Metal scrap sorting method, controller and system

Abstract

The invention discloses a metal scrap sorting method, a controller and a system. The method comprises the steps that multi-dimensional information of metal scraps on a conveying belt reaching a detection area and position information and speed information of the metal scraps on the conveying belt are obtained; obtaining an identification result according to the multi-dimensional information; the identification result is used for representing the material of the metal scrap; and based on the recognition result, the position information and the speed information, when the metal scraps reach the sorting area, the metal scraps are sorted to corresponding target areas. According to the method, the multi-dimensional information of the metal scraps is obtained, classification is carried out based on the multi-dimensional information, and different materials of metal are determined for classification. And the identification accuracy and the sorting purity level of the metal scraps are improved.

Description

Technical Field

[0001] This application relates to the field of metal sorting technology, and in particular to a method, controller and system for sorting metal waste. Background Technology

[0002] Metal scrap resources such as copper, aluminum, brass, and lead in metal waste have high recycling value. The efficient recycling and purification of recycled metal waste not only helps alleviate mineral resource shortages and reduce smelting energy consumption, but also effectively reduces environmental pollution, meeting the strategic needs of sustainable development.

[0003] In related technologies, the recycling and sorting of scrap metal mainly relies on manual picking or physical methods such as magnetic separation and eddy current separation for rough classification. However, these technologies are effective in separating magnetic materials such as steel, but have limited ability to finely sort non-ferrous metals such as aluminum, copper, and brass. Photoelectric color sorting using visible light is another related technology. While this technology is effective in sorting large-particle-size, clean-surfaced waste, its accuracy drops significantly when impurities, oxide layers, or similar colors of different alloys are present on the waste surface, making it difficult to meet the requirements of industrial production for sorting purity and efficiency.

[0004] In addition, the high reflectivity, weak surface texture, and indistinct spectral characteristics of metal materials themselves also lead to unstable recognition results and large sorting errors by a single visual inspection method, which seriously restricts the level of recycling of waste metal materials. Summary of the Invention

[0005] This application provides a method, controller, and system for sorting metal scrap, in order to at least solve the above-mentioned technical problems existing in the prior art.

[0006] According to a first aspect of this application, a method for sorting metal scrap is provided, comprising:

[0007] Acquire multidimensional information about the metal scrap on the conveyor belt that arrives at the detection area, as well as the position and speed information of the metal scrap on the conveyor belt;

[0008] Based on the multidimensional information, an identification result is obtained; the identification result is used to characterize the material of the metal scrap.

[0009] Based on the identification results, the location information, and the speed information, when the metal waste arrives at the sorting area, the metal waste is sorted into the corresponding target area.

[0010] In one possible implementation, the multidimensional information includes the shape information, texture information, and spectral reflectance information of the metal scrap; acquiring the multidimensional information of the metal scrap on the conveyor belt reaching the detection area, as well as the position and speed information of the metal scrap on the conveyor belt, includes:

[0011] The metal waste in the detection area is illuminated using an array of light sources;

[0012] Simultaneously trigger visible light cameras, hyperspectral cameras, and depth cameras to acquire shape, texture, and spectral reflectance information of the metal scrap.

[0013] The speed of the conveyor belt is obtained by an encoder; the encoder is mounted on the conveyor belt.

[0014] In one possible implementation, the simultaneous triggering of a visible light camera, a hyperspectral camera, and a depth camera to acquire shape information, texture information, and spectral reflectance information of the metal scrap includes:

[0015] A color image of the metal scrap is acquired using a visible light camera; the color image includes texture information and the position information of the metal scrap on the conveyor belt.

[0016] The hyperspectral camera is used to acquire hyperspectral images, and the spectral reflectance information of each pixel in the hyperspectral image is obtained; the spectral reflectance information is used to determine the material of the metal scrap.

[0017] The shape information of the metal scrap on the conveyor belt is obtained by using a depth camera; the shape information includes the height, volume and three-dimensional contour of the metal scrap.

