Coal rock maceral identification method and device, electronic equipment and storage medium

Abstract

The invention discloses a coal rock maceral identification method and device, electronic equipment and a storage medium. The method comprises the following steps: acquiring a coal rock microscopic image, wherein the coal rock microscopic image comprises a microscopic image under oil immersion reflected light and a microscopic image under fluorescence; the coal rock microscopic image is recognized based on a coal rock maceral recognition model, the coal rock maceral category is determined, and the coal rock maceral recognition model is composed of a multi-scale bimodal convolutional neural network structure. According to the technical scheme of the invention, the coal rock maceral is automatically identified through the idea of multi-scale bimodal fusion, and the method has the advantages of high accuracy, rapidness, simplicity, convenience and high repeatability.

Description

Technical Field

[0001] The present invention relates to the technical field of coal rock analysis, and in particular to a method, device, electronic equipment and storage medium for identifying coal rock microscopic components. Background Art

[0002] my country boasts abundant coal reserves and numerous coal varieties, but the distribution of coal quality is uneven, with high-quality coal accounting for a small proportion of the total coal supply. High-quality coking coal, in particular, is in short supply. Therefore, improving coal preparation efficiency and quality is crucial for addressing the country's coking coal shortage and improving energy efficiency. Coal's microscopic composition is closely linked to its combustion properties, CO adsorption, and adhesion. Furthermore, analysis of coal's microscopic composition is a crucial analytical parameter in geological exploration for coal, oil and gas, and geothermal energy. Therefore, coal microscopic composition analysis holds significant research significance.

[0003] Current coal rock microscopic component identification technologies are mostly based on microscopic images under oil-immersion reflected light, using traditional image processing technology to identify coal rock microscopic components.

[0004] Based on traditional image processing methods, this method requires manual selection of regions of interest and manual feature extraction, which is slow, has poor repeatability, and cannot identify microscopic components end-to-end. Summary of the Invention

[0005] The present invention provides a method, device, electronic equipment and storage medium for identifying microscopic components of coal rocks, which can realize automatic identification of microscopic components of coal rocks and have the advantages of high accuracy, rapidity, simplicity and high repeatability.

[0006] According to one aspect of the present invention, a method for identifying microscopic components of coal rocks is provided, the method comprising:

[0007] Acquire a microscopic image of the coal rock; wherein the microscopic image of the coal rock includes a microscopic image under oil-immersion reflected light and a microscopic image under fluorescence;

[0008] The coal rock microscopic image is identified based on a coal rock microscopic component identification model to determine the category of the coal rock microscopic components; wherein the coal rock microscopic component identification model is composed of a multi-scale bimodal convolutional neural network structure.

[0009] According to another aspect of the present invention, a device for identifying microscopic components of coal and rock is provided, the device comprising:

[0010] A coal rock image acquisition module, used to acquire a microscopic image of the coal rock; wherein the microscopic image of the coal rock includes a microscopic image under oil immersion reflected light and a microscopic image under fluorescence;

[0011] The coal rock image recognition module is used to identify the coal rock microscopic image based on the coal rock microscopic component recognition model and determine the coal rock microscopic component category; wherein the coal rock microscopic component recognition model is composed of a multi-scale bimodal convolutional neural network structure.

[0012] According to another aspect of the present invention, an electronic device is provided, comprising:

[0013] at least one processor; and

[0014] a memory communicatively connected to the at least one processor; wherein,

[0015] The memory stores a computer program that can be executed by the at least one processor, and the computer program is executed by the at least one processor so that the at least one processor can execute the method for identifying the microscopic components of coal rocks described in any embodiment of the present invention.

[0016] According to another aspect of the present invention, a computer-readable storage medium is provided, wherein the computer-readable storage medium stores computer instructions, and the computer instructions are used to enable a processor to implement the method for identifying the microscopic components of coal and rock according to any embodiment of the present invention when executed.

[0017] The technical solution of the embodiment of the present invention uses a dual-modal model to fuse the microscopic images of coal rock microscopic components under fluorescence and oil-immersion reflected light, and utilizes the different reactions of microscopic components to different light sources to fuse the two modal data to achieve automatic identification of coal rock microscopic components. It has the advantages of high accuracy, strong robustness, speed and simplicity.

