CN111881909A - Coal and gangue identification method and device, electronic equipment and storage medium - Google Patents

Coal and gangue identification method and device, electronic equipment and storage medium Download PDF

Info

Publication number
CN111881909A
CN111881909A CN202010729919.XA CN202010729919A CN111881909A CN 111881909 A CN111881909 A CN 111881909A CN 202010729919 A CN202010729919 A CN 202010729919A CN 111881909 A CN111881909 A CN 111881909A
Authority
CN
China
Prior art keywords
gangue
coal
target image
identification
image
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Pending
Application number
CN202010729919.XA
Other languages
Chinese (zh)
Inventor
赵存会
员娇娇
朱晓宁
吴喆峰
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Jingying Digital Technology Co Ltd
Original Assignee
Jingying Digital Technology Co Ltd
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Jingying Digital Technology Co Ltd filed Critical Jingying Digital Technology Co Ltd
Priority to CN202010729919.XA priority Critical patent/CN111881909A/en
Publication of CN111881909A publication Critical patent/CN111881909A/en
Pending legal-status Critical Current

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00—Image analysis
    • G06T7/10—Segmentation; Edge detection
    • G06T7/13—Edge detection
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F18/00—Pattern recognition
    • G06F18/20—Analysing
    • G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
    • G06F18/214—Generating training patterns; Bootstrap methods, e.g. bagging or boosting
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F18/00—Pattern recognition
    • G06F18/20—Analysing
    • G06F18/24—Classification techniques
    • G06F18/241—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches
    • G06F18/2411—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches based on the proximity to a decision surface, e.g. support vector machines
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00—Computing arrangements based on biological models
    • G06N3/02—Neural networks
    • G06N3/04—Architecture, e.g. interconnection topology
    • G06N3/045—Combinations of networks
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T5/00—Image enhancement or restoration
    • G06T5/50—Image enhancement or restoration by the use of more than one image, e.g. averaging, subtraction
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00—Image analysis
    • G06T7/40—Analysis of texture
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00—Image analysis
    • G06T7/90—Determination of colour characteristics
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V10/00—Arrangements for image or video recognition or understanding
    • G06V10/20—Image preprocessing
    • G06V10/24—Aligning, centring, orientation detection or correction of the image
    • G06V10/245—Aligning, centring, orientation detection or correction of the image by locating a pattern; Special marks for positioning

Abstract

The method comprises the steps of obtaining a target image of the coal and gangue to be identified, inputting the target image of the coal and gangue to be identified into a preset identification model when identifying the coal and gangue, obtaining at least two identification results of the target image, and fusing the at least two identification results of the target image according to a preset rule to obtain an identification result of the coal and gangue to be identified, wherein the identification result of the coal and gangue to be identified comprises the category of coal blocks and/or gangue in the target image. According to the method and the device, the target image of the coal and gangue to be identified is input into the preset identification model, the category of the coal block and/or the gangue in the target image is obtained, the automation of coal and gangue identification is realized, and the coal and gangue identification accuracy and the working efficiency are improved.

Description

Coal and gangue identification method and device, electronic equipment and storage medium
Technical Field
The application relates to the technical field of intelligent identification and automatic control, in particular to a coal and gangue identification method and device, electronic equipment and a storage medium.
Background
In recent years, with the rapid development of the national industrial industry, the supply and demand of coal have been extremely large. However, raw coal is prone to incorporation of various mineral impurities during its original formation, as well as inevitable incorporation of various rocks and other impurities during mining and transportation. In order to reduce various impurities in raw coal, divide the raw coal into various excellent products, meet the requirements of different users and reduce the pollution of the coal to the atmosphere, the coal needs to be processed. The gangue is solid waste with low carbon content and high ash content in the coal mining and processing process, along with the improvement of the coal mining mechanization degree, the content of the gangue in raw coal is increased, the coal gangue separation is used as a combustion pretreatment technology, is the premise of the optimized utilization of the coal, and can control the difficulty and the cost of the pollution in the later period to the minimum degree. Therefore, the coal and gangue identification is an essential process for coal and gangue separation in the coal mine production process, and the removal of gangue in raw coal is the basis for producing clean energy by coal; meanwhile, the washing cost can be reduced by sorting the coal gangue, the grade of finished coal is improved, and the economic benefit of the coal enterprises is improved. Therefore, coal gangue identification is one of the important links of coal production.
