Solid waste classification decision method and device, electronic equipment and storage medium

Abstract

The invention discloses a solid waste classification decision-making method and device, electronic equipment and a storage medium, according to the solid waste classification decision-making method and device, due to the fact that visual information and spectral information of materials are fused for multi-modal feature fusion analysis, a multi-dimensional perception collaborative recognition system is constructed, meanwhile, when recognition is abnormal, sorting operation is stopped in time, mistaken sorting is avoided, and therefore the sorting efficiency is improved. The technical problems that in an existing solid waste sorting method, due to the fact that a single-mode feature extraction technology is adopted, and a multi-dimensional perception and dynamic algorithm optimization cooperation mechanism is not established, recognition blind areas are caused, and the model generalization ability is insufficient can be solved, and the purposes of reducing the recognition blind areas, improving the model generalization ability and improving the sorting efficiency are achieved. And therefore, the technical effects of improving the accuracy, the reliability and the intelligent level of solid waste sorting are achieved.

Description

Technical Field

[0001] This disclosure relates to the field of data processing technology, and in particular to a solid waste classification decision-making method and apparatus, electronic equipment and storage medium. Background Technology

[0002] Solid waste treatment, as a crucial link in the resource recycling system, is widely applied in the classification and disposal of multi-source solid waste such as municipal solid waste, industrial waste, and construction waste. Among related technologies, traditional sorting systems mostly rely on mechanical screening, manual visual inspection, or single sensor detection, constructing a preliminary sorting process through the collaborative operation of physical sorting devices and basic image recognition.

[0003] Existing solid waste sorting methods directly employ single-modal feature extraction technology without establishing a collaborative mechanism for multi-dimensional perception and dynamic algorithm optimization. This may lead to problems such as blind spots in identification and insufficient model generalization ability. Summary of the Invention

[0004] This disclosure provides a solid waste classification decision-making method, apparatus, electronic device, and storage medium.

[0005] According to a first aspect of this disclosure, a solid waste classification decision-making method is provided, comprising: Collect visual and spectral information of materials located along the conveying path; Based on the visual information and the spectral information, a sorting instruction corresponding to the material category is generated through multimodal feature fusion analysis; In response to the sorting instruction, the control actuator moves the material to the collection area corresponding to the material category; If an abnormality occurs during the identification process, the sorting operation shall be stopped.

[0006] Optionally, the acquisition of visual and spectral information of materials located on the conveying path includes: Simultaneously acquire image data and spectral data of the material.

[0007] Optionally, the step of generating sorting indicators corresponding to material categories based on the visual information and the spectral information through multimodal feature fusion analysis includes: Morphological features are extracted from the visual information, and material features are extracted from the spectral information. The morphological features and material features are fused together, and the material category is determined based on the fused features.

[0008] Optionally, the step of controlling the actuator to transfer materials to the collection area corresponding to the material category in response to the sorting instruction includes: Based on the physical properties of the material, the adsorption mode, clamping mode, or scooping mode of the end effector of the actuator is adaptively selected.

[0009] Optionally, the method further includes: Record and store data related to the sorting process, and use the data to support adjustments to sorting strategies and updates to identification models.

[0010] Optionally, stopping the sorting operation when an abnormality occurs in the identification process includes: After an abnormal alarm is triggered, the drive system of the actuator is locked. Provides a human-machine interface to display identification information of abnormal materials and receive sorting instructions input by humans.

[0011] According to a second aspect of this disclosure, a solid waste classification decision-making device is provided, comprising: The acquisition unit is also used to acquire visual and spectral information of materials located on the conveying path; The analysis unit is also used to generate sorting instructions corresponding to the material category by performing multimodal feature fusion analysis based on the visual information and the spectral information. The control unit is also configured to, in response to the sorting instruction, control the actuator to transfer the material to the collection area corresponding to the material category; The exception handling unit is also used to stop the sorting operation when an exception occurs in the identification process.

[0012] Optionally, the acquisition unit is further configured to: Simultaneously acquire image data and spectral data of the material.

[0013] Optionally, the analysis unit is further configured to: Morphological features are extracted from the visual information, and material features are extracted from the spectral information. The morphological features and material features are fused together, and the material category is determined based on the fused features.

[0014] Optionally, the control unit is further configured to: Based on the physical properties of the material, the adsorption mode, clamping mode, or scooping mode of the end effector of the actuator is adaptively selected.

