Rare metal ore sorting method, device and equipment and storage medium

Abstract

The invention discloses a rare metal ore sorting method, device and equipment and a storage medium, and relates to the technical field of ore sorting, the rare metal ore sorting method comprises the steps that an ore picture is obtained, and ore corresponding to the ore picture is sorted through a preset ore sorting model, the ore sorting model is obtained through training on the basis of rare metal ore pictures with different lithofacies, sorting of the rare metal ore is conducted on the basis of the lithofacies characteristics of the ore, and for the rare metal ore with the complex lithofacies characteristics, the ore sorting model is obtained through training on the basis of the rare metal ore pictures with different lithofacies; through the ore separation model obtained through training, when the separation system carries out ore separation, adjustment can be carried out according to the lithofacies characteristics of the ore, the detection accuracy of the rare metal ore is improved, and therefore the metal recovery rate of rare metal ore separation is improved.

Description

Technical Field

[0001] This application relates to the technical field of ore sorting, and particularly to a method, device, equipment and storage medium for sorting rare metal ores. Background Art

[0002] Ore sorting is the process of separating symbiotic ores, and ore sorting can effectively improve the utilization efficiency of mineral resources. Currently, the methods for sorting rare metal ores usually rely on the physical and chemical properties of the ores for sorting. However, due to the complex petrographic characteristics of rare metal ores, the sorting of rare metal ores by current methods results in low metal recovery rates. Summary of the Invention

[0003] The main purpose of this application is to provide a method, device, equipment and storage medium for sorting rare metal ores, aiming to solve the technical problem of low metal recovery rate in the sorting of rare metal ores.

[0004] To achieve the above purpose, this application proposes a method for sorting rare metal ores, and the method includes:

[0005] Obtain an ore picture;

[0006] Sort the ore corresponding to the ore picture through a preset ore sorting model, where the ore sorting model is trained based on rare metal ore sample pictures with different petrographic characteristics, so as to sort rare metal ores based on the petrographic characteristics of the ores.

[0007] In one embodiment, before the step of obtaining the ore picture, the method further includes:

[0008] Obtain a first ore data set, where the first ore data set is composed of ore sample pictures corresponding to ores with a radius less than a preset threshold;

[0009] Detect the ore sample pictures in the first ore data set based on a pre-trained model to obtain a first detection result, where the first detection result includes whether the corresponding ore is a rare metal ore or the corresponding ore is not a rare metal ore;

[0010] Based on the first detection result, adjust the preset parameters to be updated in the pre-trained model to obtain an ore sorting model, where the parameters other than the parameters to be updated are frozen during the training process of the ore sorting model.

[0011] In one embodiment, the ore sorting model has a channel attention module and a spatial attention module, and the step of sorting the ore corresponding to the ore picture through the preset ore sorting model includes:

[0012] Feature extraction is performed on the ore picture through the channel attention module and the spatial attention module to obtain the petrographic features;

[0013] Based on the petrographic features and the detection layer of the ore sorting model, detection is performed to obtain a target detection result, where the target detection result includes that the corresponding ore is a rare metal ore or the corresponding ore is not a rare metal ore;

[0014] Based on the target detection result, sorting of rare metal ores is performed.

[0015] In one embodiment, the step of performing feature extraction on the ore picture through the channel attention module and the spatial attention module to obtain the petrographic features includes:

[0016] Based on the channel attention module, channel feature extraction is performed on the ore picture to obtain a channel feature map;

[0017] Based on the spatial attention module, max pooling and average pooling are performed on the channel feature map to obtain a max pooling feature and an average pooling feature;

[0018] The max pooling feature and the average pooling feature are concatenated to obtain a concatenated feature;

[0019] Convolution operation is performed on the concatenated feature to obtain a spatial weight of the feature map;

[0020] The spatial weight of the feature map and the channel feature map are subjected to dot product to obtain petrographic features.

[0021] In one embodiment, after the step of obtaining the ore picture, the following is further included:

[0022] The ore picture is preprocessed to obtain a preprocessed picture;

[0023] Multiple preprocessed pictures with a preset splicing quantity are spliced to obtain a spliced picture;

[0024] The ore corresponding to the spliced picture is sorted through a preset ore sorting model.

[0025] In one embodiment, before the step of obtaining the ore picture, the following is further included:

[0026] A second ore data set is obtained, where the second ore data set is composed of ore pictures corresponding to ores with different grades, and among the ores with different grades, the number of low-grade ores is higher than a preset quantity threshold, and the low-grade ore is an ore with a target metal content lower than a preset content threshold;

[0027] Detect the ore sample pictures in the second ore dataset based on the pre-trained model to obtain a second detection result, where the second detection result includes that the corresponding ore is a rare metal ore or the corresponding ore is not a rare metal ore;

[0028] Based on the second detection result, adjust the parameters in the pre-trained model to obtain an ore sorting model.

[0029] In one embodiment, before the steps before obtaining the ore pictures, the following steps are further included:

[0030] Obtain a third ore dataset, where the third ore dataset is composed of ore sample pictures corresponding to the ores in the target mining area, and the ores in the target mining area are composed of associated ores and target metals in the target mining area;

[0031] Detect the ore sample pictures in the third ore dataset based on the pre-trained model to obtain a third detection result, where the third detection result includes that the corresponding ore is a rare metal ore or the corresponding ore is not a rare metal ore;

[0032] Based on the third detection result, adjust the parameters in the pre-trained model to obtain an ore sorting model.

