Metal impurity identification method and device based on artificial intelligence, equipment and medium

Abstract

The embodiment of the invention discloses a metal impurity identification method and device based on artificial intelligence, equipment and a medium. A specific embodiment of the method comprises the following steps: controlling an associated shooting device group to shoot to-be-identified scrap metal so as to generate a scrap metal image set; preprocessing each scrap metal image in the scrap metal image set; executing the following metal identification steps: generating a segmentation mask corresponding to each pixel in the preprocessed metal image to obtain a segmentation mask group; carrying out classification processing on each pixel in the preprocessed metal image; based on the generated classification mask group set, the pixel category group set and the pixel confidence group set, performing metal identification processing to obtain metal impurity identification information; carrying out dynamic correction processing on the metal impurity identification information; and controlling the conveying equipment to carry out conveying operation on the to-be-identified scrap metal. According to the implementation mode, the time for identifying the impurities of the scrap metal is shortened, and the waste of metal resources is avoided.

Description

Technical Field

[0001] Embodiments of the present disclosure relate to the field of computer technology, and more particularly, to methods, devices, equipment, and media for identifying metal impurities based on artificial intelligence. Background Art

[0002] In the customs inspection scenario of imported recycled metals, how to determine the impurity components in waste metals is an important research topic. Currently, when determining the impurity components in waste metals, the commonly adopted method is to identify the impurities in the imported waste metals by professional staff.

[0003] However, when using the above method to determine the impurity components in waste metals, the following technical problems often exist: When professional staff identify the impurities in the imported waste metals, it takes a long time to identify the impurities, and the staff may make mistakes during the impurity identification, resulting in waste of metal resources.

[0004] The above information disclosed in this background art section is only used to enhance the understanding of the background of the inventive concept, and thus, it may include information that does not form the prior art known to those of ordinary skill in the art in this country. Summary of the Invention

[0005] This content section of the present disclosure is used to briefly introduce concepts that will be described in detail in the subsequent detailed implementation section. This content section of the present disclosure is not intended to identify the key features or essential features of the claimed technical solution, nor is it intended to limit the scope of the claimed technical solution.

[0006] Some embodiments of the present disclosure propose methods, devices, electronic devices, and computer-readable media for identifying metal impurities based on artificial intelligence to solve one or more of the technical problems mentioned in the above background art section.

[0007] In a first aspect, some embodiments of the present disclosure provide an artificial intelligence-based method for identifying metal impurities. The method includes: in response to detecting waste metal to be identified on a conveying device, controlling an associated group of photographing devices to photograph the waste metal to be identified to generate a set of waste metal images; preprocessing each waste metal image in the set of waste metal images to generate a set of preprocessed metal images; for each preprocessed metal image in the set of preprocessed metal images, performing the following metal identification steps: generating a segmentation mask corresponding to each pixel in the preprocessed metal image to obtain a set of segmentation masks; classifying each pixel in the preprocessed metal image to generate a pixel category and a pixel confidence corresponding to the pixel category to obtain a set of pixel categories and a set of pixel confidences; performing metal identification processing based on the generated set of classification mask sets, set of pixel category sets, and set of pixel confidence sets to obtain metal impurity identification information; performing dynamic correction processing on the metal impurity identification information to generate corrected metal impurity identification information; and based on the corrected metal impurity identification information, controlling the conveying device to perform a transmission operation on the waste metal to be identified.

[0008] In a second aspect, some embodiments of the present disclosure provide an artificial intelligence-based metal impurity identification device. The device includes: a first control unit configured to, in response to detecting waste metal to be identified on a conveying device, control an associated group of photographing devices to photograph the waste metal to be identified to generate a set of waste metal images; a preprocessing unit configured to preprocess each waste metal image in the set of waste metal images to generate a set of preprocessed metal images; an execution unit configured to, for each preprocessed metal image in the set of preprocessed metal images, perform the following metal identification steps: generating a segmentation mask corresponding to each pixel in the preprocessed metal image to obtain a set of segmentation masks; classifying each pixel in the preprocessed metal image to generate a pixel category and a pixel confidence corresponding to the pixel category to obtain a set of pixel categories and a set of pixel confidences; a metal identification unit configured to perform metal identification processing based on the generated set of classification mask sets, set of pixel category sets, and set of pixel confidence sets to obtain metal impurity identification information; a dynamic correction unit configured to perform dynamic correction processing on the metal impurity identification information to generate corrected metal impurity identification information; and a second control unit configured to, based on the corrected metal impurity identification information, control the conveying device to perform a transmission operation on the waste metal to be identified.

[0009] In a third aspect, some embodiments of the present disclosure provide an electronic device, including: one or more processors; a storage device storing one or more programs thereon, and when the one or more programs are executed by the one or more processors, causing the one or more processors to implement the method described in any implementation manner of the first aspect above.

[0010] In a fourth aspect, some embodiments of the present disclosure provide a computer-readable medium having a computer program stored thereon, wherein when the program is executed by a processor, the method described in any implementation manner of the above first aspect is implemented.

