Wet coal detection system and method, coal gangue separation system and computer readable medium
By constructing a deep learning model based on computer vision, the problem of reduced recognition effect due to moisture in wet coal identification was solved, achieving accurate identification and separation of wet coal and reducing economic losses.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- CHINA COAL TECH & ENG GRP SHANGHAI
- Filing Date
- 2025-07-11
- Publication Date
- 2026-04-21
AI Technical Summary
Existing technologies for wet coal identification suffer from reduced identification effectiveness due to moisture content, making it difficult to separate wet coal from gangue and resulting in economic losses.
A deep learning model based on computer vision technology is used to identify the surface features of wet coal through raw coal image acquisition, preprocessing, multi-path feature extraction and attention mechanism, and a wet coal detection system is constructed to achieve accurate identification.
It improved the accuracy of wet coal identification, reduced the phenomenon of wet coal being mistakenly disposed of as gangue, and reduced economic losses.
Smart Images
Figure CN121904418A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of coal mine gangue sorting technology; specifically, this invention relates to a wet coal detection system and method, a coal gangue sorting system, and a computer-readable medium. Background Technology
[0002] In coal mine production, raw coal needs to be sorted to remove gangue. Currently, the mainstream sorting method is to use X-rays to irradiate the coal and identify the density of the target material by the signal intensity fed back from the bottom plate, thereby determining the category of the target material and achieving the sorting effect.
[0003] During the mining process, water spraying is used. Some raw coal has a lot of moisture on its surface, which reduces signal feedback. This makes the identification signals of wet coal the same as those of gangue or other impurities, reducing the identification effect and causing some wet coal to be discharged together with gangue. Summary of the Invention
[0004] In view of the above, the present invention provides a wet coal detection system and method, a coal gangue sorting system and a computer-readable medium, thereby solving or at least alleviating one or more of the above-mentioned problems and other problems existing in the prior art.
[0005] To achieve the aforementioned objective, a first aspect of the present invention provides a wet coal detection system for a coal gangue sorting system, wherein the wet coal detection system includes a wet coal identification model, the wet coal identification model comprising: A raw coal image acquisition module, wherein the raw coal image acquisition module is used to acquire an image of the raw coal to be detected; An image preprocessing module is used to preprocess the acquired image; A multi-path feature extraction module is used to extract surface features of raw coal through multiple scale channels. An attention mechanism module, which is used to amplify the differences in surface features of raw coal; The raw coal surface feature comparison module is used to compare the surface features of the raw coal with the surface features of wet coal in a standard wet coal image, and to identify wet coal based on the similarity between the surface features of the raw coal and the surface features of the wet coal.
[0006] In the wet coal detection system described above, optionally, the wet coal identification model is a trained YOLO model or a VGG model.
[0007] In the wet coal detection system described above, optionally, the image preprocessing module is configured to sharpen the micro-texture of the image by enhancing the local contrast of the brightness channel and applying the Laplacian operator according to the following formula: in For the original image, To sharpen the intensity, It is the Laplacian operator for images.
[0008] In the wet coal detection system described above, optionally, the sharpening intensity is a value between 0.3 and 0.8.
[0009] In the wet coal detection system described above, optionally, the multi-path feature extractor is configured to extract and sample and fuse shallow, mid-level, and deep features of the image using three types of convolutional kernels to obtain a feature map with wet coal surface features. The convolutional kernels include: The large-scale channel uses a 7×7 convolution kernel to extract shallow features of the wet coal surface in the first-size area of the image, and the response function is... ,in This is the weight matrix. This represents the convolution operation, where I is the original image and b is the bias vector. The mesoscale channel uses a 5×5 convolution kernel to extract mid-layer features of the wet coal surface in the second-size area of the image; and The micro-scale channel uses a 3×3 convolution kernel to extract deep features of the wet coal surface in the third-size area of the image. And wherein the first size area is larger than the second size area, and the second size area is larger than the third size area.
[0010] In the wet coal detection system described above, optionally, the first size area is greater than [missing information]. The area of the second dimension is The area of the third dimension is smaller than .