[0018] In one possible implementation, obtaining the recognition result based on the multidimensional information includes:

[0019] Based on the multidimensional information, multidimensional feature data is obtained; the multidimensional feature data includes texture feature data, hyperspectral feature data, and shape feature data.

[0020] The texture feature data, hyperspectral feature data, and shape feature data are input into a pre-constructed multimodal feature fusion classification model to obtain the recognition result.

[0021] In one possible implementation, the step of sorting the metal waste to the corresponding target area when the metal waste arrives at the sorting area based on the identification result, the location information, and the speed information includes:

[0022] The travel distance of the conveyor belt is calculated based on the speed information;

[0023] Based on the position information of the metal waste on the conveyor belt and the moving distance, it is determined whether the metal waste has reached the sorting area;

[0024] In response to the metal waste reaching the sorting area, the metal waste is sorted to the corresponding target area based on the identification result.

[0025] In one embodiment, the multimodal feature fusion classification model includes:

[0026] A self-supervised learning mechanism is used to update the model parameters of the multimodal feature fusion classification model based on the difference between the recognition results and the true labels.

[0027] In one embodiment, the array light source is an LED array light source, and an elliptical reflector is provided outside the LED array light source;

[0028] The conveyor belt moves at a constant speed.

[0029] According to a second aspect of this application, a controller is provided, comprising:

[0030] Memory, on which executable programs are stored;

[0031] A processor for executing the executable program in the memory to implement the steps of the method described in any embodiment.

[0032] According to a third aspect of this application, a metal scrap sorting system is provided, comprising:

[0033] The controller described in any of the above embodiments; and

[0034] A conveying mechanism and a sorting mechanism, wherein the conveying mechanism and the sorting mechanism are respectively connected to the controller;

[0035] The conveying mechanism includes: a vibrating feeder, a conveyor belt, and a motor, and the conveyor belt is equipped with an encoder;

[0036] The vibrating feeder is used to evenly spread the organometallic material onto the conveyor belt.

[0037] The motor drives the conveyor belt to move at a constant speed, conveying the organometallic material to the detection area;

[0038] The encoder is used to collect the conveyor position and speed of the conveyor belt and send them to the controller;

[0039] The sorting mechanism includes: multiple sets of airflow nozzles and a sorting funnel;

[0040] Multiple sets of airflow nozzles are evenly arranged along the conveyor belt to change the trajectory of the metal waste reaching the sorting area by airflow, so that the metal waste moves to the corresponding sorting funnel.

[0041] A sorting funnel is used to sort and store metal scrap.

[0042] In one embodiment, the sorting mechanism further includes:

[0043] A robotic arm is used to grip or push scrap metal into a target sorting funnel.

[0044] By utilizing the technical solution of this application, multi-dimensional information of metal waste is acquired, and classification is performed based on this multi-dimensional information to determine the different materials of the metals for further classification. This improves the accuracy of metal waste identification and the level of sorting purity.

[0045] 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

[0046] The above and other objects, features, and advantages of exemplary embodiments of this application will become readily apparent from the following detailed description taken in conjunction with the accompanying drawings. Several embodiments of this application are illustrated in the drawings by way of example and not limitation, in which:

[0047] In the accompanying drawings, the same or corresponding reference numerals indicate the same or corresponding parts.

[0048] Figure 1 A schematic diagram illustrating the implementation process of the transmission mechanism in an embodiment of this application is shown;

[0049] Figure 2 A schematic diagram illustrating the implementation process of the sorting mechanism in an embodiment of this application is shown;

[0050] Figure 3 A schematic diagram illustrating the implementation flow of the metal scrap sorting method in an embodiment of this application is shown;

[0051] Figure 4 A schematic diagram of the array light source structure in an embodiment of this application is shown;

[0052] Figure 5 A schematic diagram of the multidimensional information fusion classification process in an embodiment of this application is shown;

[0053] Figure 6 A schematic diagram of the metal scrap sorting system in an embodiment of this application is shown;

