get wh of all data in the dataset;
[0051] The maximum values of width (w) and height (h) in each image are scaled proportionally to the specified size (img_size), and the smaller sides are also scaled accordingly.
[0052] Change bboxes from relative coordinates to absolute coordinates (multiply by the scaled wh);
[0053] n anchors are obtained using k-means clustering;
[0054] The genetic algorithm is used to randomly mutate the wh values of anchors. If the effect is better after mutation (evaluated by the fitness calculated by the anchor_fitness method), the result of the mutation is assigned to the anchors. If the effect is worse after mutation, it is skipped. The default number of mutations is 1000.
[0055] (3) Adaptive image scaling
[0056] In commonly used object detection algorithms, different images have different lengths and widths. Therefore, a common approach is to uniformly scale the original images to a standard size before feeding them into the detection network.
[0057] Backbone network:
[0058] (1) Focus structure
[0059] In YOLOv5, the original 640×640×3 image is input into the Focus structure. The slicing operation first transforms it into a 320×320×12 feature map. After a convolution operation with 32 convolution kernels, it finally becomes a 320×320×32 feature map.
[0060] (2) CSP structure
[0061] The difference in CSP structures in YOLOv5 is that YOLOv4 only used CSP structures in the backbone network, while YOLOv5 designed two CSP structures. The CSP_X structure is used in the backbone network, while the CSP2_X structure is used in the Neck network.
[0062] (3) SPP structure
[0063] Using the SPP structure, feature extraction is performed through max pooling with different kernel sizes, thereby increasing the receptive field of the network. In YOLOv4, SPP was used within FPN, while in YOLOv5, the SPP (Spatial Pyramid Pooling) module is used after the backbone and before the neck (such as FPN+PAN). Its role is to perform multi-scale fusion of feature maps extracted by the backbone, enhancing the receptive field and thus improving detection accuracy, especially the ability to detect multi-scale targets.
[0064] Neck structure:
[0065] Like YOLOv4, YOLOv5 uses the FPN+PAN structure. However, while YOLOv4's Neck structure uses ordinary convolution operations, YOLOv5's Neck uses the CSP2 structure, which is inspired by CSPnet, to enhance the network's feature fusion capabilities.
[0066] The YOLOv5 loss function consists of:
[0067] The input image of 640×640 is divided into an N×N grid (usually 80×80, 40×40, or 20×20), and then three metrics are predicted for each cell: bounding box, confidence score, and classification probability.
[0068] in:
[0069] The rectangular frame represents the size and precise location of the target;
[0070] The confidence score characterizes the credibility of the predicted bounding box, and its value ranges from 0 to 1. The higher the value, the more likely the target is to exist in the bounding box.
[0071] Classification probability represents the category of a target. Therefore, in actual detection, the first step is to determine whether the prediction confidence of each predicted bounding box exceeds a set threshold. If it does, the predicted bounding box is considered to contain a target, thus obtaining the approximate location of the target. Next, the predicted bounding boxes containing targets are filtered using a non-maximum suppression algorithm, removing duplicate rectangles corresponding to the same target. Finally, based on the classification probabilities of the filtered predicted bounding boxes, the index corresponding to the highest probability is taken as the target's classification index number, thus obtaining the target's category.
[0072] The loss function measures the distance between the neural network's predicted information and the expected information (label). The closer the predicted information is to the expected information, the smaller the loss function value. Based on the predicted information for each cell, training mainly involves three types of losses: bounding box loss (loss_rect), confidence loss (loss_obj), and classification loss (loss_cls).
[0073] Optionally, different types of garbage images are collected and labeled to form a training dataset, and the large model is trained until a preset accuracy is achieved.
[0074] Optionally, before acquiring images of the waste to be sorted on the conveying device based on the camera, the method further includes: conveying the waste to be sorted from the feeding structure to the tension screen; and, through the shaking of the tension screen, keeping the waste in a flat state as it moves toward the sorting robot.