[0011] In the wet coal detection system described above, optionally, the attention mechanism module includes: Feature evaluation module, the feature evaluation module is used for evaluating the input feature map Perform local texture analysis. ,in The kernel is 3×3 and b is the bias vector. The texture variance of each spatial location is calculated through the local receptive field. The low texture variance region is identified as the water film covered region, and the high texture variance region is identified as the dry and rough region. A spatial weight generation module is used to convert the texture features obtained by the feature evaluation module into a spatial attention map. ,in The sigmoid function compresses the range of A to... Inside, This is a 1×1 convolution operation used for cross-channel feature fusion, and the output result is... satisfy: ; A feature recalibration module is used to fuse the spatial attention map with the original features. ,in Represents element-wise multiplication. This is the enhancement coefficient.
[0012] To achieve the aforementioned objectives, a second aspect of the present invention provides a method for detecting wet coal using a wet coal detection system as described in any one of the first aspects.
[0013] Optionally, in the method for testing wet coal as described above, the method includes: Step 1: The raw coal image acquisition module acquires an image of the raw coal to be detected; Step II: The image preprocessing module preprocesses the acquired image; Step III: The multi-path feature extraction module extracts surface features of raw coal through multiple scale channels; Step IV: The attention mechanism module amplifies the differences in surface features of raw coal; Step V: The raw coal feature comparison module compares the surface features of the raw coal with the surface features of wet coal in the standard wet coal image, and identifies the wet coal based on the similarity between the surface features of the raw coal and the surface features of the wet coal.
[0014] To achieve the foregoing objective, a third aspect of the present invention provides a computer-readable medium having a computer program stored thereon, wherein the computer program, when executed by a processing device, implements the steps of the method as described in any of the second aspects.
[0015] To achieve the aforementioned objective, a fourth aspect of the present invention provides a coal gangue sorting system, wherein the coal gangue sorting system includes a wet coal detection system as described in any one of the first aspects.
[0016] The wet coal detection system for coal gangue sorting system according to the present invention is based on visual recognition. By constructing a deep learning model based on computer vision technology and a large number of labeled wet coal images, it can learn the texture and deep features of the wet coal surface. The model can then quickly capture relevant feature information of wet coal during the recognition process and distinguish the moisture content of the wet coal surface, thus achieving accurate identification of wet coal and reducing economic losses caused by insufficient recognition capabilities. Attached Figure Description
[0017] The disclosure of this invention will become more apparent from the accompanying drawings. It should be understood that these drawings are for illustrative purposes only and are not intended to limit the scope of protection of this invention. In the drawings: Figure 1 A schematic block diagram of a wet coal identification model for a wet coal detection system in a coal gangue sorting system according to the present invention; and Figure 2 This is a schematic flowchart of a method for wet coal detection using a wet coal detection system for a coal gangue sorting system according to the present invention. Detailed Implementation
[0018] Referring to the accompanying drawings and specific embodiments, the structure, composition, features, and advantages of the wet coal detection system and method, coal gangue sorting system, and computer-readable medium of the present invention will be described by way of example below. However, all descriptions should not be construed as limiting the present invention in any way.
[0019] Furthermore, for any single technical feature described or implied in the embodiments mentioned herein, or any single technical feature shown or implied in the various figures, the present invention still allows for any combination or deletion of these technical features (or their equivalents) without any technical obstacle, and thus these further embodiments according to the present invention should also be considered within the scope of this description.
[0020] Furthermore, the terms "first," "second," and "third" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of technical features indicated. Thus, a feature defined as "first," "second," or "third" may explicitly or implicitly include at least one of those features.
[0021] Figure 1 This is a schematic block diagram of a wet coal identification model for a wet coal detection system in a coal gangue sorting system according to the present invention.
[0022] In some embodiments, the wet coal detection system may include an electronic device that can provide a wet coal identification model. The electronic device suitable for embodiments of the wet coal detection system of the present invention may be a field device in a coal mine, a remote terminal device, or even a computer device. These examples should not limit the functionality and scope of the embodiments of the present invention. The electronic device may include a processing device (e.g., a central processing unit, a graphics processing unit, etc.) that can perform various appropriate actions and processes according to a program stored in a storage device or a program loaded from the storage device into a random access memory.