[0033] In addition, to achieve the above object, the present application further provides a rare metal ore sorting device, and the rare metal ore sorting device includes:

[0034] A picture acquisition module, configured to acquire ore pictures;

[0035] An ore sorting module, configured to sort the ores corresponding to the ore pictures through a preset ore sorting model, where the ore sorting model is trained based on rare metal ore sample pictures with different petrographic characteristics to sort rare metal ores based on the petrographic characteristics of the ores.

[0036] In addition, to achieve the above object, the present application further provides a rare metal ore sorting device, and the device includes: a memory, a processor, and a computer program stored on the memory and executable on the processor, and the computer program is configured to implement the steps of the rare metal ore sorting method as described above.

[0037] In addition, to achieve the above object, the present application further provides a storage medium, and the storage medium is a computer-readable storage medium, and a computer program is stored on the storage medium, and when the computer program is executed by a processor, the steps of the rare metal ore sorting method as described above are implemented.

[0038] One or more technical solutions proposed by the present application have at least the following technical effects:

[0039] Obtain an ore picture, and sort the ore corresponding to the ore picture through a preset ore sorting model, where the ore sorting model is trained based on rare metal ore pictures with different lithofacies to sort rare metal ores based on the lithofacies characteristics of the ores.

[0040] For rare metal ores with complex lithofacies characteristics, this application trains an ore sorting model based on rare metal ore pictures with different lithofacies. Through the trained ore sorting model, the sorting system can be adjusted according to the lithofacies characteristics of the ores during ore sorting, improving the detection accuracy of rare metal ores, and thus improving the metal recovery rate of rare metal ore sorting. Brief Description of the Drawings

[0041] The drawings here are incorporated into the description and form a part of this description, showing embodiments consistent with this application, and are used together with the description to explain the principles of this application.

[0042] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the following will briefly introduce the drawings required for the description of the embodiments or the prior art. Obviously, for those of ordinary skill in the art, without creative efforts, other drawings can also be obtained based on these drawings.

[0043] Figure 1 It is a schematic flowchart provided for the first embodiment of the rare metal ore sorting method of this application;

[0044] Figure 2 It is a schematic flowchart provided for the second embodiment of the rare metal ore sorting method of this application;

[0045] Figure 3 It is a schematic flowchart provided for the third embodiment of the rare metal ore sorting method of this application;

[0046] Figure 4 It is a schematic module structure diagram of the rare metal ore sorting device for the embodiments of this application;

[0047] Figure 5 It is a schematic device structure diagram of the hardware operating environment involved in the rare metal ore sorting method for the embodiments of this application.

[0048] The realization of the purpose, functional features, and advantages of this application will be further described in combination with the embodiments with reference to the drawings. Detailed Embodiments

[0049] It should be understood that the specific embodiments described here are only used to explain the technical solutions of this application and are not used to limit this application.

[0050] To better understand the technical solution of this application, the following will be described in detail in conjunction with the accompanying drawings of the specification and specific implementation manners.

[0051] It should be noted that the execution subject of this embodiment can be a computing service device with data processing, network communication, and program running functions, such as a tablet computer, a personal computer, a mobile phone, etc., or an electronic device, a rare metal ore sorting device, etc. that can implement the above functions. The following takes the rare metal ore sorting device as an example to illustrate this embodiment and the following embodiments.

[0052] Based on this, the embodiment of this application provides a rare metal ore sorting method, referring to Figure 1 , Figure 1 which is a schematic flowchart of the first embodiment of the rare metal ore sorting method of this application.

[0053] In this embodiment, the rare metal ore sorting method includes steps S10 to S20:

[0054] Step S10, obtain an ore picture;

[0055] It should be noted that the ore picture is the picture corresponding to the ore after crushing and screening.

[0056] It can be understood that most ores are composed of valuable rare metals and useless gangue. Therefore, it is necessary to crush and screen the ores to separate the valuable rare metals in the ores from the gangue as much as possible, providing favorable conditions for the subsequent sorting of rare metals.

[0057] In a feasible implementation manner, the specific implementation manner after obtaining the ore picture can also be:

[0058] Preprocess the ore picture to obtain a preprocessed picture, splice a plurality of the preprocessed pictures with a preset splicing quantity to obtain a spliced picture, and sort the ore corresponding to the spliced picture through a preset ore sorting model.

[0059] It should be noted that the preprocessing operations of the picture include operations such as distortion correction, brightness adjustment (overexposure, underexposure), white balance, etc. The preset splicing quantity can be set based on specific requirements. Usually, the preset splicing quantity is 2, 4, or 8. The picture splicing operation is to splice multiple pictures into a larger-sized picture.

[0060] It can be understood that the shooting conditions of ore pictures will affect the quality of the obtained ore pictures. For example, ore pictures taken under different lighting conditions may have different color performances. At this time, through white balance operation, the colors of different images can be made consistent. Therefore, in this embodiment, it is necessary to preprocess the obtained ore pictures first, so that the important features in the ore pictures are more obvious and consistent, and then the rare metal ores are detected and classified through the preprocessed ore pictures.