[0015] Optional, also includes: The recording unit is also used to record and store data related to the sorting process, and to support the adjustment of sorting strategies and the updating of identification models based on the data.

[0016] Optionally, the exception handling unit is further configured to: After an abnormal alarm is triggered, the drive system of the actuator is locked. Provides a human-machine interface to display identification information of abnormal materials and receive sorting instructions input by humans.

[0017] According to a third aspect of this disclosure, an electronic device is provided, comprising: At least one processor; and A memory communicatively connected to the at least one processor; wherein, The memory stores instructions that can be executed by the at least one processor to enable the at least one processor to perform the method described in the first aspect above.

[0018] According to a fourth aspect of this disclosure, a non-transitory computer-readable storage medium is provided storing computer instructions, wherein the computer instructions are configured to cause the computer to perform the method described in the first aspect above.

[0019] According to a fifth aspect of this disclosure, a computer program product is provided, comprising a computer program that, when executed by a processor, implements the method described in the first aspect above.

[0020] The solid waste classification decision-making method, device, electronic equipment, and storage medium disclosed herein, through the fusion analysis of visual and spectral information of materials to construct a multi-dimensional perception and collaborative identification system, and by promptly stopping the sorting operation when anomalies are identified to avoid missorting, can solve the technical problems of identification blind spots and insufficient model generalization ability caused by the use of single-modal feature extraction technology and the lack of a multi-dimensional perception and dynamic algorithm optimization collaborative mechanism in existing solid waste sorting methods. This achieves the technical effect of reducing identification blind spots, improving model generalization ability, and thus improving the accuracy, reliability, and intelligence level of solid waste sorting.

[0021] It should be understood that the description in this section is not intended to identify key or essential features of the embodiments of this application, nor is it intended to limit the scope of this application. Other features of this application will become readily apparent from the following description. Attached Figure Description

[0022] The accompanying drawings are provided to better understand this solution and do not constitute a limitation of this disclosure. Wherein: Figure 1 This is a flowchart illustrating a solid waste classification decision-making method provided in an embodiment of this disclosure; Figure 2 This is a schematic diagram of the structure of a solid waste classification decision device provided in an embodiment of the present disclosure; Figure 3This is a schematic diagram of the structure of a solid waste classification decision device provided in an embodiment of the present disclosure; Figure 4 A schematic block diagram of an example electronic device provided for embodiments of this disclosure. Detailed Implementation

[0023] The exemplary embodiments of this disclosure are described below with reference to the accompanying drawings, including various details of the embodiments to aid understanding, and should be considered merely exemplary. Therefore, those skilled in the art will recognize that various changes and modifications can be made to the embodiments described herein without departing from the scope and spirit of this disclosure. Similarly, for clarity and brevity, descriptions of well-known functions and structures are omitted in the following description.

[0024] The solid waste classification decision-making method, apparatus, electronic device, and storage medium of this disclosure are described below with reference to the accompanying drawings.

[0025] Figure 1 This is a flowchart illustrating a solid waste classification decision-making method provided in an embodiment of this disclosure.

[0026] like Figure 1 As shown, the method includes the following steps: Step 101: Collect visual and spectral information of the materials located on the conveying path; Visual information acquisition is accomplished by a high-resolution industrial camera, mounted directly above the conveyor path and secured with a stable bracket to ensure the lens always faces the center of the path, providing complete coverage of the main material transport area. As the material moves at a constant speed with the conveyor, the high-resolution camera continuously captures images, observing its shape, color distribution, surface texture, and other appearance features, generating clear and detailed visual information that directly reflects the material's external morphological differences. Spectral information acquisition is achieved using a near-infrared spectrometer, positioned to one side of the high-resolution camera. Its probe maintains a fixed, reasonable distance from the surface of the conveyor path to avoid collisions from excessively close proximity or compromised spectral accuracy from excessive distance.

[0027] The near-infrared spectrometer works synchronously with a high-resolution industrial camera to comprehensively scan the surface of materials as they pass through the collection area, acquiring their reflectance spectral information. This spectral information is closely related to the chemical composition and material properties of the materials, effectively distinguishing between different materials. The simultaneous acquisition of visual and spectral information achieves dual capture of both the material's external morphology and internal material characteristics. These two types of information complement and corroborate each other, providing a comprehensive data foundation for subsequent computational processing units to conduct integrated analysis, ensuring more accurate and reliable output of subsequent sorting instructions.