[0075] Step S102: Based on the type of waste to be sorted and the coordinates of the waste to be sorted, control multiple gripping mechanisms of the sorting robot to grip the waste to be sorted and place the waste to be sorted in a storage device corresponding to the type of waste to be sorted.
[0076] Optionally, the grasping mechanism is a dexterous hand; wherein, the types of dexterous hands include H-type dexterous hands and Y-type dexterous hands.
[0077] Optionally, based on the type of waste to be sorted, a corresponding dexterous hand is determined to grasp the waste to be sorted.
[0078] Optionally, based on the identified type of waste to be sorted on the conveyor, if the type of dexterous hand corresponding to the identified type is different from the type of dexterous hand currently performing the grasping task, then switch to the type of dexterous hand corresponding to the identified type.
[0079] This embodiment combines large-scale model technology with sorting robots in the waste sorting process, improving the efficiency and accuracy of waste treatment, and achieving unmanned processing, which greatly saves labor costs.
[0080] Figure 2 A schematic diagram of a waste sorting device based on a large model provided in the embodiments of this specification, such as... Figure 2 As shown, it includes a tension screen, a 2D camera + a 3D camera, a parallel robot, a dexterous hand, and a conveyor line.
[0081] Combination Figure 2 This section explains the working principle of waste sorting equipment.
[0082] 1. The waste to be sorted enters the tension screen from the feeding structure. The waste is spread out as it moves to the right by the shaking of the tension screen.
[0083] 2. The conveyor line runs from left to right;
[0084] 3. As the waste to be sorted moves along the conveyor line, it is photographed and collected by a camera. The data is then identified by a large model system, which provides the type of waste to be sorted and the coordinates of each piece of waste.
[0085] 4. The parallel robot determines the corresponding dexterous hand structure based on the coordinates and type of the waste to be sorted given by the large model, and realizes the grasping of the waste to be sorted and places it in the corresponding storage bin. Non-recyclable waste is transported to other waste storage bins in the back.
[0086] Among them, the parallel robot is equipped with a 3-claw dexterous hand at the end, which can be either an H-shaped or Y-shaped dexterous hand. It can switch modes according to the type of waste to be sorted, and can better adapt to grasping various types of waste to be sorted.
[0087] Based on the type of waste to be sorted, a corresponding dexterous hand is determined to grasp the waste.
[0088] Optionally, based on the identified type of waste to be sorted on the conveyor, if the type of dexterous hand corresponding to the identified type is different from the type of dexterous hand currently performing the grasping task, then switch to the type of dexterous hand corresponding to the identified type.
[0089] This embodiment combines large-scale model technology with sorting robots in the waste sorting process, improving the efficiency and accuracy of waste treatment, and achieving unmanned processing, which greatly saves labor costs.
[0090] Figure 3 A schematic diagram of a waste sorting device based on a large model provided in the embodiments of this specification includes: a large model recognition module 301 and a sorting module 302;
[0091] in,
[0092] The large model recognition module 301 is used to recognize images of waste to be sorted on the conveyor device acquired by the camera using a trained large model, and output the type and coordinates of the waste to be sorted; wherein, the large model is a large model that supports target detection.
[0093] The sorting module 392 is used to control multiple gripping mechanisms of the sorting robot to grip the waste to be sorted based on the type of waste to be sorted and the coordinates of the waste to be sorted, and to place the waste to be sorted in a storage device corresponding to the type of waste to be sorted.
[0094] This embodiment combines large-scale model technology with sorting robots in the waste sorting process, improving the efficiency and accuracy of waste treatment, and achieving unmanned processing, which greatly saves labor costs.
[0095] This embodiment also provides a waste sorting device based on a large model, including a processor and a memory, wherein the memory stores a program, and the processor executes the program to implement a waste sorting method based on a large model.
[0096] Figure 4 This is a schematic diagram of an electronic device provided as an embodiment of this specification. Refer to the following... Figure 4 The electronic device 300 according to this embodiment of the present invention will be described. Figure 4 The electronic device 300 shown is merely an example and should not impose any limitations on the functionality and scope of use of the embodiments of the present invention.