[0011] The above various embodiments of the present disclosure have the following beneficial effects: Through the artificial intelligence-based metal impurity identification method of some embodiments of the present disclosure, the time for identifying impurities in scrap metal is reduced, and waste of metal resources is avoided. Specifically, the reasons for the need to spend a long time on impurity identification and waste of metal resources are as follows: When professional staff identify impurities in imported scrap metal, it takes a long time for impurity identification, and errors may occur during impurity identification by the staff, resulting in waste of metal resources. Based on this, in the artificial intelligence-based metal impurity identification method of some embodiments of the present disclosure, first, in response to detecting the presence of scrap metal to be identified on the conveying device, control the associated imaging device group to perform imaging processing on the scrap metal to be identified to generate a scrap metal image set. Thus, multi-dimensional data of the scrap metal can be obtained. Secondly, preprocess each scrap metal image in the scrap metal image set to generate a preprocessed metal image set. Thus, the features of the scrap metal images can be enhanced, making the impurity identification result more accurate. Then, for each preprocessed metal image in the preprocessed metal image set, perform the following metal identification steps: generate a segmentation mask corresponding to each pixel in the preprocessed metal image to obtain a segmentation mask group; classify each pixel in the preprocessed metal image to generate a pixel category and a pixel confidence corresponding to the pixel category to obtain a pixel category group and a pixel confidence group. Thus, each pixel can be classified for metal identification and the corresponding confidence can be generated. After that, based on the generated classification mask group set, pixel category group set, and pixel confidence group set, perform metal identification processing to obtain metal impurity identification information. Thus, metal identification and impurity identification of the scrap metal can be performed. Then, perform dynamic correction processing on the metal impurity identification information to generate corrected metal impurity identification information. Thus, the identification result can be corrected to avoid misidentification. Finally, based on the corrected metal impurity identification information, control the conveying device to perform a transmission operation on the scrap metal to be identified. Thus, the identified scrap metal can be transmitted, reducing the time for identifying impurities in the scrap metal and avoiding waste of metal resources. BRIEF DESCRIPTION OF THE DRAWINGS

[0012] In combination with the accompanying drawings and with reference to the following specific embodiments, the above and other features, advantages and aspects of the various embodiments of the present disclosure will become more apparent. Throughout the drawings, the same or similar reference numerals denote the same or similar elements. It should be understood that the drawings are schematic and the elements and elements are not necessarily drawn to scale.

[0013] Figure 1 is a flowchart of some embodiments of an artificial intelligence-based metal impurity identification method according to the present disclosure; Figure 2 is a schematic structural diagram of some embodiments of an artificial intelligence-based metal impurity identification device according to the present disclosure; Figure 3 is a schematic structural diagram of an electronic device suitable for implementing some embodiments of the present disclosure. Specific Embodiments

[0014] Embodiments of the present disclosure will be described in more detail below with reference to the accompanying drawings. Although some embodiments of the present disclosure are shown in the drawings, it should be understood that the present disclosure can be implemented in various forms and should not be construed as limited to the embodiments set forth herein. On the contrary, these embodiments are provided to more thoroughly and completely understand the present disclosure. It should be understood that the drawings and embodiments of the present disclosure are for illustrative purposes only and are not used to limit the protection scope of the present disclosure.

[0015] In addition, it should be noted that, for the sake of convenience of description, only parts related to the relevant invention are shown in the drawings. Without conflict, the embodiments in the present disclosure and the features in the embodiments can be combined with each other.

[0016] It should be noted that the concepts such as "first" and "second" mentioned in the present disclosure are only used to distinguish different devices, modules or units, and are not used to limit the order of functions performed by these devices, modules or units or their interdependent relationships.

[0017] It should be noted that the modifications of "one" and "multiple" mentioned in the present disclosure are illustrative rather than restrictive. Those skilled in the art should understand that, unless otherwise clearly stated in the context, it should be understood as "one or more".

[0018] The names of the messages or information exchanged between multiple devices in the embodiments of the present disclosure are only for illustrative purposes and are not used to limit the scope of these messages or information.

[0019] Embodiments of the present disclosure will be described in detail below with reference to the drawings and in combination with the embodiments.

[0020] Figure 1Flow 100 of some embodiments of the artificial intelligence-based metal impurity identification method according to the present disclosure is shown. The artificial intelligence-based metal impurity identification method includes the following steps: Step 101, in response to detecting the presence of waste metal to be identified on the conveying device, controlling the associated shooting device group to perform shooting processing on the waste metal to be identified to generate a waste metal image set.

[0021] In some embodiments, the execution subject (such as a server) of the artificial intelligence-based metal impurity identification method can, in response to detecting the presence of waste metal to be identified on the conveying device, control the associated shooting device group to perform shooting processing on the above-mentioned waste metal to be identified to generate a waste metal image set. Among them, the above-mentioned conveying device can be a device with an item conveying function. For example, the above-mentioned conveying device can be a conveyor belt. The shooting devices in the above-mentioned associated shooting device group can be shooting devices with a shooting function that are connected to the above-mentioned execution subject by wire or wirelessly. As an example, the above-mentioned shooting device can be a high-resolution camera or a multispectral sensor. In practice, a weight sensor or an infrared sensor can be used to detect whether there is waste metal to be identified on the above-mentioned conveying device.

[0022] Optionally, before step 101, the following steps are further included: The first step is to perform parameter initialization processing on each shooting device included in the above-mentioned associated shooting device group.

[0023] In some embodiments, the above-mentioned execution subject can perform parameter initialization processing on each shooting device included in the above-mentioned associated shooting device group. The above-mentioned parameter initialization can be to restore the parameters of each shooting device to the initial parameters.

[0024] The second step is to perform time consistency processing on the time parameters corresponding to the above-mentioned each shooting device to obtain a shooting device group with consistent time.