[0028] Step 102: Based on the visual information and the spectral information, a sorting instruction corresponding to the material category is generated through multimodal feature fusion analysis; The computational processing unit, as the core processing component of this step, synchronously receives visual information from the image acquisition unit and spectral information from the spectral analysis unit via a data cable, initiating a preset multimodal feature fusion analysis process. First, the computational processing unit extracts features from both types of information. From the visual information, it separates the material's external morphological features such as its outline, color distribution, and surface texture; these features directly reflect the differences in the material's physical appearance. From the spectral information, it analyzes feature parameters related to the material's chemical composition and material properties, such as reflectance data at different wavelengths; these parameters are key to distinguishing the material's intrinsic properties.

[0029] The fusion analysis process is not a simple feature overlay, but rather a deep mining of the correspondence between two types of features using specific algorithms. For example, a certain type of plastic material possesses specific color and outline features, as well as a unique spectral reflectance curve. By associating and matching these features, misjudgments that may arise from single-feature analysis are effectively eliminated, improving the accuracy of material category identification. The deep learning algorithm integrated within the computing unit is continuously optimized based on historical sorting records and manually labeled sample sets, constantly improving its ability to fuse and analyze multimodal features, enabling accurate identification of solid waste of different categories and materials.

[0030] Once the analysis is complete and the material category is identified, the computing and processing unit will automatically generate the corresponding sorting instructions. These instructions include key information such as the location of the sorting bin corresponding to the material and the working mode to be selected for the multi-functional fixture. This ensures that the sorting execution module can accurately pick up and place the materials according to the instructions, providing core support for the efficient advancement of the entire sorting process.

[0031] Step 103: In response to the sorting instruction, control the actuator to transfer the material to the collection area corresponding to the material category; The core of the actuator consists of a robotic arm and a multi-functional gripper. The robotic arm is a six-axis robot, securely fixed to the foundation beside the conveyor via flanges. Its flexible joint design allows for a full range of motion covering the entire conveyor and collection area, ensuring no blind spots during material transfer. Upon receiving a sorting instruction, the six-axis robot quickly and accurately moves to the target material on the conveyor based on the material location information contained in the instruction. The entire movement process is precisely controlled by the control system to ensure smooth and timely operation.

[0032] As a component that directly contacts the materials, the multi-functional gripper automatically selects the appropriate working mode based on the material characteristics specified in the sorting instructions to reliably grasp different types of materials. The collection area consists of multiple independent partitions of the sorting bins, each corresponding to a specific category of materials. The electronic tag readers and writers equipped in each partition can record the sorting status of the materials in real time.

[0033] After the multi-functional gripper picks up the material, the six-axis robot carries the material along a preset path to the corresponding collection area, accurately placing the material and completing the transfer operation. Once placement is complete, the electronic tag reader immediately records the relevant results of the sorting, including material category, placement area, and sorting time, and uploads this data to the data management module in real time, providing a complete basis for subsequent data statistics, queries, and system optimization. The entire process achieves full automation from command reception and mechanism movement to material placement and result recording, effectively improving the efficiency and accuracy of material transfer, ensuring that various materials can be accurately collected in their corresponding areas, and creating favorable conditions for subsequent resource utilization.

[0034] Step 104: If an abnormality occurs during the identification process, stop the sorting operation.

[0035] During the analysis and processing of visual and spectral information of materials by the intelligent recognition module, if the computing unit cannot accurately determine the material category based on the existing data, an anomaly detection mechanism is triggered. At this time, the computing unit immediately generates an alarm signal and sends a stop command to the sorting execution module. Upon receiving the command, the sorting execution module quickly terminates all sorting-related actions, and the six-axis robot stops moving, gripping, and other operations to ensure that materials are not mistakenly placed in non-corresponding collection areas. The core purpose of stopping the sorting operation is to block the sorting process under abnormal conditions, creating safe and stable conditions for subsequent anomaly handling, avoiding mis-sorting from affecting overall classification accuracy, and ensuring the reliability of multi-source solid waste sorting.

[0036] This design enables the system to have good fault tolerance when dealing with complex or special materials. By stopping operations in a timely manner, it avoids ineffective sorting and waste of resources, and provides operators with sufficient time to intervene, check abnormal situations, and input the correct sorting instructions. This ensures that the entire sorting process can resume in an orderly manner after the abnormality is handled, maintaining the overall sorting quality and efficiency of the system.