[0054] Figure 7 A schematic diagram of the composition structure of an electronic device according to an embodiment of this application is shown. Detailed Implementation

[0055] To make the objectives, features, and advantages of this application more apparent and understandable, the technical solutions in the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0056] To make the objectives, technical solutions, and advantages of this application clearer, the application will be further described in detail below with reference to the accompanying drawings. The described embodiments should not be regarded as limitations on this application. All other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0057] In the following description, references are made to “some embodiments,” which describe a subset of all possible embodiments. However, it is understood that “some embodiments” may be the same subset or different subsets of all possible embodiments and may be combined with each other without conflict.

[0058] In the following description, the terms "first" and "second" are used merely to distinguish similar objects and do not represent a specific ordering of objects. It is understood that "first" and "second" may be interchanged in a specific order or sequence where permitted, so that the embodiments of this application described herein can be implemented in an order other than that illustrated or described herein.

[0059] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application belongs. The terminology used herein is for the purpose of describing embodiments of this application only and is not intended to limit this application.

[0060] It should be understood that in the various embodiments of this application, the sequence number of each implementation process does not imply the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of this application.

[0061] The following description, in conjunction with the accompanying drawings, introduces a metal scrap sorting method, controller, and system provided in this application.

[0062] This application provides a metal scrap sorting system, including:

[0063] Controller; and

[0064] A conveying mechanism and a sorting mechanism, wherein the conveying mechanism and the sorting mechanism are respectively connected to the controller;

[0065] The conveying mechanism includes: a vibrating feeder, a conveyor belt, and a motor, and the conveyor belt is equipped with an encoder;

[0066] The vibrating feeder is used to evenly spread the organometallic material onto the conveyor belt.

[0067] The motor drives the conveyor belt to move at a constant speed, conveying the organometallic material to the detection area;

[0068] The encoder is used to collect the conveyor position and speed of the conveyor belt and send them to the controller;

[0069] The sorting mechanism includes: multiple sets of airflow nozzles and a sorting funnel;

[0070] Multiple sets of airflow nozzles are evenly arranged along the conveyor belt to change the trajectory of the metal waste reaching the sorting area by airflow, so that the metal waste moves to the corresponding sorting funnel.

[0071] A sorting funnel is used to sort and store metal scrap.

[0072] In this application, such as Figure 1 As shown, the vibrating feeder may also include an inclined hopper containing metal scrap. The vibrating feeder receives the metal scrap falling from the inclined hopper, spreading the mixed scrap evenly to prevent stacking. The conveyor belt can be a planar belt conveyor, transporting the scrap to the detection area at a uniform speed. An encoder is installed at the end of the conveyor belt, capable of collecting the belt position and speed in real time and sending it to the controller. The controller can calculate the movement position of the metal scrap for time and position synchronization control of subsequent sorting actions. In this application, the speed of the conveyor belt can be adjusted according to actual needs and is not limited herein. For example, the conveyor belt speed can be set to 2 meters per second, or it can be flexibly set according to the size of the metal scrap and the required sorting accuracy.

[0073] like Figure 2 As shown, when the conveyor belt moves to the detection area, the controller analyzes the metal waste to determine the type of metal waste. When the metal waste moves with the conveyor belt to the target position in the sorting area, the controller sends a trigger signal to multiple sets of airflow nozzles of the sorting mechanism. The airflow nozzles spray compressed air instantly, changing the trajectory of the metal waste and accurately blowing it into the corresponding sorting hopper.

[0074] It should be noted that if the metal scrap is large or irregularly shaped, a robotic arm can be installed to precisely grip or push the metal scrap to the target sorting hopper, improving sorting accuracy. The controller in this application implements the metal scrap sorting method.

[0075] The controller provided in this application includes:

[0076] Memory, on which executable programs are stored;

[0077] A processor for executing the executable program in the memory to implement a metal scrap sorting method.