[0025] In some embodiments, the above-mentioned execution subject can perform time consistency processing on the time parameters corresponding to the above-mentioned each shooting device to obtain a shooting device group with consistent time.

[0026] Step 102, preprocess each waste metal image in the waste metal image set to generate a preprocessed metal image set.

[0027] In some embodiments, the above-mentioned execution subject can preprocess each waste metal image in the above-mentioned waste metal image set to generate a preprocessed metal image set.

[0028] In practice, the following steps can be used to preprocess each waste metal image in the above-mentioned waste metal image set: First step, perform denoising and enhancement processing on each waste metal image in the above waste metal image set to generate an enhanced waste metal image set. In practice, median filtering can be used to denoise the waste metal image, and histogram equalization can be used to enhance the contrast of the waste metal image.

[0029] Second step, perform feature point recognition processing on each waste metal image in the above waste metal image set to generate an image set after recognition. Here, algorithms such as SIFT or ORB can be used to extract feature points such as corners and edges of each image.

[0030] Third step, based on the feature points included in each recognized image in the above image set after recognition, perform registration processing on each of the above recognized images to generate a registered metal image, and obtain a registered metal image set. In practice, affine transformation can be used to align multi-angle images to a unified coordinate system for image alignment and registration.

[0031] Fourth step, for each registered metal image in the above registered metal image set, perform reflection correction processing on the above registered metal image to generate a corrected metal image as a preprocessed metal image. In practice, a lighting model can be established to separate specular reflection and diffuse reflection components and suppress metal specular interference for reflection correction.

[0032] Step 103, for each preprocessed metal image in the preprocessed metal image set, perform the following metal recognition steps: Step 1031, generate a segmentation mask corresponding to each pixel in the preprocessed metal image to obtain a group of segmentation masks.

[0033] In some embodiments, the above execution subject can generate a segmentation mask corresponding to each pixel in the preprocessed metal image to obtain a group of segmentation masks. In practice, a semantic segmentation model can be used to determine the mask label corresponding to each pixel in the preprocessed metal image as the segmentation mask. The above mask label can represent the material corresponding to the above pixel. As an example, the above segmentation mask can be "red copper", "brass", "iron", "lead impurity", etc.

[0034] Step 1032, perform classification processing on each pixel in the preprocessed metal image to generate a pixel category and the pixel confidence corresponding to the pixel category, obtaining a group of pixel categories and a group of pixel confidences.

[0035] In some embodiments, the above execution subject can perform classification processing on each pixel in the above preprocessed metal image to generate a pixel category and the pixel confidence corresponding to the above pixel category, obtaining a group of pixel categories and a group of pixel confidences.

[0036] In practice, each pixel in the preprocessed metal image can be classified through the following steps to generate pixel categories and pixel confidence levels corresponding to the pixel categories: First, perform channel splitting on the above-mentioned preprocessed metal image to generate a set of split images. In practice, it can be split according to R, G, and B in RGB to obtain three single-channel images.

[0037] Second, extract the image features of each split image in the above-mentioned set of split images to obtain a set of image feature maps. Here, multi-scale feature maps of the split images can be extracted through a pre-trained convolutional neural network.

[0038] Third, perform weighted fusion on each image feature map in the above-mentioned image feature maps to generate a fused feature map. In practice, the weight of each feature map can be determined through an attention mechanism, and a fused feature map can be generated through weighted fusion.

[0039] Fourth, based on the above-mentioned fused feature map, determine the pixel category and pixel confidence level corresponding to each pixel in the above-mentioned preprocessed metal image.

[0040] Step 104: Based on the generated set of classification masks, set of pixel categories, and set of pixel confidence levels, perform metal recognition processing to obtain metal impurity recognition information.

[0041] In some embodiments, the above-mentioned execution entity can perform metal recognition processing based on the generated set of classification masks, set of pixel categories, and set of pixel confidence levels to obtain metal impurity recognition information.

[0042] In the process of adopting technical solutions to solve the above technical problems, the following technical problems often arise: In the process of metal recognition, the colors presented by different types of metals are similar, resulting in incorrect recognition during metal recognition, misidentifying recycled metal as impurity metal, thus causing waste of metal resources. Facing the above technical problems and combining the existing technical status, the following solutions can be decided to be adopted.

[0043] In some optional implementation manners of some embodiments, the above-mentioned execution entity can perform metal recognition processing through the following steps to obtain metal impurity recognition information: First, for each preprocessed metal image in the above-mentioned set of preprocessed metal images, perform the following recognition steps: The first recognition step: Generate spatial constraint information corresponding to the above-mentioned preprocessed metal image based on the corresponding set of classification masks. In practice, each classification mask at the edge can be selected from the set of classification masks, and spatial constraint information can be generated according to the positions of the respective pixels corresponding to each classification mask in the preprocessed metal image.

[0044] A second recognition step, based on the above spatial constraint information, crops the preprocessed metal image to generate a cropped preprocessed metal image.

[0045] A third recognition step, according to each pixel category corresponding to the cropped preprocessed metal image, maps the cropped preprocessed metal image to generate at least one channel feature map.

[0046] A fourth recognition step, normalizes each pixel confidence corresponding to the cropped preprocessed metal image to generate a normalized confidence group.

[0047] A fifth recognition step, based on the above normalized confidence group, generates a weighted feature map corresponding to the cropped preprocessed metal image.

[0048] A sixth recognition step, obtains preset metal recognition rule information. Among them, the above preset metal recognition rule information can be the order information for recognizing metals set in advance. As an example, the above preset metal recognition rule information can be red copper > brass > iron > lead impurity.