[0097] like Figure 4 As shown, the electronic device 300 is presented in the form of a general-purpose computing device. The components of the electronic device 300 may include, but are not limited to: at least one processing unit 310, at least one storage unit 320, a bus 330 connecting different system components (including storage unit 320 and processing unit 310), a display unit 340, etc.
[0098] The storage unit stores program code that can be executed by the processing unit 310, causing the processing unit 310 to perform the steps described in the above-described method section of this specification according to various exemplary embodiments of the present invention. For example, the processing unit 310 can perform actions such as... Figure 1 The steps are shown.
[0099] The storage unit 320 may include a readable medium in the form of a volatile storage unit, such as a random access memory unit (RAM) 3201 and / or a cache storage unit 3202, and may further include a read-only memory unit (ROM) 3203.
[0100] The storage unit 320 may also include a program / utility 3204 having a set (at least one) program module 3205, such program module 3205 including but not limited to: an operating system, one or more application programs, other program modules and program data, each or some combination of these examples may include an implementation of a network environment.
[0101] Bus 330 can represent one or more of several types of bus structures, including a memory cell bus or memory cell controller, a peripheral bus, a graphics acceleration port, a processing unit, or a local bus using any of the various bus structures.
[0102] Electronic device 300 can also communicate with one or more external devices 400 (e.g., keyboard, pointing device, Bluetooth device, etc.), and with one or more devices that enable viewers to interact with electronic device 300, and / or with any device that enables electronic device 300 to communicate with one or more other computing devices (e.g., router, modem, etc.). This communication can be performed via input / output (I / O) interface 350. Furthermore, electronic device 300 can also communicate with one or more networks (e.g., local area network (LAN), wide area network (WAN), and / or public networks, such as the Internet) via network adapter 360. Network adapter 360 can communicate with other modules of electronic device 300 via bus 330. It should be understood that, although... Figure 4 As not shown, other hardware and / or software modules may be used in conjunction with electronic device 300, including but not limited to: microcode, device drivers, redundant processing units, external disk drive arrays, RAID systems, tape drives, and data backup storage systems.
[0103] Through the above description of the embodiments, those skilled in the art will readily understand that the exemplary embodiments described in this invention can be implemented by software or by combining software with necessary hardware. Therefore, the technical solutions according to the embodiments of this invention can be embodied in the form of a software product, which can be stored in a computer-readable storage medium (such as a CD-ROM, USB flash drive, external hard drive, etc.) or on a network, including several instructions to cause a computing device (such as a personal computer, server, or network device, etc.) to execute the method described above according to this invention. When the computer program is executed by a data processing device, it enables the computer-readable medium to implement the method described above, i.e.: as... Figure 1 The method shown.
[0104] Figure 5 This is a schematic diagram of a computer-readable medium provided for embodiments of this specification.
[0105] accomplish Figure 1The computer program of the method shown can be stored on one or more computer-readable media. A computer-readable medium can be a readable signal medium or a readable storage medium. A readable storage medium can be, for example, but not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any combination thereof. More specific examples of readable storage media (a non-exhaustive list) include: an electrical connection having one or more wires, a portable disk, a hard disk, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fiber, portable compact disk read-only memory (CD-ROM), optical storage device, magnetic storage device, or any suitable combination thereof.
[0106] The computer-readable storage medium may include data signals propagated in baseband or as part of a carrier wave, carrying readable program code. Such propagated data signals may take various forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination thereof. The readable storage medium may also be any readable medium other than a readable storage medium, capable of transmitting, propagating, or transmitting programs for use by or in connection with an instruction execution system, apparatus, or device. The program code contained on the readable storage medium may be transmitted using any suitable medium, including but not limited to wireless, wired, optical fiber, RF, etc., or any suitable combination thereof.
[0107] Program code for performing the operations of this invention can be written in any combination of one or more programming languages, including object-oriented programming languages such as Java and C++, and conventional procedural programming languages such as C or similar languages. The program code can execute entirely on the audience's computing device, partially on the audience's device, as a standalone software package, partially on the audience's computing device and partially on a remote computing device, or entirely on a remote computing device or server. In cases involving remote computing devices, the remote computing device can be connected to the audience's computing device via any type of network, including a local area network (LAN) or a wide area network (WAN), or it can be connected to an external computing device (e.g., via the Internet using an Internet service provider).