[0018] It should be understood that the content described in this section is not intended to identify the key or important features of the embodiments of the present invention, nor is it intended to limit the scope of the present invention. Other features of the present invention will become readily understood through the following description. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0020] Figure 1 This is a flow chart of a method for identifying microscopic components of coal rocks provided in accordance with the first embodiment of the present invention;

[0021] Figure 2 This is a flow chart of another method for identifying microscopic components of coal rocks provided in accordance with the first embodiment of the present invention;

[0022] Figure 3 2 is a schematic diagram of the MultiScale-VGG neural network structure provided according to the first embodiment of the present invention;

[0023] Figure 4 2. This is a schematic diagram of a pruned encoder-decoder network structure according to the first embodiment of the present invention;

[0024] Figure 5 2 is a schematic diagram of the attention mechanism structure provided according to the first embodiment of the present invention;

[0025] Figure 6 This is a flow chart of a method for identifying microscopic components of coal rocks provided in accordance with the second embodiment of the present invention;

[0026] Figure 7 This is a schematic diagram of a feature layer fusion structure provided according to the second embodiment of the present invention;

[0027] Figure 8 This is a schematic structural diagram of a device for identifying microscopic components of coal and rock according to a third embodiment of the present invention;

[0028] Figure 9 It is a structural diagram of an electronic device for implementing the method for identifying the microscopic components of coal rocks according to the fourth embodiment of the present invention. DETAILED DESCRIPTION

[0029] In order to enable those skilled in the art to better understand the solutions of the present invention, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the embodiments described are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts should fall within the scope of protection of the present invention.

[0030] It should be noted that the terms "first", "second", etc. in the description and claims of the present invention and the above-mentioned drawings are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that the numbers used in this way can be interchanged where appropriate, so that the embodiments of the present invention described herein can be implemented in an order other than those illustrated or described herein. In addition, the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusions. For example, a process, method, system, product or device that includes a series of steps or units is not necessarily limited to those steps or units clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

[0031] Example 1

[0032] Figure 1A flow chart of a method for identifying microscopic components of coal rocks is provided for the first embodiment of the present invention. This embodiment is applicable to the identification of microscopic components of coal rocks. The method can be executed by a device for identifying microscopic components of coal rocks. The device can be implemented in the form of hardware and / or software. The device can be configured in any electronic device with network communication function. Figure 1 As shown, the method includes:

[0033] S110, obtaining a microscopic image of the coal rock; wherein the microscopic image of the coal rock includes a microscopic image under oil-immersion reflected light and a microscopic image under fluorescence.

[0034] In the embodiments of the present application, microscopic image composition analysis of coal rock is one of the important analytical indicators in geological exploration for coal, oil and gas, and geothermal energy. A coal rock microscopic image is an image of the organic components of bituminous coal, identifiable under a microscope. Coal rock microscopic components include vitrinite, inertinite, exinite, and background resin. A coal rock microscopic image is a microscopic image of the microstructure of coal rock or coke, obtained primarily using a microscope. Microscopic images of the coal rock microstructure are obtained under oil-immersion reflected light and fluorescence. Coal rock microscopic components react differently to different light sources, exhibiting different colors under different light sources. Table 1 provides a comparison of the color rendering of coal rock microscopic components under different light sources.

[0035] Table 1

[0036]

[0037] S120. Identify the coal rock microscopic image based on a coal rock microscopic component identification model to determine the category of the coal rock microscopic components; wherein the coal rock microscopic component identification model is composed of a multi-scale bimodal convolutional neural network structure.

[0038] In the embodiment of the present application, an unlabeled coal rock microscopic image is input, and the coal rock microscopic component classification information is output through the coal rock microscopic component recognition model. The coal rock microscopic component recognition model is composed of a multi-scale dual-modal convolutional neural network structure, specifically a dual-modal MultiScale (multi-scale)-VGG (deep convolutional network) pruned version encoding and decoding network structure. The classic semantic segmentation network is optimized based on the idea of ​​multi-scale dual-modal fusion. The encoder part uses VGG as a feature extractor to extract features at different levels. The decoder part removes the last convolution operation of each layer and introduces an attention mechanism for feature fusion to realize automatic recognition of coal rock microscopic components and obtain coal rock microscopic image segmentation results with high accuracy, strong robustness, and fast and simple operation. See Figure 2 This is a flow chart of another method for identifying the microscopic components of coal rocks.

[0039] See also Figure 3This is a diagram of the MultiScale-VGG neural network architecture. VGG (Visual Geometry Group) is a classic convolutional neural network architecture that builds deep networks by repeatedly using simple convolutional and pooling layers. The VGG network is divided into multiple stages, each with a specific feature map output. The feature maps of these stages gradually decrease in size while increasing in number of channels. This structure helps preserve more spatial information in the network and facilitates feature extraction at both lower and higher levels.