In the related technology, the traditional coal and gangue identification is mainly completed manually, and the efficiency is low. In recent years, with the improvement of industrialization and intellectualization, how to improve the accuracy and the working efficiency of coal and gangue identification so as to improve the economic benefit of coal enterprises has become the key point of current scientific and technological research.
Disclosure of Invention
In order to overcome the problems in the related art, the application provides a coal and gangue identification method, a coal and gangue identification device, electronic equipment and a storage medium, which can improve the coal and gangue identification accuracy and the working efficiency, thereby improving the economic benefit of a coal enterprise.
The application provides a coal gangue identification method in a first aspect, which comprises the following steps:
acquiring a target image of coal and gangue to be identified, wherein the coal and gangue to be identified at least comprises coal blocks and/or gangue;
inputting the target image into a preset identification model to obtain at least two identification results of the target image;
and fusing at least two recognition results of the target image according to a preset rule to obtain a recognition result of the coal and gangue to be recognized, wherein the recognition result of the coal and gangue to be recognized comprises the category of the coal block and/or the gangue in the target image.
Further, the fusion processing is performed on at least two recognition results of the target image according to a preset rule to obtain the recognition result of the coal gangue to be recognized, and the method specifically comprises the following steps:
and performing linear weighting processing on at least two recognition results of the target image to obtain the recognition result of the coal gangue to be recognized.
Further, the target image includes: the first image of the coal gangue to be identified under the dual-energy X-ray and the second image obtained through the image acquisition device, wherein the preset identification model comprises: the first recognition model and the second recognition model input the target image into a preset recognition model to obtain at least two recognition results of the target image, specifically:
inputting the first image into the first recognition model to obtain a first recognition result of the target image, wherein the first recognition model is obtained by modeling according to the characteristics of the coal block and/or the gangue on the physical structure, which are obtained by the characteristic extraction sensor;
and inputting the second image into the second recognition model to obtain a second recognition result of the target image, wherein the second recognition model is obtained by modeling according to the characteristics of the coal block and/or the gangue on texture and/or color, which are obtained by the characteristic extraction sensor.
Further, the preset recognition model further includes: and the third identification model is used for inputting the target image into a preset identification model to obtain an identification result of the target image, and specifically comprises the following steps:
and inputting the second image into the third identification model to obtain a third identification result of the target image, wherein the third identification model is obtained by modeling the characteristics of the coal block and/or the gangue on multiple scales and multiple dimensions, which are obtained by the characteristic extraction sensor.
Further, the method further comprises: and determining the position information of the coal block and/or the gangue in the target image.
Further, the method further comprises:
and sending the identification result and the position information of the coal block and/or the gangue in the target image to a coal and gangue separation system, so that the coal and gangue separation system performs coal and gangue separation on the coal and gangue to be separated according to the identification result and the position information of the coal block and/or the gangue in the target image.
Further, the determining of the position information of the coal block and/or the gangue in the target image specifically includes:
detecting the outline of the coal block and/or the gangue by adopting a Canny edge detection method, and determining the position area of the coal block and/or the gangue in the target image;
obtaining the centroid of the position area by using a centroid method, and determining a centroid coordinate;
and taking the centroid coordinate as the position information of the coal block and/or the gangue in the target image.
The second aspect of the present application provides a coal gangue identification device, including:
the device comprises an acquisition unit, a processing unit and a processing unit, wherein the acquisition unit is used for acquiring a target image of coal and gangue to be identified, and the coal and gangue to be identified at least comprises coal blocks and/or gangue;
the identification unit is used for inputting the target image into a preset identification model to obtain at least two identification results of the target image;
and the fusion unit is used for performing fusion processing on at least two recognition results of the target image according to a preset rule to obtain a recognition result of the coal and gangue to be recognized, wherein the recognition result of the coal and gangue to be recognized comprises the type of the coal block and/or the gangue in the target image.
A third aspect of the present application provides an electronic device comprising:
a processor; and
a memory having executable code stored thereon, which when executed by the processor, causes the processor to perform the method of identifying mine refuse as described above.
A fourth aspect of the present application provides a non-transitory machine-readable storage medium having stored thereon executable code, which when executed by a processor of an electronic device, causes the processor to perform a method of identifying mine refuse as described above.