[0023] As shown in the figure, the wet coal detection system can include a wet coal identification model. For example, the wet coal identification model can be a trained YOLO model or VGG model, and other models with similar training capabilities are not excluded. Images used for training can be acquired from a camera device, and may include wet coal images, forming a database. By feeding these datasets into the model for training, the model will adjust hyperparameters based on the characteristics of wet coal and through a loss function such as cross-entropy loss, eventually converging and completing the training process. For example, the training image data can reach hundreds of thousands or more. This wet coal identification model can be used to identify wet coal to avoid discharging it along with gangue. The wet coal identification model can include a raw coal image acquisition module, an image preprocessing module, a multi-path feature extraction module, an attention mechanism module, and a raw coal surface feature comparison module. The wet coal identification model identifies the acquired wet coal image data and obtains wet coal category information and wet coal location information. By passing the wet coal category information and wet coal location information to the next execution mechanism, the wet coal sorting can be completed. Therefore, in a further embodiment, the wet coal detection system may further include a wet coal classification module, a wet coal positioning module, etc., which can classify and locate wet coal based on wet coal category information and wet coal location information.
[0024] The raw coal image acquisition module can be used to acquire images of the raw coal to be inspected. For example, in some embodiments, the raw coal image acquisition module can be a camera device used to provide image data to the wet coal inspection system. In some embodiments, the camera device can be an explosion-proof camera specifically designed for use in coal mines. The raw coal image acquisition module can be directed towards the raw coal, for example, towards the raw coal on a belt conveyor, to capture images of the raw coal. The raw coal image acquisition module can communicate with an image preprocessing module, thereby capturing images of the raw coal and transmitting the images to the subsequent image preprocessing module. In addition, to capture more suitable images, light sources can be arranged according to specific circumstances, which can effectively and conveniently improve the shooting effect.
[0025] An image preprocessing module can be used to preprocess the acquired images. Preprocessing directly captured images is advantageous for facilitating subsequent recognition. Common image processing methods include, but are not limited to, sharpening, enhancement, segmentation, and classification. This preprocessing can be performed automatically within the module. Processed images have more prominent features, facilitating subsequent recognition and feature extraction. These features could be the moisture characteristics of the raw coal surface. For example, in one embodiment, the image preprocessing module can be configured to enhance the local contrast of the lightness channel (L channel) (e.g., in the LAB color space) and sharpen the micro-texture of the image using the Laplacian operator according to the following formula: in For the original image, To sharpen the intensity, It is the Laplacian operator (second-order differential operator) for the image. This preprocessing can significantly enhance the surface texture differences between wet and dry coal. In wet coal regions, water film coverage leads to smooth texture and a weak Laplacian response. In dry coal regions, rough surfaces produce a strong Laplace response. (Grayscale level). In an optional embodiment, the sharpening intensity can be a value between 0.3 and 0.8. Images preprocessed in this way are brighter, with clearer distinctions between features, allowing the raw coal to have more prominent characteristics and be easier to identify and extract.
[0026] The multi-path feature extraction module is used to extract surface features of raw coal through multiple scale channels. For example, in some embodiments, the multi-path feature extractor is configured with three convolutional kernels to extract and sample shallow, medium, and deep features of the image to different degrees, thereby obtaining a more complete feature map with the moisture characteristics of the wet coal surface. Each feature map learns to detect different specific patterns (such as edges, corners, textures, specific shapes, object parts, etc.) in the input data. This image recognition based on multi-channel hierarchical feature extraction can utilize convolutional operations to progressively extract, combine, and refine increasingly abstract and discriminative feature representations from multi-channel input (raw pixels or feature maps) at multiple levels from shallow to deep, ultimately for accurate image classification or recognition.