[0037] In some embodiments, the acquisition of visual and spectral information of materials located on the conveying path includes: Simultaneously acquire image data and spectral data of the material.

[0038] The data acquisition equipment includes a high-resolution industrial camera and a near-infrared spectrometer. These two devices work together in pre-set positions to ensure synchronized data acquisition. The high-resolution industrial camera is fixed directly above the conveyor path, its angle adjusted via a stable bracket to ensure the lens is always focused on the center of the path, capturing the material's appearance and shape during transport. The near-infrared spectrometer is positioned to one side of the industrial camera, its probe maintained at a fixed distance from the conveyor path surface to prevent variations in distance from affecting the consistency of spectral acquisition. When the material passes through the acquisition area at a constant speed with the conveyor, the system triggers a synchronous acquisition command. The high-resolution industrial camera immediately starts continuous shooting mode, quickly capturing details such as the material's shape, color distribution, and surface texture, forming clear and continuous image data. Simultaneously, the near-infrared spectrometer initiates its scanning function, performing a comprehensive and uniform spectral scan of the material surface to obtain reflectance spectral data reflecting the material's chemical composition and material properties.

[0039] This synchronous acquisition method effectively avoids misalignment between image and spectral data caused by acquisition time differences, ensuring that each set of image data corresponds to the spectral data of the same material. This allows for a precise correlation between the material's external morphological characteristics and its internal material properties. The two types of data obtained through synchronous acquisition fully preserve the core characteristic information of the material, effectively avoiding information loss that may occur with single acquisition methods. This provides a solid foundation for subsequent computational processing units to accurately extract features and perform in-depth fusion analysis, thereby ensuring the accuracy of material category determination and the reliability of sorting instructions.

[0040] In some embodiments, generating a sorting index corresponding to a material category based on the visual information and the spectral information through multimodal feature fusion analysis includes: Morphological features are extracted from the visual information, and material features are extracted from the spectral information. The morphological features and material features are fused together, and the material category is determined based on the fused features.

[0041] After receiving the synchronously acquired visual and spectral information, the computing and processing unit first initiates the feature extraction process. The morphological features extracted from the visual information cover the material's external physical characteristics, such as its shape, color distribution, surface texture, and size. These features directly reflect the differences in appearance between different materials; for example, the difference between plastic and metal in color and outline regularity, and the difference between paper and glass in surface texture, providing a basic basis for preliminary material differentiation. The material features extracted from the spectral information focus on the material's reflectance spectral parameters, including the reflectance peaks, valleys, and spectral curve shapes at different wavelengths. These parameters are closely related to the material's chemical composition and molecular structure, effectively revealing the material's intrinsic properties. For example, different types of plastics have unique spectral curve characteristics, and the spectral reflectance of metals and non-metals differs significantly, serving as a core basis for distinguishing material properties.

[0042] After feature extraction, the computational processing unit deeply integrates morphological and material features using a pre-defined fusion algorithm. This fusion is not a simple feature overlay, but rather establishes a mapping relationship between the two types of features. For example, it matches specific color and contour features of a material with its corresponding spectral reflectance curve to form a fused feature that comprehensively characterizes the material's properties, effectively avoiding the risk of misjudgment that may occur with single feature analysis. Subsequently, the computational processing unit calls its internally integrated algorithm model to compare and analyze the fused feature with the training dataset. The training dataset comes from historical sorting records and manually labeled sample sets. Through continuous learning and iteration, the algorithm has acquired the ability to recognize the fused features of various materials, enabling it to quickly and accurately determine the material's category. Once the material category is determined, the computational processing unit immediately generates corresponding sorting instructions, specifying key information such as the collection area to which the material should be transferred and the action path of the actuator, providing precise guidance for subsequent sorting execution and ensuring the efficiency and accuracy of the sorting operation.

[0043] In some embodiments, controlling the actuator to transfer materials to the collection area corresponding to the material category in response to the sorting instruction includes: Based on the physical properties of the material, the adsorption mode, clamping mode, or scooping mode of the end effector of the actuator is adaptively selected.