The technical scheme provided by the application can comprise the following beneficial effects:
according to the method, when the coal and gangue are identified, a target image of the coal and gangue to be identified is obtained, the target image of the coal and gangue to be identified is input into a preset identification model, at least two identification results of the target image are obtained, the at least two identification results of the target image are subjected to fusion processing according to preset rules, and the identification result of the coal and gangue to be identified is obtained and contains the category of coal blocks and/or gangue in the target image. The method and the device have the advantages that the target image of the coal and gangue to be identified is input into the preset identification model, the category of the coal block and/or the gangue in the target image is obtained, the automation of coal and gangue identification is realized, the coal and gangue identification accuracy and the working efficiency are improved, and the economic benefit of a coal enterprise is further improved.
It is to be understood that both the foregoing general description and the following detailed description are exemplary and explanatory only and are not restrictive of the application.
Drawings
The foregoing and other objects, features and advantages of the application will be apparent from the following more particular descriptions of exemplary embodiments of the application, as illustrated in the accompanying drawings wherein like reference numbers generally represent like parts throughout the exemplary embodiments of the application.
FIG. 1 is a schematic structural diagram of a gangue control system shown in an embodiment of the application;
FIG. 2 is a schematic flow chart of a coal gangue identification method according to an embodiment of the application;
FIG. 3 is a schematic flow chart illustrating a method for determining position information of the coal block and/or the gangue in the target image according to the embodiment of the present application;
fig. 4 is a flowchart illustrating a method of inputting a target image into a preset recognition model to obtain a recognition result of the target image, where the recognition result of the target image includes a specific implementation of a category of a coal block and/or gangue in the target image according to an embodiment of the present application;
FIG. 5 is a dual energy X-ray imaging of coal refuse as shown in an embodiment of the present application;
FIG. 6 is a flowchart illustrating a method for constructing a specific implementation of a first recognition model according to an embodiment of the present disclosure;
FIG. 7 is a flowchart illustrating a method for constructing a specific implementation of a second recognition model according to an embodiment of the present application;
FIG. 8 is a schematic flow chart diagram illustrating one embodiment of a pre-treatment process according to an embodiment of the present disclosure;
FIG. 9 is a graph of predicted results of a pre-processing procedure as shown in an embodiment of the present application;
FIG. 10 is a flowchart illustrating a method for constructing a specific implementation of a third recognition model according to an embodiment of the present application;
FIG. 11 is a diagram of a feature fusion framework based on the Attention mechanism according to an embodiment of the present application;
FIG. 12 is a schematic structural diagram of a gangue identification device shown in an embodiment of the application;
fig. 13 is a schematic structural diagram of an electronic device shown in an embodiment of the present application.
Detailed Description
Preferred embodiments of the present application will be described in more detail below with reference to the accompanying drawings. While the preferred embodiments of the present application are shown in the drawings, it should be understood that the present application may be embodied in various forms and should not be limited to the embodiments set forth herein. Rather, these embodiments are provided so that this disclosure will be thorough and complete, and will fully convey the scope of the disclosure to those skilled in the art.
The terminology used herein is for the purpose of describing particular embodiments only and is not intended to be limiting of the application. As used in this application and the appended claims, the singular forms "a", "an", and "the" are intended to include the plural forms as well, unless the context clearly indicates otherwise. It should also be understood that the term "and/or" as used herein refers to and encompasses any and all possible combinations of one or more of the associated listed items.
It should be understood that although the terms "first," "second," "third," etc. may be used herein to describe various information, these information should not be limited to these terms. These terms are only used to distinguish one type of information from another. For example, first information may also be referred to as second information, and similarly, second information may also be referred to as first information, without departing from the scope of the present application. Thus, a feature defined as "first" or "second" may explicitly or implicitly include one or more of that feature. In the description of the present application, "a plurality" means two or more unless specifically limited otherwise.
The application provides a coal and gangue identification method, a coal and gangue identification device, an electronic device and a storage medium, which are applied to a coal and gangue control system shown in figure 1 in an operation scene of a belt, wherein the coal and gangue control system comprises: the coal and gangue identification system 10 and the coal and gangue separation system 30 are used for identifying and positioning coal and gangue in a coal and gangue mixture carried on a running belt by taking the coal and gangue as a target object through the coal and gangue identification system 10, determining the category and position information of the target object, sending the output result of the coal and gangue identification system 10 to the coal and gangue separation system 20, and realizing automatic separation of the coal and gangue according to the category and position information of the coal and gangue and/or gangue in the target image by the coal and gangue separation system 20, wherein the output result of the coal and gangue identification system 10 comprises the coordinates of the center point of the category and position information image of the coal and gangue in the target pair image.