[0061] Due to the large-scale parallel processing ability of high-performance computing hardware such as GPUs, using only a single picture for detection cannot fully utilize the parallel processing ability of the hardware. Therefore, in this embodiment, multiple pictures are spliced, and the larger-sized picture obtained by splicing is used for detection, avoiding the problem of partial hardware idleness caused by processing small pictures, thereby improving the utilization rate of computing resources and the detection efficiency.

[0062] In one embodiment, a high-performance image recognition server can be configured to provide image processing and analysis services for multiple production lines at the same time. This avoids the need for each production line to be equipped with independent high-cost computing devices, thus realizing the effective sharing of hardware resources, reducing the direct cost of equipment and the subsequent maintenance cost. And through the centralized image recognition server, it is convenient for unified management and scheduling, and the computing resources can be dynamically allocated according to the actual needs of each production line to improve the utilization rate of hardware resources.

[0063] Step S20, sorting the ore corresponding to the ore picture through a preset ore sorting model, where the ore sorting model is trained based on rare metal ore sample pictures with different petrographic characteristics to sort rare metal ores based on the petrographic characteristics of the ores.

[0064] It should be noted that the petrographic characteristics of ores are the combination, structure of the internal mineral components of the ores and their distribution in the ores.

[0065] It can be understood that the petrographic characteristics of rare metal ores are relatively complex. Existing sorting equipment usually relies on the physical and chemical properties (density and magnetism) of ores for sorting, but these properties do not fully consider the petrographic combination of ores, making traditional sorting methods (gravity separation and flotation) prone to low metal recovery rates.

[0066] In order to improve the metal recovery rate of ore separation, when performing operations such as ore crushing, screening, and separation, it is necessary to fully consider the petrographic characteristics of the ore. Therefore, in this embodiment, an ore separation model is trained with pictures of rare metal ore samples having different petrographic characteristics, so that the model can learn different petrographic phases of rare metal ore during the training process. Therefore, the ore separation model obtained through training can be adaptively adjusted according to the petrographic characteristics of the ore during separation, improving the detection accuracy of rare metal ore, and thus enhancing the metal recovery rate of rare metal ore separation.

[0067] In a feasible implementation manner, the ore separation model has a channel attention module and a spatial attention module. The specific implementation manner of separating the ore corresponding to the ore picture through the preset ore separation model can also be:

[0068] Through the channel attention module and the spatial attention module, feature extraction is performed on the ore picture to obtain the petrographic characteristics. Based on the petrographic characteristics and the detection layer of the ore separation model, detection is performed to obtain a target detection result, where the target detection result includes that the corresponding ore is rare metal ore or the corresponding ore is not rare metal ore. Based on the target detection result, separation of rare metal ore is performed.

[0069] It should be noted that the channel attention module and the spatial attention module respectively obtain the importance of each feature channel and each feature space in the input feature layer, and use the obtained importance to enhance important features and suppress features that are not important for the current task.

[0070] It can be understood that some rare metal ores have complex petrographic characteristics, and multiple different minerals are mixed together in the form of fine particles. In order to accurately separate the target rare metal from the ore with complex petrographic characteristics, it is necessary to pay attention to the important features in the ore.

[0071] When processing ore images, the channel attention module can automatically weight the feature channels that are more helpful for classification by learning which channel information is more important, and the spatial attention module focuses on highlighting the most important spatial positions or regions in the image. Therefore, in this embodiment, through the channel attention module and the spatial attention module, it is possible to pay attention to the key features in the key areas of rare metal ore, improve the accuracy of rare metal ore separation, thereby reducing the missed selection of rare metal ore and reducing the loss of ore.

[0072] In a feasible implementation manner, the specific implementation manner of performing feature extraction on the ore picture through the channel attention module and the spatial attention module to obtain the petrographic characteristics can also be:

[0073] Based on the channel attention module, channel features of the ore picture are extracted to obtain a channel feature map. Based on the spatial attention module, max pooling and average pooling are performed on the channel feature map to obtain a max pooling feature and an average pooling feature. The max pooling feature and the average pooling feature are concatenated to obtain a concatenated feature. A convolution operation is performed on the concatenated feature to obtain a spatial weight of the feature map. The spatial weight of the feature map and the channel feature map are subjected to a dot product to obtain a lithofacies feature.

[0074] It can be understood that different mineral components in the ore will produce characteristic differences in different channels of the ore picture. Therefore, in this embodiment, the channel attention module enables the model to focus on important channels in the ore picture, and the channel features after average pooling and max pooling are concatenated to highlight the most significant channel feature points while retaining global structure information.

[0075] Performing spatial convolution on the concatenated feature can enable the model to further focus on important regions of the ore picture while focusing on important channels of the ore picture. By performing a dot product operation on the calculated spatial weight of the feature map and the original channel feature map, the finally obtained lithofacies feature is a feature containing key information at the channel level and information on important lithofacies regions, thereby improving the rare metal ore detection ability of the trained model.