[0040] Figure 3 Here, f11 and f21 refer to the feature maps output by the first and second stage networks, and the same applies to f31 and f41. Because the network has two encoders, which operate on images of two modalities respectively, the encoder output feature of oil-immersed reflected light is recorded as f*1, and the encoder output feature of fluorescence is recorded as f*2. Figure 3 Shown is the encoder structure of a microscopic image of coal rock under oil-immersion reflected light.

[0041] The bimodal MultiScale-VGG pruned version of the codec network structure includes the following steps:

[0042] For dual-modal fusion, feature layer fusion is performed after extracting feature information of each modality;

[0043] For the encoder part of the original network, the VGG structure is adopted, and the idea of ​​Multi-Scale (multi-scale model) is introduced into the VGG structure;

[0044] The decoder part of the original network is pruned to filter feature information while retaining more effective information.

[0045] See also Figure 4 This is a schematic diagram of the pruned encoder-decoder network structure. Figure 3 The only difference is that the encoder structure is Figure 4 The complete network structure, namely the encoder-decoder structure, is shown. In neural networks, pruning is an optimization technique whose goal is to reduce the size of the model by removing redundant connections or neurons in the network, thereby improving the model's efficiency and inference speed while maintaining or improving the model's performance.

[0046] The encoder extracts features by stacking convolutional and pooling layers, gradually reducing the spatial resolution. The decoder restores features by stacking convolutional and upsampling layers, gradually increasing the spatial resolution. At each decoder layer, the feature maps are skip-connected with the corresponding feature maps from the encoder layer to preserve more detailed information. The final output layer is typically a convolutional layer with the number of output channels matching the number of target categories. Figure 4Here, f1, f2, f3, and f4 refer to the feature maps extracted by the network at different stages. The feature maps are the result of the fusion of the feature maps extracted by the two modalities. Figure 5 is a schematic diagram of the attention mechanism structure, corresponding to Figure 3 The "attention mechanism" module in .

[0047] As an optional but non-limiting implementation method, the process of determining the coal rock micro-component identification model includes:

[0048] Obtaining microscopic images of the coal rock to be processed;

[0049] Performing a coal rock microscopic component classification task on the coal rock microscopic image to be processed based on the coal rock microscopic component recognition model to be trained;

[0050] According to the coal rock microscopic component category identification task, the coal rock microscopic component identification model to be trained is adjusted to obtain a trained and updated coal rock microscopic component identification model.

[0051] In the embodiment of the present application, the coal rock microscopic images to be processed include microscopic images under oil-immersion reflected light and microscopic images under fluorescence. By repeatedly executing the coal rock microscopic component identification task, the identification result is obtained, and the internal parameters of the coal rock microscopic component identification model are adjusted according to the identification result evaluation index, so that the performance of the coal rock microscopic component identification model is improved. The specific operation of training the coal rock microscopic component identification model to be trained adopts the 5-fold cross-validation method, dividing the sample set into 5 equal parts, selecting 4 equal parts as the training set each time, and the remaining 1 part as the test set, and inputting the multimodal features obtained by the feature extraction network of the training set and the test set into the model to obtain the test set accuracy; repeating 5 times, averaging the 5 test set accuracy rates to obtain the final test set accuracy, and the test set accuracy reflects the accuracy of the trained coal rock microscopic component identification model.

[0052] The determination process of the coal rock micro-component identification model is to perform coal rock micro-component identification on multiple sets of training sets and test sets, adjust the internal parameters of the model according to the identification results, and improve the accuracy of the model's automatic identification of coal rock micro-components.

[0053] As an optional but non-limiting implementation, after obtaining the microscopic image of the coal rock to be processed, the method further includes:

[0054] Calibrate the microscopic image of the coal rock to be processed to determine a microscopic label image of the coal rock to be processed; wherein the microscopic label image of the coal rock to be processed includes the microscopic component area of ​​the coal rock and the category of the microscopic component of the coal rock;

[0055] Accordingly, the coal rock microscopic component classification task is performed on the coal rock microscopic image to be processed based on the coal rock microscopic component recognition model to be trained, including:

[0056] Based on the coal rock microscopic component recognition model to be trained, a coal rock microscopic component category recognition task is performed on the coal rock microscopic image to be processed and the coal rock microscopic label image to be processed.

[0057] In this embodiment, the coal microscopic image includes a microscopic image obtained under oil-immersion reflected light and a microscopic image obtained under fluorescence. The coal microscopic image is manually calibrated based on a color comparison table of coal microscopic components under different light sources, and the regions and categories of each component in the coal microscopic image are calibrated. The specific calibration process involves creating a new image of the same size as the coal microscopic image, referred to as a label image. After determining the category of any pixel in the coal microscopic image, a corresponding color is generated in the corresponding label image to indicate whether the pixel belongs to the coal microscopic component category of vitrinite, inertinite, exinite, or background resin.