[0027] In embodiments of the present invention, the convolution kernel may include large-scale channels, medium-scale channels, and micro-scale channels. For example, the large-scale channel can use a 7×7 convolution kernel to extract shallow features of the wet coal surface of a first-size area in the image, i.e., using this convolution kernel to detect specular reflection areas, and the response function can be... ,in This is the weight matrix. Representing a convolution operation, I is the original image, and b is the bias vector, which are parameters that can be learned. This operation can capture shallow features of surface moisture in wet coal, such as those greater than [value missing]. The light spot; for example, the medium-sized channel can use a 5×5 convolution kernel to extract the mid-layer features of the wet coal surface in the second-sized area of the image, that is, for Feature extraction is performed on the continuous smooth water film coverage area to obtain mid-layer feature information; for example, the micro-channel can use a 3×3 convolution kernel to extract deep features of the wet coal surface in the third size area of the image, which extracts pore-level texture, and wherein the first size area is larger than the second size area, and the second size area is larger than the third size area. In a specific embodiment, the first size area can be larger than The second dimension area can be The third dimension area can be smaller than The three features are fused through bilinear upsampling to form a feature map with a resolution of 64×64. Preserve key spatial details.
[0028] The texture attention mechanism module amplifies the differences in surface features of raw coal. It is a spatially adaptive feature enhancement module specifically designed to amplify the unique surface feature differences of wet coal. This module includes a feature evaluation module, a spatial weight generation module, and a feature recalibration module. These are achieved through three consecutive operations: feature importance evaluation, spatial weight generation, and feature recalibration. It can be used for feature enhancement or feature selection, dynamically highlighting important features and suppressing relatively less important ones. The goal is to teach the model to focus on the most relevant and informative parts of the input data (in this case, the raw coal), rather than treating all input equally.
[0029] The feature evaluation module can be used for input feature maps Perform local texture analysis. ,in The kernel is 3×3, and b is the bias vector. The texture variance of each spatial location is calculated through the local receptive field (7×7 effective area), which represents its texture saliency. High-response regions are identified as low-texture-variance regions, i.e., water film-covered areas, while low-response regions are identified as high-texture-complexity regions, i.e., dry and rough areas. This feature evaluation module provides the basis for generating spatial weights, which is used to determine why a certain feature should be focused on.
[0030] The spatial weight generation module is used to transform the texture features obtained from the feature evaluation module into spatial attention maps. ,in The sigmoid function compresses the range of A to... Inside, This is a 1×1 convolution operation used for cross-channel feature fusion, and the output result is... satisfy: .
[0031] As can be seen, the spatial weight generation module generates a weight map in a spatial dimension based on the output of the feature evaluation module. Each value in this weight map represents the importance weight of the corresponding spatial location.
[0032] The feature recalibration module is used to fuse the spatial attention map with the original features. ,in Represents element-wise multiplication. This is the enhancement coefficient. The default value for this enhancement system is 0.8. Its principle is to enhance the uniform response of the texture by amplifying the water film area by 1.8 times, while maintaining the original feature values of the dry areas. The feature recalibration module, as the final step in the attention mechanism, uses the weight map generated by the spatial weighting module to perform weighting operations on the original input feature map (or its transformed form), amplifying important features and maintaining or weakening unimportant features, thus completing the dynamic selection and enhancement of features.
[0033] The raw coal feature comparison module compares the surface features of raw coal with the surface features of wet coal in a standard wet coal image, and identifies the wet coal based on the similarity between the two surface features. After identifying the wet coal, subsequent classification and localization processing are performed.
[0034] The design concept of this invention is to use a wet coal identification network that focuses on changes in surface characteristics caused by moisture, achieving accurate classification through texture analysis and reflection feature capture. The network design here is based on three physical phenomena, which can be used to identify the moisture characteristics of the wet coal surface: (1) Texture smoothing effect: Moisture fills the pores on the surface of the coal block, reducing the micro-roughness, which manifests as a decrease in local gray-scale variance, as shown in the following formula, where Represents local texture variance. ; (2) Enhanced specular reflection: The water film forms a specular reflection, producing a highlight area at a specific angle, as shown in the following formula, where It is a reverberatory furnace. For refractive index, ; (3) Edge blurring effect: Moisture diffusion weakens the contour gradient. The Frobenius norm of the image gradient. .
[0035] Compared to the method of detecting raw coal by radiation, the wet coal detection system of this invention uses visual recognition to detect and sort wet raw coal with a particle size of 300mm or larger.
[0036] Figure 2 This is a schematic flowchart illustrating a method for wet coal detection using the wet coal detection system for a coal gangue sorting system according to the present invention. In the figure, the upper part shows a camera device that provides wet coal image data. This device can train a deep learning model based on wet coal identification using the labeled image data. The deep learning model can then identify the wet coal image data and finally classify and locate the identified wet coal. The lower part of the figure shows marked wet coal. The deep learning model can be trained on these wet coal samples and learn the ability to identify wet coal features.