[0076] In summary, in this embodiment, an ore picture is obtained, and the ore corresponding to the ore picture is sorted by a preset ore sorting model, where the ore sorting model is trained based on rare metal ore pictures with different lithofacies to sort rare metal ores based on the lithofacies characteristics of the ores.

[0077] For rare metal ores with complex lithofacies characteristics, in this embodiment, an ore sorting model is trained based on rare metal ore pictures with different lithofacies. Through the trained ore sorting model, the sorting system can be adjusted according to the lithofacies characteristics of the ores during ore sorting, improving the detection accuracy of rare metal ores, thereby enhancing the metal recovery rate of rare metal ore sorting.

[0078] Based on the first embodiment of the present application, in the second embodiment of the present application, for content that is the same as or similar to the above-mentioned first embodiment, reference may be made to the above introduction and will not be repeated hereinafter. On this basis, please refer to Figure 2 Before step S10, the rare metal ore sorting method further includes steps S01 to S03:

[0079] Step S01, obtaining a first ore data set, where the first ore data set is composed of ore sample pictures corresponding to ores with a radius smaller than a preset threshold;

[0080] It should be noted that the preset threshold is the radius threshold of the rare metal ore after crushing, which can be set based on specific requirements. The preset threshold in this embodiment is set to 20 mm.

[0081] It can be understood that the radius of the ore after the rare metal ore is crushed is between 20 mm and 150 mm, and the radius of some small-grained ores will be less than 20 mm. When detecting the above small-grained ores by the target detection method, due to reasons such as the insufficient ability of the image to capture the details of the small-grained ores, the accuracy of the model in detecting the small-grained ores is relatively low. Therefore, in this embodiment, the model is trained with a small-grained ore dataset to improve the model's detection ability for small-grained ores.

[0082] Step S02: Detect the ore sample pictures in the first ore dataset based on the pre-trained model to obtain a first detection result, where the first detection result includes whether the corresponding ore is a rare metal ore or the corresponding ore is not a rare metal ore.

[0083] It should be noted that the pre-trained model is a model obtained after training on a large target detection dataset.

[0084] It can be understood that the target detection dataset used for pre-training contains a small number of small target objects. Therefore, in this embodiment, a small-grained rare mineral dataset is used to further train the pre-trained model for the task of small target mineral detection to obtain an ore sorting model.

[0085] Since the pre-trained model is trained in a large target detection dataset, the obtained pre-trained model can recognize various general features such as edges and textures. Therefore, training based on the pre-trained model can reduce the resource consumption required for training and improve the accuracy of the final model.

[0086] Step S03: Based on the first detection result, adjust the preset parameters to be updated in the pre-trained model to obtain an ore sorting model, where the parameters other than the parameters to be updated are frozen during the training process of the ore sorting model.

[0087] It should be noted that the frozen layers are the layers containing the general features of the model and the layers that are insensitive to the small target dataset. In this embodiment, mainly the shallow layers of the neural network close to the input end and the layers in the deep neural network that are insensitive to the small target dataset are frozen.

[0088] It can be understood that in the pre-trained deep learning model, the shallow layers usually learn more basic and general features. By freezing these layers, it can be ensured that the model does not lose the general features that have been learned, so as to maintain its performance in the detection task of small-grained rare metal ores.

[0089] Moreover, since this embodiment fine-tunes based on a pre-trained model using a small-grained dataset, if the entire network is fine-tuned comprehensively, it may lead to model instability. Therefore, in this embodiment, the shallow layers of the neural network and the layers insensitive to the small-target dataset are frozen during training, thereby avoiding unnecessary parameter updates and maintaining the stability of the model. And since only a small number of parameters participate in the training, the number of parameters to be optimized is reduced, so that the model can reach the convergence state faster, improving the training efficiency of the model.

[0090] In summary, in this embodiment, a first ore dataset is obtained, where the first ore dataset is composed of ore sample pictures corresponding to ores with a radius smaller than a preset threshold. The ore sample pictures in the first ore dataset are detected based on a pre-trained model to obtain a first detection result, where the first detection result includes that the corresponding ore is a rare metal ore or the corresponding ore is not a rare metal ore. Based on the first detection result, the preset parameters to be updated in the pre-trained model are adjusted to obtain an ore sorting model, where the parameters other than the parameters to be updated are frozen during the training process of the ore sorting model.

[0091] The pre-trained model is obtained by training in a large-scale object detection dataset. Therefore, in this embodiment, by fine-tuning the model based on the pre-trained model using a small-grained rare mineral dataset, the detection accuracy of the model for small-grained ores can be improved. And since this embodiment fine-tunes based on the pre-trained model using a small-grained dataset, in order to avoid model instability caused by comprehensive fine-tuning, in this embodiment, the shallow layers containing general features and the layers insensitive to the small-target dataset in the model are frozen during fine-tuning, thereby enhancing the stability during model fine-tuning and improving the training efficiency of the model.