[0078] Specifically, such as Figure 3 As shown, this application provides a method for sorting metal scrap, including:

[0079] S301, acquire multi-dimensional information of the metal scrap on the conveyor belt that has arrived at the detection area, as well as the position and speed information of the metal scrap on the conveyor belt;

[0080] In this application, multiple sensors or cameras can be installed above the conveyor belt. Multiple sensors acquire multi-dimensional information about the metal scrap on the conveyor belt, as well as its position on the conveyor belt. An encoder at the end of the conveyor belt obtains the conveyor belt's speed.

[0081] S302, Based on the multidimensional information, an identification result is obtained; the identification result is used to characterize the material of the metal scrap.

[0082] This application utilizes a classification model to categorize multidimensional information and obtain the material classification of metal scrap. Alternatively, other methods can also be used to classify the material of metal scrap.

[0083] S303, based on the identification result, the location information and the speed information, when the metal waste arrives at the sorting area, the metal waste is sorted to the corresponding target area.

[0084] After determining the material of the metal scrap, the conveyor belt travels a certain distance based on its position and speed. This allows the system to determine whether the metal scrap has reached the correct position. If it has, the sorting mechanism sorts the metal scrap into the target sorting hopper.

[0085] The metal scrap sorting method provided in this application acquires multidimensional information about the metal scrap, its position on the conveyor belt, and the speed of the conveyor belt when the scrap reaches the detection area. The multidimensional information is analyzed to obtain identification results. Combining the identification results, position information, and speed information, the metal scrap is sorted into the target classification hopper when it is determined that it has reached the sorting area. This application improves the accuracy of metal scrap identification and the purity of sorting by acquiring multidimensional information about the metal scrap and classifying it based on this information to determine different metal materials.

[0086] In some embodiments, such as Figure 4As shown, the multidimensional information includes the shape information, texture information, and spectral reflectance information of the metal scrap; the acquisition of multidimensional information of the metal scrap on the conveyor belt reaching the detection area, as well as the position and speed information of the metal scrap on the conveyor belt, includes:

[0087] The metal waste in the detection area is illuminated using an array of light sources;

[0088] Simultaneously trigger visible light cameras, hyperspectral cameras, and depth cameras to acquire shape, texture, and spectral reflectance information of the metal scrap.

[0089] The speed of the conveyor belt is obtained by an encoder; the encoder is mounted on the conveyor belt.

[0090] like Figure 5 As shown, the array light source in this application uses an LED array light source, which can eliminate ambient light interference and suppress the influence of metal surface reflection on image quality. The visible light camera provided in this application refers to an imaging device that captures the surface texture features of metal scrap. Specifically, it can be an industrial camera with a CMOS sensor or CCD sensor, or an image acquisition device capable of acquiring color distribution and surface texture feature data of metal scrap. In this application, the hyperspectral camera refers to an imaging device that acquires the spectral reflectance characteristics of metal scrap. For example, it can be a pushbroom hyperspectral imager or a snapshot hyperspectral imager. The hyperspectral imager can acquire the continuous spectral curve of each pixel to distinguish material differences. The depth camera in this application can be an imaging device that acquires three-dimensional spatial information of metal scrap. The depth camera can be a structured light camera or a time-of-flight camera, capable of acquiring the height, volume, and three-dimensional contour parameters of the metal scrap. The encoder can be a sensing device that detects the motion state of the conveyor belt, for example, it can be implemented as a rotary encoder or a linear encoder ruler, capable of acquiring the linear velocity and displacement parameters of the conveyor belt in real time.

[0091] For example, such as Figure 6As shown, this application installs an LED array light source above the detection area of ​​the conveyor belt. The elliptical reflector on the LED array light source can uniformly project light onto the detection area. A visible light camera, a hyperspectral camera, and a depth camera are arranged in a triangle above the conveyor belt, with their optical axes intersecting at the center point of the detection area. In this application, an encoder is installed at the end of the conveyor belt drive roller. When the vibrating feeder evenly spreads the metal scrap onto the conveyor belt and it enters the detection area, the PLC controller simultaneously triggers the three cameras to acquire images and records the encoder's pulse count value at that moment. The visible light camera acquires RGB images and extracts texture features; the hyperspectral camera acquires spectral data from 400-2500nm and extracts reflectivity in characteristic bands; the depth camera acquires point cloud data, which is then filtered to calculate volume parameters. The encoder pulse signal is converted to obtain the real-time speed of the conveyor belt, and combined with the image acquisition time, the dynamic coordinates of the metal scrap are determined.