[0049] A seventh recognition step, based on the above preset metal recognition rule information, inputs the above at least one channel feature map and the above weighted feature map into a preset classifier chain to perform metal recognition processing on the cropped preprocessed metal image, and obtains a metal recognition result.

[0050] An eighth recognition step, according to the above metal recognition result, determines impurity region information corresponding to metal impurities.

[0051] The second step, according to the generated metal recognition results and each impurity region information, generates metal impurity recognition information. In practice, regions representing impurities can be screened out from the above metal recognition results and the above impurity region information, and the area ratio of the impurity region is determined. The above area ratio can be the ratio of the number of impurity pixels to the total number of metal pixels.

[0052] The relevant content from the first step to the second step, as an inventive point of the embodiments of the present disclosure, in combination with the following step "Step 106", solves the technical problem: "During the process of metal identification, the colors presented by different types of metals are similar, resulting in incorrect identification during metal identification, misidentifying recycled metal as impurity metal, thus causing waste of metal resources." The reasons for the waste of metal resources are as follows: During the process of metal identification, the colors presented by different types of metals are similar, resulting in incorrect identification during metal identification, misidentifying recycled metal as impurity metal, thus causing waste of metal resources. If the above factors are solved, the effect of avoiding waste of metal resources can be achieved. To achieve this effect, first, for each preprocessed metal image in the above preprocessed metal image set, perform the following identification steps: Generate spatial constraint information corresponding to the above preprocessed metal image based on the corresponding classification mask group. Thus, the range of impurity metal in the preprocessed metal image can be determined. Second, based on the above spatial constraint information, perform a cropping process on the above preprocessed metal image to generate a cropped preprocessed metal image. Thus, the impurity metal area can be cropped to reduce the identification range. Third, according to the respective pixel categories corresponding to the above cropped preprocessed metal image, map the above cropped preprocessed metal image to generate at least one channel feature map. Thus, the pixel categories can be mapped to channel feature maps. Fourth, perform a normalization process on the respective pixel confidences corresponding to the above cropped preprocessed metal image to generate a normalized confidence group; based on the above normalized confidence group, generate a weighted feature map corresponding to the above cropped preprocessed metal image. Thus, the confidences can be weighted and fused. Fifth, obtain preset metal identification rule information; based on the above preset metal identification rule information, input the above at least one channel feature map and the above weighted feature map into a preset classifier chain to perform metal identification processing on the above cropped preprocessed metal image to obtain a metal identification result. Thus, the metal can be identified to obtain an identification result. Sixth, according to the above metal identification result, determine the impurity area information corresponding to the metal impurity; according to the generated respective metal identification results and respective impurity area information, generate metal impurity identification information. Thus, the identified metal information and impurity information can be determined. In combination with the following step "Step 106", based on the above corrected metal impurity identification information, control the above conveying device to perform a transmission operation on the waste metal to be identified. Thus, the situation of incorrect metal identification can be avoided, thereby avoiding misidentifying recycled metal as impurity metal and avoiding waste of metal resources.

[0053] Step 105: Perform a dynamic correction process on the metal impurity identification information to generate corrected metal impurity identification information.

[0054] In some embodiments, the above-mentioned execution entity may perform dynamic correction processing on the above-mentioned metal impurity identification information to generate corrected metal impurity identification information.

[0055] In the process of adopting technical solutions to solve the above-mentioned technical problems, the following technical problems often arise: Due to the different light intensities and shooting angles when shooting the passing metals, the captured metal images may not conform to the actual situation, resulting in incorrect metal identification. As a result, metals with more impurities may be allowed to pass or metals with fewer impurities may be prohibited from passing through customs, leading to a waste of metal resources. Facing the above-mentioned technical problems and combining the existing technical status, the following solutions can be decided to be adopted.

[0056] In practice, the dynamic correction processing of the metal impurity identification information can be carried out through the following steps to generate corrected metal impurity identification information: First step, in response to the impurity confidence level corresponding to the above-mentioned metal impurity identification information satisfying the preset confidence level condition, select a shooting device that meets the preset shooting conditions from the above-mentioned set of shooting devices as the target shooting device. Among them, the above-mentioned preset confidence level condition may be that the impurity awareness is less than or equal to the preset confidence level threshold. As an example, the above-mentioned preset confidence level threshold may be 80%. The above-mentioned preset shooting condition may be the shooting device closest to the waste metal to be identified.

[0057] Second step, control the above-mentioned conveying device to perform transmission processing on the above-mentioned waste metal to be identified according to the preset transmission distance. Among them, the above-mentioned preset transmission distance may be the distance preset for transmitting the above-mentioned waste metal to be identified.

[0058] Third step, based on the spatial alignment algorithm, control the above-mentioned target shooting device to perform multi-frame shooting on the above-mentioned waste metal to be identified to obtain a waste metal frame sequence. Among them, the above-mentioned spatial alignment algorithm may be the flow method. Here, the metal position can be tracked through the spatial alignment algorithm to ensure shooting at the same position. The above-mentioned multi-frame shooting may be carried out according to the preset shooting quantity and shooting interval.

[0059] Fourth step, perform impurity identification processing on each waste metal frame in the above-mentioned waste metal frame sequence to generate multiple impurity identification results, and obtain an impurity identification result set.