[0044] In response to sorting instructions, the control mechanism moves materials to the corresponding collection area. The core principle lies in adaptively selecting the suction, clamping, or scooping mode of the actuator's end effector based on the material's physical properties. This ensures stable and reliable material transfer, preventing detachment, damage, or transfer deviation. The actuator's end effector integrates multiple operating modes, with mode selection dynamically adjusted based entirely on the material's physical properties. These physical properties primarily include key characteristics such as weight, density, surface flatness, shape regularity, and looseness.

[0045] For lightweight materials with relatively flat surfaces and no obvious protrusions, such as thin plastic sheets and paper waste, the end effector automatically switches to adsorption mode. The pneumatic actuator adjusts the air pressure to generate a stable adsorption force, firmly adhering to the material surface and forming a secure connection, ensuring it won't fall off during movement and delivery. For materials of moderate weight with a certain degree of hardness and relatively regular shape, such as blocky metal parts or hard plastic blocks, the end effector switches to clamping mode. By adjusting the opening angle of the clamping claws, it precisely conforms to the material's shape and applies appropriate clamping force to fix the material in place. The clamping force can be adaptively adjusted according to the material's hardness and weight, ensuring the material doesn't loosen or slip, while also preventing damage due to excessive clamping force.

[0046] For heavier, irregularly shaped, or loose materials that are easily scattered, such as large pieces of loose granular construction waste, the end effector selects a shoveling mode. Using a shovel structure, it lifts the material from the bottom, smoothly scooping it up. The shovel's containment function prevents the material from scattering during transport. The end effector's mode switching is driven by a pneumatic actuator. Pressure sensors in the air circuit monitor air pressure changes in real time, ensuring smooth and continuous switching between modes and precise matching of pressure parameters with the material's physical properties. This adaptive mode selection based on material physical properties allows the actuator to adapt to different types of materials, significantly improving the stability and safety of gripping and transporting. Combined with the precise movement of the actuator, it ensures that materials are transported smoothly and efficiently to the corresponding collection area, guaranteeing smooth progress and accurate sorting.

[0047] In some embodiments, the method further includes: Record and store data related to the sorting process, and use the data to support adjustments to sorting strategies and updates to identification models.

[0048] Data related to the sorting process covers multiple dimensions, including loading volume data and material transmission speed data in the material receiving stage, raw image data, reflectance spectral information, and feature extraction results in the intelligent identification stage, sorting instruction content, material category determination results, placement zoning information, end-user operator mode selection records in the sorting execution stage, and abnormal material data and manual intervention instructions in the abnormal handling stage.

[0049] This data is transmitted to the data management module in real time and stored on the local server using a distributed storage architecture. This supports multi-node parallel processing, ensuring the storage security and retrieval efficiency of massive amounts of data. Simultaneously, it generates complete and continuous sorting logs, providing comprehensive evidence for subsequent traceability and analysis. The cloud platform connects to the local server via the internet, deploying a load balancing mechanism to ensure the stability of data transmission under high concurrency and synchronously backing up locally stored data, achieving dual data protection.

[0050] Based on the stored data, users can access the cloud platform via remote terminals to query various detailed data during the sorting process. By considering factors such as actual sorting efficiency, classification accuracy, and changes in material types, users can flexibly adjust sorting strategies, such as optimizing material placement priorities, adjusting the correspondence between sorting bin zones, and setting sorting speed parameters for different materials. Simultaneously, the stored data contains a large number of historical sorting records and labeled samples. This data can serve as a training data source for the recognition model. Users can upload new sample data according to actual needs, triggering the model update process. The model continuously optimizes its algorithm parameters through learning and iteration on the new data, improving the accuracy and adaptability of recognizing various materials, especially new materials. This allows the system to continuously adapt to the complex changes in multi-source solid waste, maintaining a consistently high level of efficient and accurate sorting.

[0051] In some embodiments, stopping the sorting operation when an anomaly occurs in the identification process includes: After an abnormal alarm is triggered, the drive system of the actuator is locked. Provides a human-machine interface to display identification information of abnormal materials and receive sorting instructions input by humans.

[0052] When the computing unit cannot accurately determine the material category based on visual and spectral information, it will immediately trigger an abnormal alarm signal. This signal is simultaneously transmitted to the drive system of the actuator, enabling the drive system to lock quickly. The drive system of the actuator includes the servo motor drive of the six-axis robot and the pneumatic actuator of the multi-functional gripper. After the locking mechanism is activated, the servo motor stops power output, the angles of each joint are kept fixed by the encoder, and the pneumatic actuator cuts off the air supply to prevent the gripper from malfunctioning. This fundamentally prevents problems such as material missorting, equipment collisions, or material damage caused by the actuator continuing to operate under abnormal conditions, ensuring the safety and order of the sorting site.