[0108] In summary, the present invention can be implemented in hardware, or as software modules running on one or more processors, or a combination thereof. Those skilled in the art will understand that in practice, general-purpose data processing devices such as microprocessors or digital signal processors (DSPs) can be used to implement some or all of the functions of some or all of the components according to the embodiments of the present invention. The present invention can also be implemented as a device or apparatus program (e.g., a computer program and computer program product) for performing part or all of the methods described herein. Such programs implementing the present invention can be stored on a computer-readable medium or can be in the form of one or more signals. Such signals can be downloaded from an Internet website, provided on a carrier signal, or provided in any other form.
[0109] The specific embodiments described above further illustrate the purpose, technical solution, and beneficial effects of the present invention. It should be understood that the present invention is not inherently related to any specific computer, virtual device, or electronic device, and various general-purpose devices can also implement the present invention. The above descriptions are merely specific embodiments of the present invention and are not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.
[0110] The various embodiments in this specification are described in a progressive manner. The same or similar parts between the various embodiments can be referred to each other. Each embodiment focuses on describing the differences from other embodiments.
[0111] The above description is merely an embodiment of this application and is not intended to limit the scope of this application. Various modifications and variations can be made to this application by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the scope of the claims of this application.
Claims
1. A waste sorting method based on a large model, characterized in that, include: Based on images of waste to be sorted on the conveyor system acquired by the camera, a trained large model is used for identification, and the type and coordinates of the waste to be sorted are output; wherein, the large model is a large model that supports target detection. Based on the type of waste to be sorted and the coordinates of the waste to be sorted, multiple gripping mechanisms of the sorting robot are controlled to grip the waste to be sorted and place it in a storage device corresponding to the type of waste to be sorted.
2. The method according to claim 1, characterized in that, The grasping mechanism is a dexterous hand; wherein, the types of dexterous hands include H-type dexterous hands and Y-type dexterous hands.
3. The method according to claim 2, characterized in that, Based on the type of waste to be sorted, a corresponding dexterous hand is determined to grasp the waste.
4. The method according to claim 3, characterized in that, Based on the identified type of waste to be sorted on the conveyor, if the type of dexterous hand corresponding to the identified type is different from the type of dexterous hand currently performing the grasping task, the process switches to the type of dexterous hand corresponding to the identified type.
5. The method according to claim 1, characterized in that, The large models supporting object detection include at least one of the following: YOLOv5, YOLOv8, YOLOv9, Detectron2, RT-DETR, and SAM.
6. The method according to claim 1, characterized in that, Collect and label images of different types of garbage to form a training dataset, and train the large model until the preset accuracy is achieved.
7. The method according to claim 1, characterized in that, Before obtaining images of the waste to be sorted on the conveyor based on the camera, the process also includes: conveying the waste to be sorted from the feeding structure to the tension screen; and, through the shaking of the tension screen, keeping the waste in a flat state as it moves toward the sorting robot.
8. A waste sorting device based on a large model, characterized in that, include: Large model recognition module, sorting module; in, The large model recognition module is used to identify the waste to be sorted on the conveyor based on the image of the waste to be sorted on the camera, and output the type of waste to be sorted and the coordinates of the waste to be sorted; wherein, the large model is a large model that supports target detection; The sorting module is used to control multiple gripping mechanisms of the sorting robot to grip the waste to be sorted based on the type of waste to be sorted and the coordinates of the waste to be sorted, and to place the waste to be sorted in a storage device corresponding to the type of waste to be sorted.
9. A waste sorting device based on a large model, comprising a processor and a memory, wherein the memory stores a program, characterized in that, When the processor executes the program, it implements the method of any one of claims 1 to 7.
10. A readable storage medium storing a program, characterized in that, When the program is executed, it implements the method of any one of claims 1 to 7.