[0092] Based on the first embodiment and the second embodiment of the present application, in the third embodiment of the present application, the same or similar content as in the above-mentioned first embodiment and second embodiment can be referred to the above introduction and will not be elaborated hereinafter. On this basis, please refer to Figure 3 , before step S10, the rare metal ore sorting method further includes steps A01 to A03:

[0093] Step A01, obtain a second ore dataset, where the second ore dataset is composed of ore pictures corresponding to ores with different grades. Among the ores with different grades, the number of low-grade ores is higher than a preset quantity threshold, and the low-grade ores are ores with a target metal content lower than a preset content threshold;

[0094] It should be noted that the preset quantity threshold and the preset content threshold are set based on specific task requirements.

[0095] It is understandable that the actual ore sorting process requires sorting ores of different grades, and the characteristics between low-grade ores and high-grade ores are usually different. The model trained with high-grade ores cannot accurately detect low-grade ores. Therefore, in order to enable the trained ore sorting model to have the generalization ability to sort ores of different grades, it is necessary to train the model with ore pictures of different grades.

[0096] Moreover, since the number of low-grade ores is usually higher than that of high-grade ores, it is necessary to make the number of low-grade ores higher than the preset quantity threshold, so that the trained model can improve the detection and sorting accuracy of low-grade ores. And in this embodiment, the model is trained with ore pictures of different grades, which can also prevent overfitting of the model.

[0097] Step A02: Detect the ore sample pictures in the second ore data set based on the pre-trained model to obtain a second detection result, where the second detection result includes that the corresponding ore is a rare metal ore or the corresponding ore is not a rare metal ore;

[0098] It is understandable that the pre-trained model is a model that has learned the general features of the object detection task. In this embodiment, based on the pre-trained model, fine-tuning pre-training is performed with rare metal ores of different grades, and the number of low-grade ores is made higher than the preset quantity threshold, so that the model can learn the ore characteristics of different grades, thereby improving the detection and sorting accuracy of low-grade ores and preventing overfitting.

[0099] Step A03: Adjust the parameters in the pre-trained model based on the second detection result to obtain an ore sorting model.

[0100] In a feasible implementation manner, the specific implementation manner before obtaining the ore pictures may also be:

[0101] Obtain a third ore data set, where the third ore data set is composed of ore sample pictures corresponding to the ores in the target mining area. The ores in the target mining area are composed of associated ores and target metals in the target mining area. Detect the ore sample pictures in the third ore data set based on the pre-trained model to obtain a third detection result, where the third detection result includes that the corresponding ore is a rare metal ore or the corresponding ore is not a rare metal ore. Adjust the parameters in the pre-trained model based on the third detection result to obtain an ore sorting model.

[0102] It can be understood that in this embodiment, rare metal ores are detected and sorted based on the petrographic characteristics of the ores. Since the geological conditions in different regions are different, the types of associated ores and the appearance characteristics of rare metal ores in different regions are different, that is, the petrographic characteristics of the ores are related to the regions of the ores and ore veins.

[0103] Therefore, when sorting rare metal ores, in this embodiment, the model is specifically trained through the ore sample pictures of the target mining area, so that the model can better learn the petrographic characteristics of the ores in the target mining area. When the ore sorting model obtained through training is used to detect the ores in the target mining area, the accuracy of rare metal ore detection can be improved.

[0104] In a feasible implementation manner, the specific implementation manner before obtaining the ore pictures may also be:

[0105] Obtain a background data set and an ore sample data set. The background data set is composed of pictures of conveyor belts for transporting ores. Based on a pre-trained model, the pictures in the background data set and the ore sample data set are detected to obtain a fourth detection result, where the fourth detection result includes that the corresponding ore is a rare metal ore or the corresponding ore is not a rare metal ore. Based on the fourth detection result, the parameters in the pre-trained model are adjusted to obtain the ore sorting model.

[0106] It can be understood that the background of rare metal ore sorting is relatively single, so that the ore sorting model cannot obtain relevant information of the ore through the background of the ore. Therefore, when sorting the ore, more inferences are required, which increases the computing consumption of the image recognition server.

[0107] Therefore, in this embodiment, by increasing the background data set and fine-tuning the pre-training through the background data set, the trained ore sorting model can detect the ore through the background information of the ore picture, thereby reducing the computational complexity of image detection and image classification algorithm inference, and correspondingly reducing the computing power consumption of the image recognition server.

[0108] In summary, in this embodiment, a second ore data set is obtained, where the second ore data set is composed of ore pictures corresponding to ores of different grades. Among the ores of different grades, the number of low-grade ores is higher than a preset number threshold, and the low-grade ores are ores with a target metal content lower than a preset content threshold. Based on a pre-trained model, the ore sample pictures in the second ore data set are detected to obtain a second detection result, where the second detection result includes that the corresponding ore is a rare metal ore or the corresponding ore is not a rare metal ore. Based on the second detection result, the parameters in the pre-trained model are adjusted to obtain the ore sorting model.

[0109] In the actual ore sorting process, the rare metal ores to be sorted are ores of different grades from different mining areas, with corresponding petrographic characteristics. Therefore, in this embodiment, the model is trained with ore datasets from different mining areas and different grades, enabling the model to better learn the petrographic characteristics of the corresponding ores, thereby improving the detection accuracy of the ore sorting model for rare metal ores. Moreover, in this embodiment, the model is also fine-tuned and pre-trained by adding a background dataset, enabling the model to perform ore sorting based on background information, thereby reducing the computing power consumption of the image recognition server.