[0053] Simultaneously, the system automatically activates the human-machine interface, which retrieves and displays the identification information of abnormal materials in real time, including synchronously acquired image data and spectral information. The image data clearly presents the material's external features such as shape, color distribution, and surface texture, while the spectral information visually displays the material's reflectance parameters at different wavelengths in the form of curves, providing operators with comprehensive and detailed reference for quickly and accurately determining the material category. Operators manually input corresponding sorting instructions through the operation buttons or input modules on the human-machine interface. The instructions include key information such as the material's category and the designated collection area.

[0054] The human-machine interface boasts excellent ease of operation and clear data display, rapidly responding to operator input and transmitting manual sorting instructions to the computing unit in real time. The computing unit then issues these instructions to the execution mechanism, ensuring that abnormal materials are promptly and correctly sorted. The entire process is secured by a drive system lock, and the human-machine interface enables efficient integration of manual intervention. This not only solves the problem of sorting abnormal materials but also provides comprehensive data support for subsequent abnormal data labeling, storage, and algorithm optimization, further enhancing the system's fault tolerance and intelligence.

[0055] Corresponding to the solid waste classification decision-making method described above, this invention also proposes a solid waste classification decision-making device. Since the device embodiments of this invention correspond to the method embodiments described above, details not disclosed in the device embodiments can be referred to in the method embodiments described above, and will not be repeated here.

[0056] Figure 2 This is a schematic diagram of the structure of a solid waste classification decision device provided in an embodiment of this disclosure, as shown below. Figure 2 As shown, it includes: The acquisition unit 21 is also used to acquire visual and spectral information of materials located on the conveying path; The analysis unit 22 is also used to generate sorting instructions corresponding to the material category by multimodal feature fusion analysis based on the visual information and the spectral information; Control unit 23 is also configured to, in response to the sorting instruction, control the actuator to transfer the material to the collection area corresponding to the material category; The exception handling unit 24 is also used to stop the sorting operation when an exception occurs in the identification process.

[0057] Furthermore, in one possible implementation of this disclosure, the acquisition unit 21 is further configured to: Simultaneously acquire image data and spectral data of the material.

[0058] Furthermore, in one possible implementation of this disclosure, the analysis unit 22 is further configured to: Morphological features are extracted from the visual information, and material features are extracted from the spectral information. The morphological features and material features are fused together, and the material category is determined based on the fused features.

[0059] Furthermore, in one possible implementation of this disclosure, the control unit 23 is further configured to: Based on the physical properties of the material, the adsorption mode, clamping mode, or scooping mode of the end effector of the actuator is adaptively selected.

[0060] Furthermore, in one possible implementation of the embodiments of this disclosure, such as Figure 3 As shown, it also includes: The recording unit 25 is also used to record and store data related to the sorting process, and to support the adjustment of sorting strategies and the updating of identification models based on the data.

[0061] Furthermore, in one possible implementation of this disclosure embodiment, the exception handling unit 24 is further configured to: After an abnormal alarm is triggered, the drive system of the actuator is locked. Provides a human-machine interface to display identification information of abnormal materials and receive sorting instructions input by humans.

[0062] It should be noted that the foregoing explanation of the method embodiments also applies to the apparatus of the embodiments of this disclosure, and the principle is the same. Therefore, the embodiments of this disclosure are not limited thereto.

[0063] According to embodiments of this disclosure, this disclosure also provides an electronic device, a readable storage medium, and a computer program product.

[0064] Figure 4 A schematic block diagram of an example electronic device 400 that can be used to implement embodiments of the present disclosure is shown. The electronic device is intended to represent various forms of digital computers, such as laptop computers, desktop computers, workstations, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. The electronic device may also represent various forms of mobile devices, such as personal digital assistants, cellular phones, smartphones, wearable devices, and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely illustrative and are not intended to limit the implementation of the present disclosure described and / or claimed herein.