[0092] This application achieves simultaneous acquisition of multi-dimensional features from metal scrap. Uniform illumination from the array light source ensures image quality, and synchronous triggering of the three cameras ensures consistency of multi-modal data. This application overcomes the limitations of a single sensor, realizing collaborative analysis of multi-dimensional features, thereby improving the identification accuracy of complex-material metal scrap and significantly enhancing the environmental adaptability and sorting precision of the sorting system.

[0093] In some embodiments, the simultaneous triggering of a visible light camera, a hyperspectral camera, and a depth camera to acquire shape information, texture information, and spectral reflectance information of the metal scrap includes:

[0094] A color image of the metal scrap is acquired using a visible light camera; the color image includes texture information and the position information of the metal scrap on the conveyor belt.

[0095] The hyperspectral camera is used to acquire hyperspectral images, and the spectral reflectance information of each pixel in the hyperspectral image is obtained; the spectral reflectance information is used to determine the material of the metal scrap.

[0096] The shape information of the metal scrap on the conveyor belt is obtained by using a depth camera; the shape information includes the height, volume and three-dimensional contour of the metal scrap.

[0097] In this application, a visible light camera can capture electromagnetic radiation in the visible light band (400-700nm), such as a CMOS or CCD sensor, to acquire texture features formed by oxide layers, coatings, etc., on metal surfaces. A hyperspectral camera can acquire continuous narrow-band spectral data. This camera can be a pushbroom or snapshot imaging camera, capturing characteristic absorption peaks of different metals in specific bands by analyzing spectral reflectance pixel-by-pixel. For example, copper exhibits a sharp drop in reflectance near 800nm, and aluminum shows a flat spectral characteristic in the 400-700nm range. A depth camera is capable of acquiring three-dimensional spatial information of an object. A depth camera can be a Time-of-Flight (TOF) camera, and the point cloud data obtained from the TOF camera can be used to construct the height, volume, and three-dimensional contour parameters of the metal scrap.

[0098] This application can simultaneously trigger a visible light camera, a hyperspectral camera, and a depth camera to acquire information at the same time, ensuring that color images, hyperspectral data, and 3D morphological information remain synchronized at the time of acquisition. The visible light camera, hyperspectral camera, and depth camera in this application can synchronously image via a controller's trigger signal, ensuring strict alignment of multi-source data in timestamps and spatial coordinates.

[0099] This application achieves multi-source data synergistic optimization by simultaneously activating visible light cameras, hyperspectral cameras, and depth cameras, aligning the data in time and space. This improves the accuracy of metal scrap identification and effectively solves the problem of misjudging metals with similar colors, thus enhancing the intelligence level of the sorting system.

[0100] In some embodiments, obtaining the recognition result based on the multidimensional information includes:

[0101] Based on the multidimensional information, multidimensional feature data is obtained; the multidimensional feature data includes texture feature data, hyperspectral feature data, and shape feature data.

[0102] The texture feature data, hyperspectral feature data, and shape feature data are input into a pre-constructed multimodal feature fusion classification model to obtain the recognition result.

[0103] It should be noted that the multimodal feature fusion classification model in this application can be a deep learning model based on deep learning, a hybrid model combining convolutional neural networks and attention mechanisms, a multimodal feature interaction network based on the Transformer architecture, or other network training structures; this application does not limit these possibilities. Specifically, texture feature data refers to the microstructural features of metal surfaces acquired by a visible light camera, which can be quantized using local binarization patterns or gray-level co-occurrence matrices; hyperspectral feature data refers to the spectral reflectance characteristics of metal materials collected by a hyperspectral camera, which can be used for feature dimensionality reduction and classification using principal component analysis or support vector machines; shape feature data refers to the three-dimensional geometric morphology parameters of metal scrap acquired by a depth camera, which can be used to extract volume, curvature, and contour features using point cloud data processing techniques.