[0110] It should be noted that the above examples are only for understanding this application and do not constitute a limitation on the rare metal ore sorting method of this application. Based on this technical concept, more forms of simple transformations are within the protection scope of this application.

[0111] This application also provides a rare metal ore sorting device. Please refer to Figure 4 , the rare metal ore sorting device includes:

[0112] An image acquisition module 10 for acquiring ore images;

[0113] An ore sorting module 20 for sorting the ore corresponding to the ore image through a preset ore sorting model, where the ore sorting model is trained based on rare metal ore sample images with different petrographic characteristics to sort rare metal ores based on the petrographic characteristics of the ores.

[0114] In one embodiment, the rare metal ore sorting device further includes:

[0115] A first dataset acquisition module for acquiring a first ore dataset, where the first ore dataset is composed of ore sample images corresponding to ores with a radius smaller than a preset threshold;

[0116] A first detection module for detecting the ore sample images in the first ore dataset based on a pre-trained model to obtain a first detection result, where the first detection result includes whether the corresponding ore is a rare metal ore or the corresponding ore is not a rare metal ore;

[0117] A first parameter adjustment module for adjusting the preset parameters to be updated in the pre-trained model based on the first detection result to obtain an ore sorting model, where the parameters other than the parameters to be updated are frozen during the training process of the ore sorting model.

[0118] In one embodiment, the ore sorting module further includes:

[0119] A feature extraction sub-module, which is used to extract features from the ore picture through the channel attention module and the spatial attention module to obtain the petrographic features;

[0120] An ore detection sub-module, which is used to perform detection based on the petrographic features and the detection layer of the ore sorting model to obtain a target detection result, where the target detection result includes that the corresponding ore is a rare metal ore or the corresponding ore is not a rare metal ore;

[0121] An ore sorting sub-module, which is used to sort rare metal ores based on the target detection result.

[0122] In one embodiment, the feature extraction sub-module further includes:

[0123] A channel feature extraction unit, which is used to extract channel features from the ore picture based on the channel attention module to obtain a channel feature map;

[0124] A spatial feature extraction unit, which is used to perform max pooling and average pooling on the channel feature map based on the spatial attention module to obtain a max pooling feature and an average pooling feature;

[0125] A feature connection unit, which is used to connect the max pooling feature and the average pooling feature to obtain a connection feature;

[0126] A weight calculation unit, which is used to perform a convolution operation on the connection feature to obtain a feature map spatial weight;

[0127] A feature dot product unit, which is used to perform a dot product on the feature map spatial weight and the channel feature map to obtain the petrographic features.

[0128] In one embodiment, the rare metal ore sorting device further includes:

[0129] A picture processing module, which is used to preprocess the ore picture to obtain a preprocessed picture;

[0130] A picture splicing module, which is used to splice a plurality of the preprocessed pictures with a preset splicing quantity to obtain a spliced picture;

[0131] A splicing and sorting module, which sorts the ores corresponding to the spliced picture through a preset ore sorting model.

[0132] In one embodiment, the rare metal ore sorting device further includes:

[0133] A second data set acquisition module for acquiring a second ore data set, where the second ore data set is composed of ore pictures corresponding to ores of different grades, and among the ores of different grades, the number of low-grade ores is higher than a preset quantity threshold, and the low-grade ores are ores with a target metal content lower than a preset content threshold;

[0134] A second detection module for detecting ore sample pictures in the second ore data set based on a pre-trained model to obtain a second detection result, where the second detection result includes that the corresponding ore is a rare metal ore or the corresponding ore is not a rare metal ore;

[0135] A second parameter adjustment module for adjusting parameters in the pre-trained model based on the second detection result to obtain an ore sorting model.

[0136] In one embodiment, the rare metal ore sorting device further includes:

[0137] A third data set acquisition module for acquiring a third ore data set, where the third ore data set is composed of ore sample pictures corresponding to ores in a target mining area, and the ores in the target mining area are composed of associated ores and target metals in the target mining area;

[0138] A third detection module for detecting ore sample pictures in the third ore data set based on a pre-trained model to obtain a third detection result, where the third detection result includes that the corresponding ore is a rare metal ore or the corresponding ore is not a rare metal ore;

[0139] A third parameter adjustment module for adjusting parameters in the pre-trained model based on the third detection result to obtain an ore sorting model.

[0140] The rare metal ore sorting device provided by the present application adopts the rare metal ore sorting method in the above embodiment, and can solve the technical problem of low metal recovery rate in rare metal ore sorting. Compared with the prior art, the beneficial effects of the rare metal ore sorting device provided by the present application are the same as those of the rare metal ore sorting method provided by the above embodiment, and other technical features in the rare metal ore sorting device are the same as the features disclosed in the above embodiment method, and will not be elaborated here.

[0141] The present application provides a rare metal ore sorting device, which includes: at least one processor; and a memory communicatively connected to the at least one processor; wherein, the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor so that the at least one processor can execute the rare metal ore sorting method in the first embodiment above.