[0060] Fifth step, in response to the above-mentioned impurity identification result set not satisfying the preset condition, generate correction request information corresponding to the above-mentioned waste metal to be identified, and send the above-mentioned correction information to the above-mentioned customs terminal for display. Among them, the above-mentioned preset condition may be that the number of impurity identification results indicating impurities as the above-mentioned metal impurity identification information in the above-mentioned impurity identification result set is greater than or equal to the preset number.

[0061] Step 6: In response to receiving the calibration information for the to-be-identified scrap metal sent by the above-mentioned customs terminal, update the metal impurity identification information to generate updated metal impurity identification information as the calibrated metal impurity identification information.

[0062] The relevant content of the above first to sixth steps, as an inventive point of the embodiments of the present disclosure, in combination with the following step "step 106", solves the technical problem: "Due to different light intensities and shooting angles for shooting the passing metal, the captured metal image may not conform to the actual situation, resulting in incorrect metal recognition, and further allowing metals with more impurities to pass or prohibiting metals with fewer impurities from passing through customs, leading to waste of metal resources." The reasons for the waste of metal resources are as follows: Due to different light intensities and shooting angles for shooting the passing metal, the captured metal image may not conform to the actual situation, resulting in incorrect metal recognition, and further allowing metals with more impurities to pass or prohibiting metals with fewer impurities from passing through customs, leading to waste of metal resources. If the above factors are solved, the effect of avoiding waste of metal resources can be achieved. To achieve this effect, the present disclosure first, in response to the impurity confidence corresponding to the above metal impurity recognition information satisfying the preset confidence condition, selects a shooting device that meets the preset shooting conditions from the above shooting device set as the target shooting device. Thus, when the confidence level does not reach the threshold, the shooting device closest to the metal to be recognized can be selected. Second, control the above conveying device to perform a transmission process on the above waste metal to be recognized according to a preset transmission distance; based on the spatial alignment algorithm, control the above target shooting device to perform multi-frame shooting on the above waste metal to be recognized to obtain a waste metal frame sequence. Thus, the waste metal to be recognized can be tracked and shot, and the metal can be shot from multiple angles. Third, perform impurity recognition processing on each waste metal frame in the above waste metal frame sequence to generate a plurality of impurity recognition results to obtain an impurity recognition result set. Thus, it can be determined whether there are impurities in each captured image. Fourth, in response to the above impurity recognition result set not meeting the preset conditions, generate a correction request information corresponding to the above waste metal to be recognized, and send the above correction request information to the above customs terminal for display. Thus, when the number of images showing impurities is less than the preset threshold, the correction request information can be sent to the customs terminal. Fifth, in response to receiving the correction information sent by the above customs terminal for the above waste metal to be recognized, perform an update process on the above metal impurity recognition information to generate updated metal impurity recognition information as the corrected metal impurity recognition information. Thus, the information on the impurities included in the waste metal to be recognized can be corrected by receiving the correction information sent by the customs terminal. In combination with the following step "step 106", based on the above corrected metal impurity recognition information, control the above conveying device to perform a transmission operation on the above waste metal to be recognized. Thus, by correcting the recognized impurity information, the situation of allowing metals with more impurities to pass or prohibiting metals with fewer impurities from passing through customs is avoided, thereby avoiding waste of metal resources.

[0063] Step 106, based on the corrected metal impurity recognition information, control the conveying device to perform a transmission operation on the waste metal to be recognized.

[0064] In some embodiments, the above-mentioned execution entity may, based on the above-mentioned corrected metal impurity identification information, control the above-mentioned conveying device to perform a transmission operation on the above-mentioned waste metal to be identified.

[0065] In some optional implementation manners of some embodiments, the above-mentioned execution entity may control the conveying device to perform a transmission operation on the waste metal to be identified through the following steps: First step, obtain the customs clearance rule information corresponding to the customs terminal. Among them, the above-mentioned customs clearance rule information may represent the relationship between the impurities in the waste metal to be identified and customs clearance. As an example, the above-mentioned customs clearance rule information may be that when the impurities are greater than or equal to 20%, the waste metal to be identified is prohibited from customs clearance.

[0066] Second step, determine whether the above-mentioned corrected metal impurity identification information meets the above-mentioned customs clearance rule information.

[0067] Third step, in response to the above-mentioned corrected metal impurity identification information meeting the above-mentioned customs clearance rule information, generate a customs clearance mark corresponding to the above-mentioned waste metal to be identified.

[0068] Fourth step, transfer the above-mentioned waste metal to be identified to the metal storage area corresponding to the above-mentioned customs clearance mark. Among them, the above-mentioned metal storage area may be a pre-set storage area for metals representing customs clearance.

[0069] Fifth step, in response to the above-mentioned corrected metal impurity identification information not meeting the above-mentioned customs clearance rule information, generate an alarm message corresponding to the above-mentioned waste metal to be identified, and send the above-mentioned alarm message to the above-mentioned customs terminal for display.

[0070] Sixth step, control the above-mentioned conveying device to stop transmitting the above-mentioned waste metal to be identified.