[0058] The coal rock microscopic component recognition model to be trained refers to a dual-modal MultiScale-VGG pruned version encoding and decoding network structure. The dual-modal MultiScale-VGG pruned version encoding and decoding network structure is trained on calibrated coal rock microscopic image data. The unlabeled coal rock microscopic images are input into the trained dual-modal model to obtain the component category to which each coal rock microscopic image belongs.

[0059] By calibrating the component areas and categories of coal rock microscopic images and using the coal rock microscopic component recognition model to perform recognition tasks, the component categories to which the coal rock microscopic images belong are obtained, thereby realizing automatic recognition of coal rock microscopic components.

[0060] As an optional but non-limiting implementation, the to-be-trained coal rock micro-component identification model is adjusted according to the coal rock micro-component classification identification task, including:

[0061] Determining a loss function value corresponding to the coal rock microscopic component classification task;

[0062] The network parameters in the coal rock micro-component identification model to be trained are adjusted according to the loss function value.

[0063] In the embodiments of this application, the loss function of the coal rock micro-component identification model is used to measure the distance between the output of the neural network model and the expected value, facilitating control and parameter adjustment to minimize the difference between the output and the expected value. Mean square error and cross entropy error are commonly used as loss functions. The loss function value is used to adjust the parameters of the coal rock micro-component identification model, thereby improving the accuracy of the identification results.

[0064] As an optional but non-limiting implementation, after obtaining the trained and updated coal rock microscopic component identification model, the method further includes:

[0065] The output result of the trained and updated coal rock micro-component identification model is evaluated based on a predetermined evaluation method; wherein the evaluation method includes pixel accuracy, average pixel accuracy, intersection-over-union ratio and average intersection-over-union ratio.

[0066] In the embodiments of this application, evaluating the output of a trained coal rock microscopic component identification model refers to the process of evaluating the performance of the segmentation results obtained after performing an image segmentation task on coal rock microscopic images. Evaluating segmentation results is crucial for understanding the performance of the coal rock microscopic component identification model, adjusting parameters, selecting appropriate algorithms, and comparing the performance of different models.

[0067] The evaluation criteria for the coal rock micro-component recognition model include pixel accuracy (PA), mean pixel accuracy (MPA), intersection over union (IoU), and mean intersection over union (MIoU). The specific calculation formulas are as follows:

[0068]

[0069]

[0070]

[0071]

[0072] Where C is the total number of classes, p mij It refers to the total number of pixels in the mth sample that belong to class i but are predicted to be class j, p mji It refers to the total number of pixels in the mth sample that belong to class j but are predicted to be class i, p mii It refers to the total number of pixels in the mth sample that belong to class i and are also predicted to be class i, and M refers to the total number of coal rock microscopic image samples used.

[0073] The performance of the coal rock micro-component identification model is obtained by calculating the output results of the coal rock micro-component identification model according to multiple parameters.

[0074] The present invention discloses a method for identifying coal rock microscopic components. The method comprises: obtaining a coal rock microscopic image, wherein the coal rock microscopic image includes a microscopic image under oil immersion reflected light and a microscopic image under fluorescence; and identifying the coal rock microscopic image based on a coal rock microscopic component identification model to determine the coal rock microscopic component category. The coal rock microscopic component identification model comprises a multi-scale bimodal convolutional neural network structure. The technical solution of the present invention uses a bimodal model to fuse microscopic images of coal rock microscopic components under fluorescence and oil immersion reflected light. By utilizing the different responses of microscopic components to different light sources, the two modal data are integrated to achieve automatic identification of coal rock microscopic components. This method has the advantages of high accuracy, strong robustness, and rapidity and simplicity.

[0075] Example 2

[0076] Figure 6 This is a flow chart of a method for identifying the microscopic components of coal rocks provided in the second embodiment of the present invention. This embodiment is optimized based on the above embodiment. For solutions not fully described in the embodiments of this application, please refer to the above embodiment. Figure 6 As shown, the method includes:

[0077] S610, obtaining a microscopic image of the coal rock; wherein the microscopic image of the coal rock includes a microscopic image under oil-immersion reflected light and a microscopic image under fluorescence.

[0078] S620. Based on the coal rock microscopic component identification model, feature extraction is performed on the microscopic image under oil immersion reflected light to obtain an oil immersion reflected light feature map, and feature extraction is performed on the microscopic image under fluorescence to obtain a fluorescence feature map.