[0142] The following refers to Figure 5 , which shows a schematic structural diagram of a rare metal ore sorting device suitable for implementing the embodiments of the present application. The rare metal ore sorting device in the embodiments of the present application may include, but is not limited to, mobile terminals such as mobile phones, tablet computers, laptop computers, digital broadcast receivers, PDAs (Personal Digital Assistants), PMPs (Portable Media Players), vehicle-mounted terminals (such as vehicle-mounted navigation terminals), etc., and fixed terminals such as digital TVs, desktop computers, etc. Figure 5 The shown rare metal ore sorting device is merely an example and should not impose any limitations on the functions and usage scope of the embodiments of the present application.

[0143] As Figure 5 shown, the rare metal ore sorting device may include a processing device 1001 (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 a read-only memory (ROM: Read Only Memory) 1002 or a program loaded from a storage device 1003 into a random access memory (RAM: Random Access Memory) 1004. In the RAM 1004, various programs and data required for the operation of the rare metal ore sorting device are also stored. The processing device 1001, the ROM 1002, and the RAM 1004 are connected to each other through a bus 1005. An input / output (I / O) interface 1006 is also connected to the bus. Generally, the following systems may be connected to the I / O interface 1006: an input device 1007 including, for example, a touch screen, a touchpad, a keyboard, a mouse, an image sensor, a microphone, an accelerometer, a gyroscope, etc.; an output device 1008 including, for example, a liquid crystal display (LCD: Liquid Crystal Display), a speaker, a vibrator, etc.; a storage device 1003 including, for example, magnetic tapes, hard disks, etc.; and a communication device 1009. The communication device 1009 may allow the rare metal ore sorting device to communicate with other devices wirelessly or wiredly to exchange data. Although the figure shows a rare metal ore sorting device having various systems, it should be understood that it is not required to implement or have all the shown systems. More or fewer systems may be alternatively implemented or had.

[0144] In particular, according to the embodiments disclosed in the present application, the processes described above with reference to the flowcharts can be implemented as computer software programs. For example, the embodiments disclosed in the present application include a computer program product, which includes a computer program carried on a computer-readable medium, and the computer program contains program codes for executing the methods shown in the flowcharts. In such an embodiment, the computer program can be downloaded and installed from a network through a communication device, or installed from a storage device 1003, or installed from a ROM 1002. When the computer program is executed by a processing device 1001, the above functions defined in the methods of the embodiments disclosed in the present application are executed.

[0145] The rare metal ore sorting equipment provided by the present application adopts the rare metal ore sorting method in the above embodiment, and can solve the technical problem of low metal recovery rate in rare metal ore sorting. Compared with the prior art, the beneficial effects of the rare metal ore sorting equipment provided by the present application are the same as those of the rare metal ore sorting method provided by the above embodiment, and other technical features in the rare metal ore sorting equipment are the same as those disclosed in the method of the previous embodiment, and will not be described in detail here.

[0146] It should be understood that the various parts disclosed in the present application can be implemented by hardware, software, firmware or a combination thereof. In the description of the above embodiments, specific features, structures, materials or characteristics can be combined in a suitable manner in any one or more embodiments or examples.

[0147] As described above, only the specific embodiments of the present application are provided, but the protection scope of the present application is not limited thereto. Any person skilled in the art can easily think of changes or substitutions within the technical scope disclosed in the present application, and all of them should be covered by the protection scope of the present application. Therefore, the protection scope of the present application should be subject to the protection scope of the claims.

[0148] The present application provides a computer-readable storage medium, on which computer-readable program instructions (i.e., computer programs) are stored, and the computer-readable program instructions are used to execute the rare metal ore sorting method in the above embodiment.

[0149] The computer-readable storage medium provided by this application can be, for example, a USB flash drive, but is not limited to electrical, magnetic, optical, electromagnetic, infrared, or semiconductor systems, devices, or components, or any combination of the above. More specific examples of computer-readable storage media can include, but are not limited to: electrical connections with 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 above. In this embodiment, the computer-readable storage medium can be any tangible medium that contains or stores a program, and this program can be used by or in combination with an instruction execution system, device, or component. The program code contained on the computer-readable storage medium can be transmitted using any suitable medium, including but not limited to: wires, optical cables, RF (radio frequency), etc., or any suitable combination of the above.

[0150] The above computer-readable storage medium can be included in the rare metal ore sorting equipment; it can also exist independently without being assembled into the rare metal ore sorting equipment.

[0151] The above computer-readable storage medium carries one or more programs. When the above one or more programs are executed by the rare metal ore sorting equipment, the rare metal ore sorting equipment is caused to: execute the above rare metal ore sorting method.

[0152] Computer program code for performing the operations of this application can be written in one or more programming languages or combinations thereof. The above 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 can be executed entirely on the user's computer, partially on the user's computer, executed as an independent software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In the case of a remote computer, the remote computer can 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 can be connected to an external computer (for example, by using an Internet service provider to connect through the Internet).

[0153] 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 application. In this regard, each box in the flowchart or block diagram may represent a module, a program segment, or a portion of code that contains one or more executable instructions for implementing a specified logical function. It should also be noted that in some alternative implementations, the functions marked in the boxes may occur in a different order than that marked in the accompanying drawings. For example, two consecutive boxes 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 box in the block diagram and / or flowchart, and combinations of boxes 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.