[0065] like Figure 4 As shown, device 400 includes a computing unit 401, which can perform various appropriate actions and processes based on a computer program stored in ROM (Read-Only Memory) 402 or a computer program loaded from storage unit 408 into RAM (Random Access Memory) 403. RAM 403 may also store various programs and data required for the operation of device 400. The computing unit 401, ROM 402, and RAM 403 are interconnected via bus 404. I / O (Input / Output) interface 405 is also connected to bus 404.

[0066] Multiple components in device 400 are connected to I / O interface 405, including: input unit 406, such as keyboard, mouse, etc.; output unit 407, such as various types of monitors, speakers, etc.; storage unit 408, such as disk, optical disk, etc.; and communication unit 409, such as network card, modem, wireless transceiver, etc. Communication unit 409 allows device 400 to exchange information / data with other devices through computer networks such as the Internet and / or various telecommunications networks.

[0067] The computing unit 401 can be a variety of general-purpose and / or special-purpose processing components with processing and computing capabilities. Some examples of the computing unit 401 include, but are not limited to, CPUs (Central Processing Units), GPUs (Graphics Processing Units), various special-purpose AI (Artificial Intelligence) computing chips, various computing units running machine learning model algorithms, DSPs (Digital Signal Processors), and any suitable processor, controller, microcontroller, etc. The computing unit 401 performs the various methods and processes described above, such as solid waste classification decision methods. For example, in some embodiments, the solid waste classification decision method may be implemented as a computer software program tangibly contained in a machine-readable medium, such as storage unit 408. In some embodiments, part or all of the computer program may be loaded and / or installed on device 400 via ROM 402 and / or communication unit 409. When the computer program is loaded into RAM 403 and executed by the computing unit 401, one or more steps of the methods described above may be performed. Alternatively, in other embodiments, the computing unit 401 may be configured to perform the aforementioned solid waste classification decision method by any other suitable means (e.g., by means of firmware).

[0068] Various implementations of the systems and techniques described above herein can be implemented in digital electronic circuit systems, integrated circuit systems, FPGAs (Field Programmable Gate Arrays), ASICs (Application-Specific Integrated Circuits), ASSPs (Application-Specific Standard Products), SOCs (System-on-Chips), CPLDs (Complex Programmable Logic Devices), computer hardware, firmware, software, and / or combinations thereof. These various implementations may include implementations in one or more computer programs that can be executed and / or interpreted on a programmable system including at least one programmable processor, which may be a dedicated or general-purpose programmable processor, capable of receiving data and instructions from a storage system, at least one input device, and at least one output device, and transmitting data and instructions to the storage system, the at least one input device, and the at least one output device.

[0069] The program code used to implement the methods of this disclosure may be written in any combination of one or more programming languages. This program code may be provided to a processor or controller of a general-purpose computer, special-purpose computer, or other programmable data processing apparatus, such that when executed by the processor or controller, the program code causes the functions / operations specified in the flowcharts and / or block diagrams to be implemented. The program code may be executed entirely on a machine, partially on a machine, as a standalone software package partially on a machine and partially on a remote machine, or entirely on a remote machine or server.

[0070] In the context of this disclosure, a machine-readable medium can be a tangible medium that may contain or store a program for use by or in conjunction with an instruction execution system, apparatus, or device. A machine-readable medium can be a machine-readable signal medium or a machine-readable storage medium. A machine-readable medium can be, but is not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, apparatus, or devices, or any suitable combination of the foregoing. More specific examples of machine-readable storage media include electrical connections based on one or more wires, portable computer disks, hard disks, RAM, ROM, EPROM (Electrically Programmable Read-Only Memory) or flash memory, optical fiber, CD-ROM (Compact Disc Read-Only Memory), optical storage devices, magnetic storage devices, or any suitable combination of the foregoing.

[0071] To provide interaction with a user, the systems and techniques described herein can be implemented on a computer having: a display device for displaying information to the user (e.g., a CRT (Cathode-Ray Tube) or LCD (Liquid Crystal Display) monitor); and a keyboard and pointing device (e.g., a mouse or trackball) through which the user provides input to the computer. Other types of devices can also be used to provide interaction with the user; for example, feedback provided to the user can be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user can be received in any form (including sound input, voice input, or tactile input).