[0071] The above-mentioned various embodiments of the present disclosure have the following beneficial effects: Through the artificial intelligence-based metal impurity identification method of some embodiments of the present disclosure, the time for identifying impurities in waste metal is reduced, and the waste of metal resources is avoided. Specifically, the reasons for the long time required for impurity identification and the waste of metal resources are as follows: When professional staff identify impurities in imported waste metal, it takes a long time for impurity identification, and errors may occur during the impurity identification by the staff, resulting in the waste of metal resources. Based on this, in the artificial intelligence-based metal impurity identification method of some embodiments of the present disclosure, first, in response to detecting the presence of waste metal to be identified on the conveying device, control the associated imaging device group to perform imaging processing on the waste metal to be identified to generate a waste metal image set. Thus, multi-dimensional data of the waste metal can be obtained. Secondly, preprocess each waste metal image in the waste metal image set to generate a preprocessed metal image set. Thus, the features of the waste metal image can be enhanced, making the result of impurity identification more accurate. Then, for each preprocessed metal image in the preprocessed metal image set, perform the following metal identification steps: generate a segmentation mask corresponding to each pixel in the preprocessed metal image to obtain a segmentation mask group; classify each pixel in the preprocessed metal image to generate a pixel category and a pixel confidence corresponding to the pixel category to obtain a pixel category group and a pixel confidence group. Thus, each pixel can be classified for metal identification and the corresponding confidence can be generated. After that, based on the generated classification mask group set, pixel category group set, and pixel confidence group set, perform metal identification processing to obtain metal impurity identification information. Thus, metal identification and impurity identification of the waste metal can be performed. Then, perform dynamic correction processing on the metal impurity identification information to generate corrected metal impurity identification information. Thus, the identification result can be corrected to avoid misidentification. Finally, based on the corrected metal impurity identification information, control the conveying device to perform a transmission operation on the waste metal to be identified. Thus, the identified waste metal can be transmitted, reducing the time for identifying impurities in the waste metal and avoiding the waste of metal resources.

[0072] Further referring to Figure 2 , as an implementation of the methods shown in the above figures, the present disclosure provides some embodiments of an artificial intelligence-based metal impurity identification device. These device embodiments correspond to Figure 1 the method embodiments shown, and the artificial intelligence-based metal impurity identification device can be specifically applied to various electronic devices.

[0073] As Figure 2As shown, the artificial intelligence-based metal impurity identification device 200 of some embodiments includes: a first control unit 201, a preprocessing unit 202, an execution unit 203, a metal identification unit 204, a dynamic correction unit 205, and a second control unit 206. Among them, the first control unit 201 is configured to control the associated imaging device group to perform imaging processing on the waste metal to be identified in response to detecting the presence of waste metal to be identified on the conveying device, so as to generate a waste metal image set; the preprocessing unit 202 is configured to perform preprocessing on each waste metal image in the waste metal image set to generate a preprocessed metal image set; the execution unit 203 is configured to perform the following metal identification steps for each preprocessed metal image in the preprocessed metal image set: generate a segmentation mask corresponding to each pixel in the preprocessed metal image to obtain a segmentation mask group; perform classification processing on each pixel in the preprocessed metal image to generate a pixel category and a pixel confidence corresponding to the pixel category to obtain a pixel category group and a pixel confidence group; the metal identification unit 204 is configured to perform metal identification processing based on the generated classification mask group set, pixel category group set, and pixel confidence group set to obtain metal impurity identification information; the dynamic correction unit 205 is configured to perform dynamic correction processing on the metal impurity identification information to generate corrected metal impurity identification information; the second control unit 206 is configured to control the conveying device to perform a conveying operation on the waste metal to be identified based on the corrected metal impurity identification information.

[0074] It can be understood that the units described in the artificial intelligence-based metal impurity identification device 200 correspond to the respective steps in the method described in the reference Figure 1 Therefore, the operations, features, and beneficial effects described above for the method also apply to the artificial intelligence-based metal impurity identification device 200 and the units included therein, and will not be elaborated here.

[0075] Next, refer to Figure 3 , which shows a schematic structural diagram of an electronic device 300 suitable for implementing some embodiments of the present disclosure. The electronic devices in some embodiments of the present disclosure may include, but are not limited to, mobile terminals such as mobile phones, laptop computers, digital broadcast receivers, PDAs (Personal Digital Assistants), PADs (Tablet Computers), PMPs (Portable Multimedia Players), vehicle terminals (such as vehicle navigation terminals), etc., and fixed terminals such as digital TVs, desktop computers, etc. Figure 3 The electronic device shown is only an example and should not impose any limitations on the functions and usage scopes of the embodiments of the present disclosure.

[0076] As Figure 3As shown, the electronic device 300 may include a processing device 301 (such as a central processing unit, a graphics processing unit, etc.), which may perform various appropriate actions and processes according to a program stored in the read-only memory (ROM) 302 or a program loaded from the storage device 308 into the random access memory (RAM) 303. In the RAM 303, various programs and data required for the operation of the electronic device 300 are also stored. The processing device 301, the ROM 302, and the RAM 303 are connected to each other through a bus 304. The input / output (I / O) interface 305 is also connected to the bus 304.

[0077] Generally, the following devices may be connected to the I / O interface 305: an input device 306 including, for example, a touch screen, a touchpad, a keyboard, a mouse, a camera, a microphone, an accelerometer, a gyroscope, etc.; an output device 307 including, for example, a liquid crystal display (LCD), a speaker, a vibrator, etc.; a storage device 308 including, for example, a magnetic tape, a hard disk, etc.; and a communication device 309. The communication device 309 may allow the electronic device 300 to communicate with other devices wirelessly or wireline to exchange data. Although Figure 3 the electronic device 300 with various devices is shown, it should be understood that it is not required to implement or have all the shown devices. More or fewer devices may be implemented or had alternatively. Figure 3 Each block shown in the figure may represent one device or, as required, multiple devices.