[0154] The modules described in the embodiments of the present application can be implemented in software or in hardware. Among them, the name of the module does not constitute a limitation to the unit itself in some cases.

[0155] The readable storage medium provided by the present application is a computer-readable storage medium, and the computer-readable storage medium stores computer-readable program instructions (i.e., computer programs) for executing the above-mentioned rare metal ore sorting method, which can solve the technical problem of low metal recovery rate in rare metal ore sorting. Compared with the prior art, the beneficial effects of the computer-readable storage medium provided by the present application are the same as those of the rare metal ore sorting method provided by the above embodiments, and will not be elaborated here.

[0156] The above are only some embodiments of the present application, and do not limit the patent scope of the present application. All equivalent structural transformations made under the technical concept of the present application by using the content of the specification and drawings of the present application, or direct / indirect applications in other related technical fields are included in the patent protection scope of the present application.

Claims

1. A method for separating rare metal ores, characterized in that: The method includes: Get ore pictures; The ore corresponding to the ore image is sorted by a preset ore sorting model, wherein the ore sorting model is trained based on rare metal ore sample images with different lithological characteristics, so as to sort the rare metal ore based on the lithological characteristics of the ore.

2. The method according to claim 1, characterized in that Before the step of obtaining the ore picture, the method further includes: Acquire a first ore data set, wherein the first ore data set is composed of ore sample images corresponding to ores with a radius smaller than a preset threshold; Detecting the ore sample images in the first ore data set based on the pre-trained model to obtain a first detection result, wherein the first detection result includes that the corresponding ore is a rare metal ore or that the corresponding ore is not a rare metal ore; Based on the first detection result, the parameters to be updated preset in the pre-trained model are adjusted to obtain an ore sorting model, wherein parameters other than the parameters to be updated are frozen during the training process of the ore sorting model.

3. The method according to claim 1, characterized in that The ore sorting model has a channel attention module and a spatial attention module. The step of sorting the ore corresponding to the ore picture by using the preset ore sorting model includes: By using the channel attention module and the spatial attention module, feature extraction is performed on the ore image to obtain the lithofacies feature; Performing detection based on the petrographic characteristics and the detection layer of the ore sorting model to obtain a target detection result, wherein the target detection result includes that the corresponding ore is a rare metal ore or that the corresponding ore is not a rare metal ore; Based on the target detection results, rare metal ores are sorted.

4. The method according to claim 3, characterized in that The step of extracting features from the ore image through the channel attention module and the spatial attention module to obtain the lithofacies features comprises: Based on the channel attention module, channel features are extracted from the ore image to obtain a channel feature map; Based on the spatial attention module, performing maximum pooling and average pooling on the channel feature map to obtain maximum pooling features and average pooling features; Connecting the maximum pooling feature and the average pooling feature to obtain a connection feature; Performing a convolution operation on the connection features to obtain a feature map spatial weight; The spatial weight of the feature map and the channel feature map are dot-producted to obtain the lithofacies feature.

5. The method according to claim 1, characterized in that After the step of obtaining the ore picture, the method further includes: Preprocessing the ore image to obtain a preprocessed image; Splicing a preset number of pre-processed images to obtain a spliced ​​image; The ore corresponding to the spliced ​​image is sorted using a preset ore sorting model.

6. The method according to claim 1, characterized in that: Before the step of obtaining the ore picture, the method further includes: Acquire a second ore data set, wherein the second ore data set is composed of ore images corresponding to ores of different grades, wherein the number of low-grade ores among the ores of different grades is higher than a preset number threshold, and the low-grade ores are ores having a target metal content lower than a preset content threshold; Detecting the ore sample images in the second ore data set based on the pre-trained model to obtain a second detection result, wherein the second detection result includes that the corresponding ore is a rare metal ore or that the corresponding ore is not a rare metal ore; Based on the second detection result, the parameters in the pre-trained model are adjusted to obtain an ore separation model.

7. The method according to claim 1, characterized in that: Before the step of obtaining the ore picture, the method further includes: Acquire a third ore data set, wherein the third ore data set is composed of ore sample images corresponding to ore in a target mining area, and the ore in the target mining area is composed of associated ore and target metal in the target mining area; Detecting the ore sample images in the third ore data set based on the pre-trained model to obtain a third detection result, wherein the third detection result includes that the corresponding ore is a rare metal ore or that the corresponding ore is not a rare metal ore; Based on the third detection result, the parameters in the pre-trained model are adjusted to obtain an ore separation model.

8. A rare metal ore separation device, characterized in that: The device comprises: Image acquisition module, used to obtain ore images; The ore sorting module is used to sort the ore corresponding to the ore image through a preset ore sorting model, wherein the ore sorting model is trained based on rare metal ore sample images with different lithological characteristics, so as to sort the rare metal ore based on the lithological characteristics of the ore.

9. A rare metal ore sorting device, characterized in that: The device comprises: a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the computer program is configured to implement the steps of the rare metal ore separation method according to any one of claims 1 to 6.

10. A storage medium, characterized in that: The storage medium is a computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, the steps of the rare metal ore separation method according to any one of claims 1 to 6 are implemented.

Images