[0079] In the examples of this application, see Figure 7 The schematic diagram of the characteristic layer fusion structure is shown in Figure 2. The coal rock microscopic image obtained under oil-immersion reflected light is obtained by Figure 3 The VGG feature extractor network shown extracts an oil-immersion reflected light feature map. A fluorescence feature map is also extracted from a fluorescent coal microscopic image using the VGG feature extractor network. The extracted feature map information is automatically extracted from the coal microscopic image by the VGG neural network model. Since the neural network is a black box model, the extracted feature map information has no specific meaning and is the result of the neural network model's automatic learning.

[0080] S630: Fusing the oil-immersion reflected light characteristic map and the fluorescence characteristic map to obtain a fused characteristic map.

[0081] In the embodiment of the present application, fusing the oil immersion reflected light characteristic map and the fluorescence characteristic map means splicing the two characteristic maps. The specific splicing process corresponds to Figure 4 The "Sticking" module in .

[0082] S640: Identify the fused feature map to obtain the category of coal rock microscopic components.

[0083] In an embodiment of the present application, the fused feature map is identified by the coal rock microscopic component recognition model to obtain the coal rock microscopic image segmentation result. The image segmentation result is divided into 4 parts to display the areas and categories of each coal rock microscopic component, including: background resin, exinite, vitrinite and inertinite, which are displayed in the figure according to different colors.

[0084] Based on the above algorithm, a coal rock microscopic component analysis software can be developed, which integrates multiple algorithms. Users can obtain image segmentation results by submitting coal rock microscopic images (coalization degree R0 <1.0%).

[0085] See Table 2 for the evaluation index results of clustering algorithm and semantic segmentation network under single modality.

[0086] Table 2

[0087]

[0088] See Table 3 for a performance comparison of the dual-modal idea and the MultiScale-VGG structure.

[0089] Table 3

[0090]

[0091]

[0092] The performance and accuracy of the dual-modal MultiScale-VGG pruned version codec network model of the present invention are better than other models.

[0093] The present invention discloses a method for identifying coal rock microscopic components, the method comprising: obtaining a coal rock microscopic image, wherein the coal rock microscopic image includes a microscopic image under oil immersion reflected light and a microscopic image under fluorescence; performing feature extraction on the microscopic image under oil immersion reflected light based on a coal rock microscopic component identification model to obtain an oil immersion reflected light feature map, and performing feature extraction on the microscopic image under fluorescence based on a coal rock microscopic component identification model to obtain a fluorescence feature map; fusing the oil immersion reflected light feature map and the fluorescence feature map to obtain a fused feature map; and identifying the fused feature map to obtain the coal rock microscopic component category. The technical solution of the present invention fuses the microscopic images of the coal rock microscopic components under fluorescence and oil immersion reflected light through a dual-modal model, utilizes the different responses of microscopic components to different light sources, extracts and fuses the microscopic images under the two light sources based on the coal rock microscopic component identification model, and identifies the fused feature map to obtain the coal rock microscopic component category, thereby achieving automatic identification of the coal rock microscopic components, with the advantages of high accuracy, strong robustness, and rapidity and simplicity.

[0094] Example 3

[0095] Figure 8 This is a schematic diagram of the structure of a device for identifying microscopic components of coal rocks provided in the third embodiment of the present invention. This embodiment is applicable to the identification of microscopic components of coal rocks. The device for identifying microscopic components of coal rocks can be implemented in the form of hardware and / or software. The device for identifying microscopic components of coal rocks can be configured in any electronic device with network communication function. Figure 8 As shown, the device includes:

[0096] The coal rock image acquisition module 810 is used to acquire a microscopic image of the coal rock; wherein the microscopic image of the coal rock includes a microscopic image under oil immersion reflected light and a microscopic image under fluorescence;

[0097] The coal rock image recognition module 820 is used to identify the coal rock microscopic image based on the coal rock microscopic component recognition model to determine the coal rock microscopic component category; wherein the coal rock microscopic component recognition model is composed of a multi-scale bimodal convolutional neural network structure.

[0098] Optionally, the coal rock image acquisition module 810 includes:

[0099] a coal rock microscopic image calibration unit, configured to calibrate the coal rock microscopic image to be processed and determine a coal rock microscopic label image to be processed; wherein the coal rock microscopic label image to be processed includes coal rock microscopic component regions and coal rock microscopic component categories;

[0100] The coal rock microscopic component identification unit is used to perform the coal rock microscopic component category identification task on the coal rock microscopic image to be processed based on the coal rock microscopic component identification model to be trained, including:

[0101] Based on the coal rock microscopic component recognition model to be trained, a coal rock microscopic component category recognition task is performed on the coal rock microscopic image to be processed and the coal rock microscopic label image to be processed.