[0078] In particular, according to some embodiments of the present disclosure, the processes described above with reference to the flowcharts may be implemented as computer software programs. For example, some embodiments of the present disclosure include a computer program product, which includes a computer program carried on a computer-readable medium, and the computer program contains program codes for performing the methods shown in the flowcharts. In such some embodiments, the computer program may be downloaded and installed from the network through the communication device 309, or installed from the storage device 308, or installed from the ROM 302. When the computer program is executed by the processing device 301, the above functions defined in the methods of some embodiments of the present disclosure are performed.

[0079] It should be noted that the computer-readable media described in some embodiments of the present disclosure may be a computer-readable signal medium, a computer-readable storage medium, or any combination of the two. A computer-readable storage medium may be, for example, but not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any combination of the above. More specific examples of the computer-readable storage medium may include, but are not limited to: an electrical connection having one or more wires, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the above. In some embodiments of the present disclosure, the computer-readable storage medium may be any tangible medium that contains or stores a program, and the program may be used by or in conjunction with an instruction execution system, apparatus, or device. In some embodiments of the present disclosure, the computer-readable signal medium may include a data signal propagated in a baseband or as part of a carrier wave, which carries computer-readable program code. Such a propagated data signal may take various forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination of the above. The computer-readable signal medium may also be any computer-readable medium other than the computer-readable storage medium, and the computer-readable signal medium may send, propagate, or transmit a program for use by or in conjunction with an instruction execution system, apparatus, or device. The program code contained on the computer-readable medium may be transmitted by any appropriate medium, including but not limited to: wires, optical cables, RF (radio frequency), etc., or any suitable combination of the above.

[0080] In some embodiments, the client and the server may communicate using any currently known or future-developed network protocol such as HTTP (HyperText Transfer Protocol), and may be interconnected with digital data communication in any form or medium (e.g., a communication network). Examples of communication networks include local area networks ("LANs"), wide area networks ("WANs"), the Internet (e.g., the Internet), and end-to-end networks (e.g., ad hoc end-to-end networks), as well as any currently known or future-developed networks.

[0081] The above computer-readable medium may be included in the above electronic device; or it may exist separately and not be assembled into the electronic device. The above computer-readable medium carries one or more programs, and when the above one or more programs are executed by the electronic device, the electronic device is caused to: in response to detecting the waste metal to be identified on the conveying device, control the associated imaging device group to perform imaging processing on the waste metal to be identified to generate a waste metal image set. Preprocess each waste metal image in the waste metal image set to generate a preprocessed metal image set. For each preprocessed metal image in the preprocessed metal image set, perform the following metal identification steps: generate a segmentation mask corresponding to each pixel in the preprocessed metal image to obtain a segmentation mask group; classify each pixel in the preprocessed metal image to generate a pixel category and a pixel confidence corresponding to the pixel category to obtain a pixel category group and a pixel confidence group. Based on the generated classification mask group set, pixel category group set, and pixel confidence group set, perform metal identification processing to obtain metal impurity identification information. Perform dynamic correction processing on the metal impurity identification information to generate corrected metal impurity identification information. Based on the corrected metal impurity identification information, control the conveying device to perform a conveying operation on the waste metal to be identified.

[0082] Computer program code for performing the operations of some embodiments of the present disclosure may be written in one or more programming languages or combinations thereof. The programming languages include object-oriented programming languages such as Java, Smalltalk, C++, and also include conventional procedural programming languages such as the "C" language or similar programming languages. The program code may execute entirely on the user's computer, partially on the user's computer, execute as a stand-alone software package, execute partially on the user's computer and partially on a remote computer, or execute entirely on a remote computer or server. In the case of a remote computer, the remote computer may be connected to the user's computer through any type of network, including a local area network (LAN) or a wide area network (WAN), or may be connected to an external computer (e.g., by using an Internet service provider to connect through the Internet).

[0083] The flowcharts and block diagrams in the accompanying drawings illustrate the possible architectures, functions, and operations of systems, methods, and computer program products according to various embodiments of the present disclosure. In this regard, each block in the flowchart or block diagram may represent a module, a segment of a program, or a part of code that contains one or more executable instructions for implementing the specified logical function. It should also be noted that in some alternative implementations, the functions marked in the blocks may occur in a different order than that marked in the accompanying drawings. For example, two consecutive blocks shown may actually be executed substantially in parallel, and they may sometimes be executed in the reverse order, depending on the functions involved. It should also be noted that each block in the block diagram and / or flowchart, as well as combinations of blocks in the block diagram and / or flowchart, can be implemented by a dedicated hardware-based system that performs the specified functions or operations, or can be implemented by a combination of dedicated hardware and computer instructions.

[0084] The units described in some embodiments of the present disclosure can be implemented in software or in hardware. The described units can also be provided in a processor. For example, it can be described as: a processor includes a first control unit, a preprocessing unit, an execution unit, an input unit, a dynamic correction unit, and a second control unit. Among them, the names of these units do not constitute a limitation on the unit itself in some cases. For example, the first control unit can also be described as "a unit that controls the associated imaging device group to image the waste metal to be identified on the conveying device to generate a set of waste metal images in response to detecting the presence of waste metal to be identified on the conveying device".

[0085] The functions described above can be performed, at least in part, by one or more hardware logic components. For example, without limitation, exemplary types of hardware logic components that can be used include: Field Programmable Gate Array (FPGA), Application Specific Integrated Circuit (ASIC), Application Specific Standard Product (ASSP), System on Chip (SOC), Complex Programmable Logic Device (CPLD), and so on.