[0037] Combination Figure 1 and Figure 2 As described above, the method for detecting wet coal using the wet coal detection system in the foregoing embodiments of the present invention may include: Step I: A raw coal image acquisition module acquires an image of the raw coal to be detected; Step II: An image preprocessing module preprocesses the acquired image; Step III: A multi-path feature extraction module extracts surface features of the raw coal through multiple scale channels; Step IV: An attention mechanism module amplifies the differences in surface features of the raw coal; Step V: A raw coal feature comparison module compares the surface features of the raw coal with the surface features of the wet coal in a standard wet coal image, and identifies the wet coal based on the similarity between the surface features of the raw coal and the surface features of the wet coal. For details of each step, please refer to the description combined with the wet coal identification model. As can be seen from the above, the wet coal detection method according to the present invention is based on visual recognition. By constructing a deep learning model based on computer vision technology and a large number of labeled wet coal images, it can learn the texture and deep features of the wet coal surface. The model can then quickly capture relevant feature information of the wet coal during the identification process and distinguish the moisture content of the wet coal surface, achieving accurate identification of wet coal and reducing economic losses due to insufficient identification capabilities.
[0038] Another aspect of the present invention provides a computer-readable medium having a computer program stored thereon, wherein the computer program, when executed by a processing device, implements the steps of the method as described in any of the second aspects.
[0039] In particular, the processes described in the flowcharts above can be implemented as computer software programs. For example, embodiments of the present invention may include a computer program product comprising a computer program carried on a non-transitory computer-readable medium, the computer program containing program code for performing the methods shown in the flowcharts. In such embodiments, the computer program may be downloaded and installed from a network via a communication device, or installed from a storage device. When the computer program is executed by a processing device, it performs the functions defined in the methods of embodiments of the present invention.
[0040] It should be noted that the computer-readable medium described above in this invention can be a computer-readable signal medium, a computer-readable storage medium, or any combination thereof. A computer-readable storage medium can be, for example,—but not limited to—an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any combination thereof. More specific examples of a computer-readable storage medium may include, but are not limited to: an electrical connection having one or more wires, a portable computer disk, a hard disk, a random access memory, a read-only memory, an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination thereof. In this invention, a computer-readable storage medium can be any tangible medium containing or storing a program that can be used by or in conjunction with an instruction execution system, apparatus, or device. In this invention, a computer-readable signal medium can include a data signal propagated in baseband or as part of a carrier wave, carrying computer-readable program code. Such propagated data signals can take various forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination thereof. A computer-readable signal medium can be any computer-readable medium other than a computer-readable storage medium, which can send, propagate, or transmit a program for use by or in connection with an instruction execution system, apparatus, or device. The program code contained on the computer-readable medium can be transmitted using any suitable medium, including but not limited to: wires, optical fibers, RF (radio frequency), etc., or any suitable combination thereof.
[0041] The aforementioned computer-readable medium may be included in the aforementioned electronic device; or it may exist independently and not assembled into the electronic device.
[0042] The aforementioned computer-readable medium carries one or more programs. When the aforementioned one or more programs are executed by the electronic device, the electronic device: uses a visual recognition-based detection method, employs a deep learning model based on computer vision technology and a large number of labeled wet coal images to learn the texture and deep features of the wet coal surface, and uses the model to quickly capture relevant feature information of wet coal during the recognition process, distinguish the moisture content of the wet coal surface, thereby accurately identifying wet coal and reducing economic losses caused by insufficient recognition capabilities.
[0043] The modules described in the embodiments of the present invention can be implemented in software or hardware. The names of the modules are not, in some cases, limiting the scope of the module itself.
[0044] Furthermore, one aspect of the present invention also provides a coal gangue sorting system. It is understood that a coal gangue sorting system is an equipment system used to effectively separate raw coal from gangue generated during coal mining. The coal gangue sorting system of the present invention may include a wet coal detection system as described in any of the foregoing embodiments, which uses a computer vision recognition and deep learning model system to accurately detect wet coal mixed in with gangue, preventing wet coal from being misidentified as gangue and discharged, thus reducing economic losses due to insufficient identification capabilities.