[0102] Optionally, the coal rock image recognition module 820 includes:

[0103] a feature map extraction unit, configured to perform feature extraction on the microscopic image under oil immersion reflected light based on a coal rock microscopic component identification model to obtain an oil immersion reflected light feature map, and to perform feature extraction on the microscopic image under fluorescence to obtain a fluorescence feature map;

[0104] a feature map fusion unit, configured to fuse the oil immersion reflected light feature map and the fluorescence feature map to obtain a fused feature map;

[0105] The microscopic component identification unit is used to identify the fused feature map to obtain the microscopic component category of the coal rock.

[0106] Optionally, the coal rock image recognition module 820 may determine the coal rock microscopic component recognition model by:

[0107] A microscopic image acquisition unit for the coal rock to be processed, used for acquiring a microscopic image of the coal rock to be processed;

[0108] a coal rock microscopic component classification task execution unit, configured to execute the coal rock microscopic component classification task on the coal rock microscopic image to be processed based on the coal rock microscopic component classification model to be trained;

[0109] The coal rock micro-component identification model training and updating unit is used to adjust the coal rock micro-component identification model to be trained according to the coal rock micro-component category identification task to obtain a trained and updated coal rock micro-component identification model.

[0110] Optionally, the coal rock image recognition module 820 further includes:

[0111] a unit for determining a microscopic label image of the coal rock to be processed, configured to calibrate the microscopic image of the coal rock to be processed and determine the microscopic label image of the coal rock to be processed; wherein the microscopic label image of the coal rock to be processed includes the microscopic component region and the microscopic component category of the coal rock;

[0112] Correspondingly, the coal rock microscopic component classification identification task execution unit is specifically used to:

[0113] Based on the coal rock microscopic component recognition model to be trained, a coal rock microscopic component category recognition task is performed on the coal rock microscopic image to be processed and the coal rock microscopic label image to be processed.

[0114] Optional, coal rock micro-component identification model training and updating unit, specifically used for:

[0115] Determining a loss function value corresponding to the coal rock microscopic component classification task;

[0116] The network parameters in the coal rock micro-component identification model to be trained are adjusted according to the loss function value.

[0117] Optionally, the coal rock image recognition module 820 further includes:

[0118] An output result evaluation unit is used to evaluate the output result of the trained and updated coal rock micro-component identification model based on a predetermined evaluation method; wherein the evaluation method includes pixel accuracy, average pixel accuracy, intersection-over-union ratio and average intersection-over-union ratio.

[0119] The device for identifying the microscopic components of coal and rock provided in the embodiment of the present invention can execute the method for identifying the microscopic components of coal and rock provided in any embodiment of the present invention, and has the corresponding functional modules and beneficial effects of the execution method.

[0120] Example 4

[0121] Figure 9 A schematic diagram of the structure of an electronic device 10 that can be used to implement an embodiment of the present invention is shown. The electronic device is intended to represent various forms of digital computers, such as laptop computers, desktop computers, workstations, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. The electronic device can also represent various forms of mobile devices, such as personal digital processing, cellular phones, smart phones, wearable devices (such as helmets, glasses, watches, etc.) and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely examples and are not intended to limit the implementation of the present invention described and / or claimed herein.

[0122] like Figure 9 As shown, the electronic device 10 includes at least one processor 11 and a memory, such as a read-only memory (ROM) 12, a random access memory (RAM) 13, etc., which is communicatively connected to the at least one processor 11. The memory stores a computer program that can be executed by the at least one processor. The processor 11 can perform various appropriate actions and processes according to the computer program stored in the read-only memory (ROM) 12 or the computer program loaded from the storage unit 18 into the random access memory (RAM) 13. Various programs and data required for the operation of the electronic device 10 can also be stored in the RAM 13. The processor 11, ROM 12, and RAM 13 are connected to each other via a bus 14. An input / output (I / O) interface 15 is also connected to the bus 14.

[0123] Multiple components in the electronic device 10 are connected to the I / O interface 15, including an input unit 16, such as a keyboard, a mouse, etc.; an output unit 17, such as various types of displays, speakers, etc.; a storage unit 18, such as a magnetic disk, an optical disk, etc.; and a communication unit 19, such as a network card, a modem, a wireless communication transceiver, etc. The communication unit 19 allows the electronic device 10 to exchange information / data with other devices via a computer network such as the Internet and / or various telecommunication networks.