The technical solutions of the embodiments of the present application are described in detail below with reference to the accompanying drawings.
Fig. 2 is a schematic flow chart of a coal gangue identification method according to an embodiment of the present application.
Referring to fig. 2, an embodiment of the present application provides a coal gangue identification method, which specifically includes the following steps:
s201: and acquiring a target image of the coal gangue to be identified, wherein the coal gangue to be identified at least comprises coal blocks and/or gangue.
The target image includes: the method comprises the steps that a first image of the coal gangue to be identified under the dual-energy X-ray and a second image obtained through an image acquisition device participate in subsequent identification of the coal gangue to be identified by taking the first image and the second image as target images. In the embodiment of the application, in the operation scene of the belt, the irradiation of the dual-energy X-ray is carried out on the coal blocks and the gangue in the coal gangue mixture carried on the belt in operation, and then the first image under the dual-energy X-ray can be obtained. And in the running process of the belt, monitoring equipment is used for monitoring in real time all the time, and when coal and gangue are separated, a second image can be obtained by acquiring a monitoring video corresponding to the coal and gangue to be separated and monitored and intercepting a video image of the coal and gangue to be separated from the monitoring video. In a specific embodiment, the monitoring device may be configured to obtain the second image by a camera.
In the embodiment of the application, a preset identification model needs to be constructed in the coal and gangue identification system 10 in advance, and the preset identification model is obtained by modeling according to the characteristics of the coal blocks and/or the gangue in physical structure, texture and/or color, which are obtained by the characteristic extraction sensor. It should be noted that the feature extraction sensor mainly combines sensors such as an X-ray sensor and a camera to obtain features of the coal briquette and/or the gangue on physical structure, texture and/or color, and respectively establishes corresponding models, thereby realizing automatic identification of the coal briquette and/or the gangue.
S202: and inputting the target image into a preset identification model to obtain at least two identification results of the target image.
In an embodiment of the application, the preset identification model includes: the first recognition model and the second recognition model are used for inputting the target image into a preset recognition model to obtain at least two recognition results of the target image, and the method specifically comprises the following steps:
and inputting the first image into a first recognition model to obtain a first recognition result of the target image, wherein the first recognition model is obtained by modeling according to the characteristics of the coal blocks and/or the gangue on the physical structure, which are obtained by the characteristic extraction sensor.
And inputting the second image into a second recognition model to obtain a second recognition result of the target image, wherein the second recognition model is obtained by modeling according to the characteristics of the coal blocks and/or the gangue on the texture and/or the color, which are obtained by the characteristic extraction sensor.
Further, the preset identification model further includes: and the third identification model is used for inputting the target image into the preset identification model to obtain an identification result of the target image, and specifically comprises the following steps:
and inputting the second image into the third identification model to obtain a third identification result of the target image, wherein the third identification model is obtained by modeling the characteristics of the coal block and/or the gangue on multiple scales and multiple dimensions, which are obtained by the characteristic extraction sensor.
It should be noted that, in the embodiment of the present application, the preset identification model may further include other types of identification models, and model training is performed through other dimensions or other features, so that identification of the coal block and/or the gangue can be achieved, which is not described in detail herein.
In the embodiment of the application, different attributes of the coal briquette and the gangue are respectively obtained by using various sensors, and models are respectively constructed for the coal briquette and the gangue by using technologies such as a neural network and deep learning, so that the coal briquette and/or the gangue are identified, and a corresponding identification result of a target image is obtained, wherein the identification result of the target image comprises the category of the coal briquette and/or the gangue.
S203: and fusing the recognition results of the target image according to a preset rule to obtain a recognition result of the coal and gangue to be recognized, wherein the recognition result of the coal and gangue to be recognized comprises the category of the coal blocks and/or the gangue in the target image.
It should be noted that, in order to obtain a recognition model with good performance and meeting the requirements of practical applications, the embodiment of the present application provides a recognition strategy of multi-model fusion based on the above multiple recognition models, specifically, training is performed on each recognition model based on the acquired data, and recognition and classification are performed to obtain corresponding recognition results, and finally, the recognition results of each model are fused to obtain a final recognition result of the coal and gangue to be recognized.