[0104] For example, this application sets up a multimodal sensing array consisting of a visible light camera, a hyperspectral camera, and a depth camera in the detection area to acquire texture, spectral, and three-dimensional morphological data of the metal scrap through a synchronous triggering mechanism. Subsequently, the aforementioned feature data is input into a fusion model containing a cross-modal attention module, and weighted fusion is performed using the correlation between features. The final model outputs the material recognition result. Based on the recognition result and the conveyor belt position information fed back by the encoder, the controller precisely controls the opening and closing sequence of the airflow nozzles to achieve directional sorting of the metal scrap.

[0105] In some embodiments, the step of sorting the metal waste to the corresponding target area when the metal waste arrives at the sorting area based on the identification result, the location information, and the speed information includes:

[0106] The travel distance of the conveyor belt is calculated based on the speed information;

[0107] Based on the position information of the metal waste on the conveyor belt and the moving distance, it is determined whether the metal waste has reached the sorting area;

[0108] In response to the metal waste reaching the sorting area, the metal waste is sorted to the corresponding target area based on the identification result.

[0109] This application calculates the movement distance by correlating speed and location information, thereby determining the precise coordinates of the metal scrap during dynamic conveying, effectively eliminating positioning deviations caused by conveyor belt vibration and speed fluctuations. Finally, when the judgment confirms that the metal scrap has entered the sorting area, the sorting execution mechanism is activated to sort scrap of different materials to the preset target area. This application's technical solution solves the sorting failure problem caused by delayed response in traditional methods, and also avoids path conflicts during multi-target sorting through spatial coordinate mapping.

[0110] This application achieves precise positioning and sorting of metal scrap in dynamic conveying environments. By dynamically calculating the correlation between speed and position information, the impact of conveyor belt speed fluctuations on positioning accuracy is effectively eliminated. At the same time, based on a spatial coordinate mapping mechanism, path conflicts during multi-target sorting are avoided, significantly improving the sorting success rate and system stability under complex working conditions.

[0111] In some embodiments, the multimodal feature fusion classification model includes:

[0112] A self-supervised learning mechanism is used to update the model parameters of the multimodal feature fusion classification model based on the difference between the recognition results and the true labels.

[0113] Self-supervised learning mechanisms are optimization mechanisms that automatically adjust model parameters based on the error signal between the recognition output and the manually labeled true labels. They can also update parameters using a combination of backpropagation algorithms, dynamic gradient descent, or online incremental learning strategies.

[0114] This application utilizes a multimodal feature fusion model to effectively address the performance degradation issue caused by differences in the characteristics of different batches of waste. The multimodal feature fusion model significantly improves the ability to distinguish between copper and aluminum alloys of similar color schemes, enhances sorting purity stability, and reduces the impact of surface contamination, oxide layer changes, and other factors on recognition accuracy.

[0115] This application provides a stable and continuous illumination base through an LED array light source. An elliptical reflector redistributes the light emitted from the LED array light source, forming a uniformly diffused annular light field. The LED array light source provided by this application can eliminate overexposure caused by highly reflective areas on the surface of metal scrap, and enhance the light coverage of details on uneven surfaces. This improves the clarity of texture information acquired by a visible light camera, the signal-to-noise ratio of data acquired by a hyperspectral camera, and the completeness of the three-dimensional contours acquired by a depth camera.

[0116] For example, the LED array light source provided in this application can be a ring array composed of 24 groups of 5W high-power LED beads. The elliptical reflector is made of anodized aluminum, and its inner surface is polished at the nanoscale to enhance the reflection efficiency.

[0117] According to embodiments of this application, this application also provides an electronic device and a readable storage medium.