[0086] The above description is only some preferred embodiments of the present disclosure and an explanation of the applied technical principles. Those skilled in the art should understand that the scope of the invention involved in the embodiments of the present disclosure is not limited to the technical solutions formed by the specific combination of the above technical features, and should also cover other technical solutions formed by any combination of the above technical features or their equivalent features without departing from the above inventive concept. For example, technical solutions formed by mutually replacing the above features with (but not limited to) technical features having similar functions disclosed in the embodiments of the present disclosure.

Claims

1. An artificial intelligence-based method for identifying metal impurities, comprising: In response to detecting scrap metal to be identified on a conveying device, controlling an associated set of imaging devices to perform imaging processing on the scrap metal to be identified to generate a set of scrap metal images; Preprocessing each scrap metal image in the set of scrap metal images to generate a set of preprocessed metal images; For each preprocessed metal image in the set of preprocessed metal images, performing the following metal identification steps: Generating a segmentation mask corresponding to each pixel in the preprocessed metal image to obtain a set of segmentation masks; Classifying each pixel in the preprocessed metal image to generate a pixel category and a pixel confidence corresponding to the pixel category, obtaining a set of pixel categories and a set of pixel confidences; Based on the generated set of classification mask sets, set of pixel category sets, and set of pixel confidence sets, performing metal identification processing to obtain metal impurity identification information; Performing dynamic correction processing on the metal impurity identification information to generate corrected metal impurity identification information; Based on the corrected metal impurity identification information, controlling the conveying device to perform a conveying operation on the scrap metal to be identified.

2. The method according to claim 1, wherein Before the step of, in response to detecting scrap metal to be identified on a conveying device, controlling an associated set of imaging devices to perform imaging processing on the scrap metal to be identified to generate a set of scrap metal images, the method further includes: Performing parameter initialization processing on each imaging device included in the associated set of imaging devices; Performing time consistency processing on the time parameters corresponding to each imaging device to obtain a set of imaging devices with consistent time.

3. The method according to claim 1, wherein, The step of preprocessing each scrap metal image in the set of scrap metal images to generate a set of preprocessed metal images includes: Performing denoising and enhancement processing on each scrap metal image in the set of scrap metal images to generate an enhanced set of scrap metal images; Performing feature point identification processing on each scrap metal image in the set of scrap metal images to generate a set of identified images; Based on the respective feature points included in each identified image in the set of identified images, performing registration processing on each identified image to generate a registered metal image, obtaining a set of registered metal images; For each registered metal image in the set of registered metal images, performing reflection correction processing on the registered metal image to generate a corrected metal image as the preprocessed metal image.

4. The method according to claim 1, wherein The step of classifying each pixel in the preprocessed metal image to generate a pixel category and a pixel confidence corresponding to the pixel category, obtaining a set of pixel categories and a set of pixel confidences, includes: Performing channel splitting processing on the preprocessed metal image to generate a set of split images; Extracting image features of each split image in the set of split images to obtain a set of image feature maps; Performing weighted fusion processing on each image feature map in the set of image feature maps to generate a fused feature map; Based on the fused feature map, determining the pixel category and pixel confidence corresponding to each pixel in the preprocessed metal image.

5. The method according to claim 1, wherein, Based on the corrected metal impurity identification information, controlling the transfer device to perform a transfer operation on the waste metal to be identified, including: Obtaining customs clearance rule information corresponding to the customs terminal; Determining whether the corrected metal impurity identification information meets the customs clearance rule information; In response to the corrected metal impurity identification information meeting the customs clearance rule information, generating a customs clearance mark corresponding to the waste metal to be identified; Transferring the waste metal to be identified to the metal storage area corresponding to the customs clearance mark; In response to the corrected metal impurity identification information not meeting the customs clearance rule information, generating an alarm message corresponding to the waste metal to be identified, and sending the alarm message to the customs terminal for display; Controlling the transfer device to stop transferring the waste metal to be identified.

6. An artificial intelligence-based metal impurity identification device, including: A first control unit, configured to control an associated set of imaging devices to perform imaging processing on the waste metal to be identified in response to detecting the presence of waste metal to be identified on the transfer device, so as to generate a waste metal image set; A preprocessing unit, configured to preprocess each waste metal image in the waste metal image set to generate a preprocessed metal image set; An execution unit, configured to, for each preprocessed metal image in the preprocessed metal image set, perform the following metal identification steps: generating a segmentation mask corresponding to each pixel in the preprocessed metal image to obtain a set of segmentation masks; classifying each pixel in the preprocessed metal image to generate a pixel category and a pixel confidence corresponding to the pixel category, to obtain a set of pixel categories and a set of pixel confidences; A metal identification unit, configured to perform metal identification processing based on the generated set of classification mask sets, set of pixel category sets, and set of pixel confidence sets to obtain metal impurity identification information; A dynamic correction unit, configured to perform dynamic correction processing on the metal impurity identification information to generate corrected metal impurity identification information; A second control unit, configured to control the transfer device to perform a transfer operation on the waste metal to be identified based on the corrected metal impurity identification information.

7. An electronic device, including: One or more processors; A storage device on which one or more programs are stored; When the one or more programs are executed by the one or more processors, the one or more processors implement the method according to any one of claims 1 to 5.

8. A computer-readable medium having a computer program stored thereon, wherein, The program, when executed by the processor, implements the method according to any one of claims 1 to 5.

Images