[0124] The processor 11 can be any general-purpose and / or specialized processing component with processing and computing capabilities. Some examples of the processor 11 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various specialized artificial intelligence (AI) computing chips, various processors running machine learning model algorithms, a digital signal processor (DSP), and any suitable processor, controller, microcontroller, etc. The processor 11 executes the various methods and processes described above, such as the method for identifying the microscopic components of coal and rock.

[0125] In some embodiments, the coal rock microscopic component identification method can be implemented as a computer program, which is tangibly contained in a computer-readable storage medium, such as storage unit 18. In some embodiments, part or all of the computer program can be loaded and / or installed on the electronic device 10 via ROM 12 and / or communication unit 19. When the computer program is loaded into RAM 13 and executed by processor 11, one or more steps of the coal rock microscopic component identification method described above can be performed. Alternatively, in other embodiments, processor 11 can be configured to execute the coal rock microscopic component identification method in any other appropriate manner (e.g., via firmware).

[0126] Various embodiments of the systems and techniques described herein can be implemented in digital electronic circuit systems, integrated circuit systems, field programmable gate arrays (FPGAs), application specific integrated circuits (ASICs), application specific standard products (ASSPs), system-on-chip systems (SOCs), programmable logic devices (CPLDs), computer hardware, firmware, software, and / or combinations thereof. These various embodiments can include being implemented in one or more computer programs that are executable and / or interpreted on a programmable system that includes at least one programmable processor, which can be a special purpose or general purpose programmable processor that can receive data and instructions from a storage system, at least one input device, and at least one output device, and transmit data and instructions to the storage system, the at least one input device, and the at least one output device.

[0127] Computer programs for implementing the methods of the present invention may be written in any combination of one or more programming languages. These computer programs may be provided to a processor of a general-purpose computer, a special-purpose computer, or other programmable data processing device, such that when the computer program is executed by the processor, the functions / operations specified in the flowcharts and / or block diagrams are implemented. The computer program may be executed entirely on the machine, partially on the machine, as a stand-alone software package, partially on the machine and partially on a remote machine, or entirely on a remote machine or server.

[0128] In the context of the present invention, computer-readable storage media can be tangible media that can contain or store a computer program for use with an instruction execution system, device or equipment or used in combination with an instruction execution system, device or equipment. Computer-readable storage media can include but are not limited to electronic, magnetic, optical, electromagnetic, infrared or semiconductor systems, devices or equipment, or any suitable combination of the foregoing. Alternatively, computer-readable storage media can be machine-readable signal media. More specific examples of machine-readable storage media can include electrical connections based on one or more lines, portable computer disks, hard disks, random access memories (RAM), read-only memories (ROM), erasable programmable read-only memories (EPROM or flash memory), optical fibers, portable compact disk read-only memories (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination of the foregoing.

[0129] To provide interaction with a user, the systems and techniques described herein can be implemented on an electronic device having: a display device (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor) for displaying information to the user; and a keyboard and pointing device (e.g., a mouse or trackball) through which the user can provide input to the electronic device. Other types of devices can also be used to provide interaction with the user; for example, the feedback provided to the user can be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user can be received in any form (including acoustic input, voice input, or tactile input).

[0130] The systems and techniques described herein can be implemented in a computing system that includes back-end components (e.g., as a data server), or a computing system that includes middleware components (e.g., an application server), or a computing system that includes front-end components (e.g., a user computer with a graphical user interface or web browser through which a user can interact with implementations of the systems and techniques described herein), or a computing system that includes any combination of such back-end components, middleware components, or front-end components. The components of the system can be interconnected by any form or medium of digital data communication (e.g., a communication network). Examples of communication networks include: a local area network (LAN), a wide area network (WAN), a blockchain network, and the Internet.

[0131] A computing system may include clients and servers. The clients and servers are typically remote from each other and typically interact via a communication network. This client-server relationship arises through computer programs running on the respective computers, creating a client-server relationship. The server may be a cloud server, also known as a cloud computing server or cloud host. This server is a hosting product within the cloud computing service ecosystem that addresses the management difficulties and limited scalability of traditional physical hosting and VPS services.

[0132] It should be understood that the various forms of the processes shown above can be used to reorder, add, or delete steps. For example, the steps described in the present invention can be performed in parallel, sequentially, or in a different order, as long as the desired results of the technical solution of the present invention can be achieved. This is not limited herein.