[0118] The electronic device includes at least one processor and a memory communicatively connected to the at least one processor. The memory stores instructions executable by the at least one processor, which, when executed, enable the at least one processor to perform the metal scrap sorting method described in this application. The computer instructions are used to cause the computer to perform the metal scrap sorting method described in this application.

[0119] This application also provides a computer program product, including a computer program / instructions, which, when executed by a processor, implement the metal scrap sorting method of this application.

[0120] Figure 7 A schematic block diagram of an example electronic device 800 that can be used to implement embodiments of this application 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 processors, 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 application described and / or claimed herein.

[0121] like Figure 7 As shown, device 800 includes a computing unit 801, which can perform various appropriate actions and processes based on a computer program stored in read-only memory (ROM) 802 or a computer program loaded from storage unit 808 into random access memory (RAM) 803. RAM 803 may also store various programs and data required for the operation of device 800. The computing unit 801, ROM 802, and RAM 803 are interconnected via bus 804. Input / output (I / O) interface 805 is also connected to bus 804.

[0122] Multiple components in device 800 are connected to I / O interface 805, including: input unit 806, such as keyboard, mouse, etc.; output unit 807, such as various types of monitors, speakers, etc.; storage unit 808, such as disk, optical disk, etc.; and communication unit 809, such as network card, modem, wireless transceiver, etc. Communication unit 809 allows device 800 to exchange information / data with other devices through computer networks such as the Internet and / or various telecommunications networks.

[0123] The computing unit 801 can be a variety of general-purpose and / or special-purpose processing components with processing and computing capabilities. Some examples of the computing unit 801 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various special-purpose artificial intelligence (AI) computing chips, various computing units running machine learning model algorithms, a digital signal processor (DSP), and any suitable processor, controller, microcontroller, etc. The computing unit 801 performs the various methods and processes described above, such as a metal scrap sorting method. For example, in some embodiments, the metal scrap sorting method can be implemented as a computer software program tangibly contained in a machine-readable medium, such as storage unit 808. In some embodiments, part or all of the computer program can be loaded and / or installed on device 800 via ROM 802 and / or communication unit 809. When the computer program is loaded into RAM 803 and executed by the computing unit 801, one or more steps of the metal scrap sorting method described above can be performed. Alternatively, in other embodiments, the computing unit 801 can be configured to perform the metal scrap sorting method by any other suitable means (e.g., by means of firmware).

[0124] Various embodiments of the systems and techniques described above herein can be implemented in digital electronic circuit systems, integrated circuit systems, field-programmable gate arrays (FPGAs), application-specific integrated circuits (ASICs), application-specific standard products (ASSPs), systems-on-a-chip (SoCs), payload-programmable logic devices (CPLDs), computer hardware, firmware, software, and / or combinations thereof. These various embodiments 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.

[0125] The program code used to implement the methods of this application 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 device, such that when executed by the processor or controller, the functions / operations specified in the flowcharts and / or block diagrams are 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.

[0126] In the context of this application, 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. Machine-readable media 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, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fibers, portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination of the foregoing.

[0127] 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).

[0128] The systems and technologies described herein can be implemented in computing systems that include backend components (e.g., as a data server), or computing systems that include middleware components (e.g., an application server), or computing systems that include frontend components (e.g., a user computer with a graphical user interface or web browser through which a user can interact with embodiments 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., a communication network). Examples of communication networks include local area networks (LANs), wide area networks (WANs), and the Internet.

[0129] Computer systems can include clients and servers. Clients and servers are generally located far apart and typically interact via communication networks. Client-server relationships are created by computer programs running on the respective computers and having a client-server relationship with each other. Servers can be cloud servers, servers in distributed systems, or servers incorporating blockchain technology.