[0045] The technical scope of this invention is not limited to the contents of the above specification. Those skilled in the art can make various modifications and variations to the above embodiments without departing from the technical concept of this invention, and all such modifications and variations should fall within the scope of this invention.
Claims
1. A wet coal detection system for a coal gangue sorting system, characterized in that, The wet coal detection system includes a wet coal identification model, which includes: A raw coal image acquisition module, wherein the raw coal image acquisition module is used to acquire an image of the raw coal to be detected; An image preprocessing module is used to preprocess the acquired image; A multi-path feature extraction module is used to extract surface features of raw coal through multiple scale channels. An attention mechanism module, which is used to amplify the differences in surface features of raw coal; The raw coal surface feature comparison module is used to compare the surface features of the raw coal with the surface features of wet coal in a standard wet coal image, and to identify wet coal based on the similarity between the surface features of the raw coal and the surface features of the wet coal.
2. The wet coal detection system as described in claim 1, characterized in that, The wet coal identification model is a trained YOLO model or VGG model.
3. The wet coal detection system as described in claim 1, characterized in that, The image preprocessing module is configured to sharpen the micro-texture of the image by enhancing the local contrast of the luminance channel and applying the Laplacian operator using the following formula: in For the original image, To sharpen the intensity, It is the Laplacian operator for images.
4. The wet coal detection system as described in claim 4, characterized in that, The sharpening intensity is a value between 0.3 and 0.
8.
5. The wet coal detection system as described in claim 1, characterized in that, The multi-path feature extractor is configured to extract and sample and fuse shallow, mid-level, and deep features of the image using three types of convolutional kernels to obtain a feature map with wet coal surface features. The convolutional kernels include: The large-scale channel uses a 7×7 convolution kernel to extract shallow features of the wet coal surface in the first-size area of the image, and the response function is... ,in This is the weight matrix. This represents the convolution operation, where I is the original image and b is the bias vector. The mesoscale channel uses a 5×5 convolution kernel to extract mid-layer features of the wet coal surface in the second-size area of the image; and The micro-scale channel uses a 3×3 convolution kernel to extract deep features of the wet coal surface in the third-size area of the image. And wherein the first size area is larger than the second size area, and the second size area is larger than the third size area.
6. The wet coal detection system as described in claim 5, characterized in that, The area of the first dimension is greater than The area of the second dimension is The area of the third dimension is smaller than .
7. The wet coal detection system as described in claim 1, characterized in that, The attention mechanism module includes: Feature evaluation module, the feature evaluation module is used for evaluating the input feature map Perform local texture analysis. ,in The kernel is 3×3 and b is the bias vector. The texture variance of each spatial location is calculated through the local receptive field. The low texture variance region is identified as the water film covered region, and the high texture variance region is identified as the dry and rough region. A spatial weight generation module is used to convert the texture features obtained by the feature evaluation module into a spatial attention map. ,in The sigmoid function compresses the range of A to... Inside, This is a 1×1 convolution operation used for cross-channel feature fusion, and the output result is... satisfy: ； A feature recalibration module is used to fuse the spatial attention map with the original features. ,in Represents element-wise multiplication. This is the enhancement coefficient.
8. A method for detecting wet coal using the wet coal detection system as described in any one of claims 1 to 7.
9. The method for detecting wet coal as described in claim 8, characterized in that, The method includes: Step 1: The raw coal image acquisition module acquires an image of the raw coal to be detected; Step II: The image preprocessing module preprocesses the acquired image; Step III: The multi-path feature extraction module extracts surface features of raw coal through multiple scale channels; Step IV: The attention mechanism module amplifies the differences in surface features of raw coal; Step V: The raw coal feature comparison module compares the surface features of the raw coal with the surface features of wet coal in the standard wet coal image, and identifies the wet coal based on the similarity between the surface features of the raw coal and the surface features of the wet coal.
10. A computer-readable medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processing device, it implements the steps of the method as described in claim 8 or 9.
11. A coal gangue sorting system, characterized in that, The coal gangue sorting system includes a wet coal detection system as described in any one of claims 1 to 7.