[0072] The systems and technologies described herein can be implemented in computing systems that include backend components (e.g., as data servers), or computing systems that include middleware components (e.g., application servers), or computing systems that include frontend components (e.g., user computers with graphical user interfaces or web browsers through which users can interact with implementations of the systems and technologies described herein), or any combination of such backend, middleware, or frontend components. The components of the system can be interconnected via digital data communication of any form or medium (e.g., communication networks). Examples of communication networks include LANs (Local Area Networks), WANs (Wide Area Networks), the Internet, and blockchain networks.

[0073] Computer systems can include clients and servers. Clients and servers are generally geographically separated and typically interact via communication networks. The client-server relationship is created by computer programs running on the respective computers and having a client-server relationship with each other. A server can be a cloud server, also known as a cloud computing server or cloud host, a hosting product within the cloud computing service system that addresses the shortcomings of traditional physical hosts and VPS (Virtual Private Server) services, such as high management difficulty and weak business scalability. Servers can also be servers for distributed systems or servers incorporating blockchain technology.

[0074] It's important to note that artificial intelligence (AI) is the study of enabling computers to simulate certain human thought processes and intelligent behaviors (such as learning, reasoning, thinking, and planning). It encompasses both hardware and software technologies. AI hardware technologies generally include sensors, dedicated AI chips, cloud computing, distributed storage, and big data processing. AI software technologies primarily include computer vision, speech recognition, natural language processing, machine learning / deep learning, big data processing, and knowledge graph technologies.

[0075] It should be understood that the various forms of processes shown above can be used to rearrange, add, or delete steps. For example, the steps described in this disclosure can be executed in parallel, sequentially, or in different orders, as long as the desired result of the technical solution disclosed in this disclosure can be achieved, and this is not limited herein.

[0076] The specific embodiments described above do not constitute a limitation on the scope of protection of this disclosure. Those skilled in the art should understand that various modifications, combinations, sub-combinations, and substitutions can be made according to design requirements and other factors. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of this disclosure should be included within the scope of protection of this disclosure.

Claims

1. A solid waste classification decision-making method, characterized in that, include: Collect visual and spectral information of materials located along the conveying path; Based on the visual information and the spectral information, a sorting instruction corresponding to the material category is generated through multimodal feature fusion analysis; In response to the sorting instruction, the control actuator moves the material to the collection area corresponding to the material category; If an abnormality occurs during the identification process, the sorting operation shall be stopped.

2. The method according to claim 1, characterized in that, The acquisition of visual and spectral information of materials located on the conveying path includes: Simultaneously acquire image data and spectral data of the material.

3. The method according to claim 1 or 2, characterized in that, The step of generating sorting indicators corresponding to material categories based on the visual information and the spectral information through multimodal feature fusion analysis includes: Morphological features are extracted from the visual information, and material features are extracted from the spectral information. The morphological features and material features are fused together, and the material category is determined based on the fused features.

4. The method according to claim 1, characterized in that, The step of controlling the actuator to transfer materials to the collection area corresponding to the material category in response to the sorting instruction includes: Based on the physical properties of the material, the adsorption mode, clamping mode, or scooping mode of the end effector of the actuator is adaptively selected.

5. The method according to claim 1, characterized in that, The method further includes: Record and store data related to the sorting process, and use the data to support adjustments to sorting strategies and updates to identification models.

6. The method according to claim 1, characterized in that, The step of stopping the sorting operation when an abnormality occurs in the identification process includes: After an abnormal alarm is triggered, the drive system of the actuator is locked. Provides a human-machine interface to display identification information of abnormal materials and receive sorting instructions input by humans.

7. A solid waste classification decision-making device, characterized in that, include: The acquisition unit is also used to acquire visual and spectral information of materials located on the conveying path; The analysis unit is also used to generate sorting instructions corresponding to the material category by performing multimodal feature fusion analysis based on the visual information and the spectral information. The control unit is also configured to, in response to the sorting instruction, control the actuator to transfer the material to the collection area corresponding to the material category; The exception handling unit is also used to stop the sorting operation when an exception occurs in the identification process.

8. An electronic device, characterized in that, include: At least one processor; as well as A memory communicatively connected to the at least one processor; wherein, The memory stores instructions that can be executed by the at least one processor to enable the at least one processor to perform the method of any one of claims 1-6.

9. A non-transitory computer-readable storage medium storing computer instructions, characterized in that, The computer instructions are used to cause the computer to perform the method according to any one of claims 1-6.

10. A computer program product, characterized in that, Includes a computer program that, when executed by a processor, implements the method according to any one of claims 1-6.

Images