[0130] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

Claims

1. A method for sorting metal scrap, characterized in that, include: Acquire multidimensional information about the metal scrap on the conveyor belt that arrives at the detection area, as well as the position and speed information of the metal scrap on the conveyor belt; Based on the multidimensional information, an identification result is obtained; the identification result is used to characterize the material of the metal scrap. Based on the identification results, the location information, and the speed information, when the metal waste arrives at the sorting area, the metal waste is sorted into the corresponding target area.

2. The method according to claim 1, characterized in that, The multidimensional information includes the shape, texture, and spectral reflectance information of the metal scrap; acquiring the multidimensional information of the metal scrap on the conveyor belt arriving at the detection area, as well as the position and speed information of the metal scrap on the conveyor belt, includes: The metal waste in the detection area is illuminated using an array of light sources; Simultaneously trigger visible light cameras, hyperspectral cameras, and depth cameras to acquire shape, texture, and spectral reflectance information of the metal scrap. The speed of the conveyor belt is obtained by an encoder; the encoder is mounted on the conveyor belt.

3. The method according to claim 2, characterized in that, The simultaneous triggering of a visible light camera, a hyperspectral camera, and a depth camera to acquire shape, texture, and spectral reflectance information of the metal scrap includes: A color image of the metal scrap is acquired using a visible light camera; the color image includes texture information and the position information of the metal scrap on the conveyor belt. The hyperspectral camera is used to acquire hyperspectral images, and the spectral reflectance information of each pixel in the hyperspectral image is obtained; the spectral reflectance information is used to determine the material of the metal scrap. The shape information of the metal scrap on the conveyor belt is obtained by using a depth camera; the shape information includes the height, volume and three-dimensional contour of the metal scrap.

4. The method according to claim 2, characterized in that, The step of obtaining the recognition result based on the multidimensional information includes: Based on the multidimensional information, multidimensional feature data is obtained; the multidimensional feature data includes texture feature data, hyperspectral feature data, and shape feature data. The texture feature data, hyperspectral feature data, and shape feature data are input into a pre-constructed multimodal feature fusion classification model to obtain the recognition result.

5. The method according to claim 4, characterized in that, Based on the identification result, the location information, and the speed information, when the metal waste arrives at the sorting area, it is sorted to the corresponding target area, including: The travel distance of the conveyor belt is calculated based on the speed information; Based on the position information of the metal waste on the conveyor belt and the moving distance, it is determined whether the metal waste has reached the sorting area; In response to the metal waste reaching the sorting area, the metal waste is sorted to the corresponding target area based on the identification result.

6. The method according to claim 4, characterized in that, The multimodal feature fusion classification model includes: A self-supervised learning mechanism is used to update the model parameters of the multimodal feature fusion classification model based on the difference between the recognition results and the true labels.

7. The method according to claim 4, characterized in that, The array light source is an LED array light source, and the LED array light source is provided with an elliptical reflector. The conveyor belt moves at a constant speed.

8. A controller, characterized in that, include: Memory, on which executable programs are stored; A processor for executing the executable program in the memory to implement the steps of the method according to any one of claims 1-7.

9. A metal scrap sorting system, characterized in that, include: The controller as described in claim 8; as well as A conveying mechanism and a sorting mechanism, wherein the conveying mechanism and the sorting mechanism are respectively connected to the controller; The conveying mechanism includes: a vibrating feeder, a conveyor belt, and a motor, and the conveyor belt is equipped with an encoder; The vibrating feeder is used to evenly spread the organometallic material onto the conveyor belt. The motor drives the conveyor belt to move at a constant speed, conveying the organometallic material to the detection area; The encoder is used to collect the conveyor position and speed of the conveyor belt and send them to the controller; The sorting mechanism includes: multiple sets of airflow nozzles and a sorting funnel; Multiple sets of airflow nozzles are evenly arranged along the conveyor belt to change the movement trajectory of the metal waste reaching the sorting area through airflow, so that the metal waste moves to the target sorting funnel; A sorting funnel is used to sort and store metal scrap.

10. The system according to claim 9, characterized in that, The sorting mechanism also includes: A robotic arm is used to grip or push scrap metal into a target sorting funnel.

Images