[0133] The above specific embodiments do not limit the scope of protection of the present invention. Those skilled in the art will appreciate that various modifications, combinations, sub-combinations, and substitutions may be made based on design requirements and other factors. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of the present invention are intended to be included within the scope of protection of the present invention.

Claims

1. A method for identifying microscopic components of coal rock, characterized in that: include: Acquire a microscopic image of the coal rock; wherein the microscopic image of the coal rock includes a microscopic image under oil-immersion reflected light and a microscopic image under fluorescence; The coal rock microscopic image is identified based on a coal rock microscopic component identification model to determine the category of the coal rock microscopic components; wherein the coal rock microscopic component identification model is composed of a multi-scale bimodal convolutional neural network structure.

2. The method according to claim 1, characterized in that Identifying the coal rock microscopic image based on the coal rock microscopic component identification model to determine the type of coal rock microscopic components includes: performing feature extraction on the microscopic image under oil immersion reflected light based on a coal rock microscopic component recognition model to obtain an oil immersion reflected light feature map, and performing feature extraction on the microscopic image under fluorescence to obtain a fluorescence feature map; fusing the oil-immersion reflected light characteristic map and the fluorescence characteristic map to obtain a fused characteristic map; The fused feature map is identified to obtain the category of coal rock microscopic components.

3. The method according to claim 1, characterized in that The process of determining the coal rock micro-component identification model includes: Obtaining microscopic images of the coal rock to be processed; Performing a coal rock microscopic component classification task on the coal rock microscopic image to be processed based on the coal rock microscopic component recognition model to be trained; According to the coal rock microscopic component category identification task, the coal rock microscopic component identification model to be trained is adjusted to obtain a trained and updated coal rock microscopic component identification model.

4. The method according to claim 3, characterized in that After obtaining the microscopic image of the coal rock to be processed, the method further includes: Calibrate the microscopic image of the coal rock to be processed to determine a microscopic label image of the coal rock to be processed; wherein the microscopic label image of the coal rock to be processed includes the microscopic component area of ​​the coal rock and the category of the microscopic component of the coal rock; Accordingly, the coal rock microscopic component classification task is performed on the coal rock microscopic image to be processed based on the coal rock microscopic component recognition model to be trained, including: Based on the coal rock microscopic component recognition model to be trained, a coal rock microscopic component category recognition task is performed on the coal rock microscopic image to be processed and the coal rock microscopic label image to be processed.

5. The method according to claim 3, characterized in that According to the coal rock microscopic component classification identification task, the coal rock microscopic component identification model to be trained is adjusted, including: Determining a loss function value corresponding to the coal rock microscopic component classification task; The network parameters in the coal rock micro-component identification model to be trained are adjusted according to the loss function value.

6. The method according to claim 3, characterized in that After obtaining the trained and updated coal rock micro-component identification model, the method further includes: The output result of the trained and updated coal rock micro-component identification model is evaluated based on a predetermined evaluation method; wherein the evaluation method includes pixel accuracy, average pixel accuracy, intersection-over-union ratio and average intersection-over-union ratio.

7. A device for identifying microscopic components of coal and rock, characterized in that: include: A coal rock image acquisition module, used to acquire a microscopic image of the coal rock; wherein the microscopic image of the coal rock includes a microscopic image under oil immersion reflected light and a microscopic image under fluorescence; The coal rock image recognition module is used to identify the coal rock microscopic image based on the coal rock microscopic component recognition model and determine the coal rock microscopic component category; wherein the coal rock microscopic component recognition model is composed of a multi-scale bimodal convolutional neural network structure.

8. The device according to claim 7, characterized in that The coal rock image recognition module is specifically used for: performing feature extraction on the microscopic image under oil immersion reflected light based on a coal rock microscopic component recognition model to obtain an oil immersion reflected light feature map, and performing feature extraction on the microscopic image under fluorescence to obtain a fluorescence feature map; fusing the oil-immersion reflected light characteristic map and the fluorescence characteristic map to obtain a fused characteristic map; The fused feature map is identified to obtain the category of coal rock microscopic components.

9. An electronic device, characterized in that: The electronic device comprises: at least one processor; and a memory communicatively connected to the at least one processor; wherein, The memory stores a computer program that can be executed by the at least one processor, and the computer program is executed by the at least one processor so that the at least one processor can execute a method for identifying microscopic components of coal rocks according to any one of claims 1-6.

10. A computer-readable storage medium, characterized in that The computer-readable storage medium stores computer instructions, and the computer instructions are used to enable a processor to implement a method for identifying microscopic components of coal rocks according to any one of claims 1 to 6 when executed.

Images