In the embodiment of the application, the preset rule performs fusion processing on the recognition result of the target image to obtain the recognition result of the coal and gangue to be recognized, and specifically may be:
and carrying out linear weighting processing on the recognition result of the target image to obtain the recognition result of the coal and gangue to be recognized.
It should be noted that, since it is not clear which recognition model obtains a more reliable recognition result, in the process of the linear weighting, the weights of the models may all be set to 1, and further, if it can be clearly determined which recognition model obtains a more reliable recognition result, the weights may be appropriately increased, and the setting of the specific weights may be adaptively adjusted according to actual requirements, which is not described in detail herein.
In the embodiment of the present application, the above-mentioned multiple recognition results may be fused in other manners, for example, recognition probability values of the recognition results may be calculated respectively, and which recognition result is determined as the final recognition result according to the calculated recognition probability values.
It should be noted that the above fusion processing can be implemented in various ways, and specific implementation manners thereof may refer to technical means commonly used in the related art, and are not described in detail herein.
According to the coal and gangue identification method provided by the embodiment of the application, when coal and gangue identification is carried out, a target image of coal and gangue to be identified is obtained, at least two identification results of the target image are obtained by inputting the target image of the coal and gangue to be identified into a preset identification model, at least two identification results of the target image are subjected to fusion processing according to a preset rule, and the identification result of the coal and gangue to be identified is obtained, wherein the identification result of the coal and gangue to be identified comprises the category of coal blocks and/or gangue in the target image. According to the method and the device, the target image of the coal and gangue to be identified is input into the preset identification model, the category of the coal block and/or the gangue in the target image is obtained, the automation of coal and gangue identification is realized, the coal and gangue identification accuracy and the working efficiency are improved, and further the economic benefit of a coal enterprise is improved.
After the identification result of the target image is obtained, the coal gangue and/or gangue needs to be positioned so as to facilitate the subsequent separation step. Thus, on the basis of fig. 2, the method comprises:
and inputting the second image into a second recognition model to obtain the position information of the coal block and/or the gangue in the target image.
And inputting the second image into the second recognition model, and performing a positioning processing process of the position information of the coal block and/or the gangue in the target image in the second recognition model.
As shown in fig. 3, in the embodiment of the present application, determining the position information of the coal block and/or the gangue in the target image specifically includes the following steps:
s301: and detecting the outline of the coal block and/or the gangue by adopting a Canny edge detection method, and determining the position area of the coal block and/or the gangue in the target image.
S302: and obtaining the centroid of the position area by using a centroid method, and determining the coordinates of the centroid.
S303: and taking the mass center coordinate as the position information of the coal block and/or the gangue in the target image.
After the identification result of the coal block and/or the gangue is obtained, the coal block and/or the gangue needs to be positioned so as to facilitate the subsequent separation step. Firstly, the contour of the coal block and/or the gangue is obtained by adopting Canny edge detection, and the position area of the coal block and/or the gangue in the target image is determined. Then, the centroid method is used to obtain the centroid of the location area and determine the centroid coordinates. And finally, taking the centroid coordinates as the position information of the coal blocks and/or the gangue in the target image. Thus, the coal blocks and/or the gangue are positioned in the target image.
And sending the identification result of the coal and gangue to be identified and the position information of the coal block and/or the gangue in the target image to a coal and gangue separation system, so that the coal and gangue separation system performs coal and gangue separation on the coal and gangue to be separated according to the identification result of the coal and gangue to be identified and the position information of the coal block and/or the gangue in the target image. .
In the embodiment of the application, the to-be-identified gangue identification result (i.e., the category of the coal block and/or gangue in the target image) of the gangue identification system 10 and the position information of the coal block and/or gangue in the target image are sent to the gangue separation system 20, so that the gangue separation system 20 separates the coal blocks and/or gangue of different categories to different tracks according to the to-be-identified gangue identification result and the position information of the coal block and/or gangue in the target image, thereby realizing the automatic separation of the coal block and/or gangue.
It should be noted that the coal and gangue separation system 20 may be configured as an intelligent manipulator, and sort the coal blocks and/or the gangue according to the position information of the coal blocks and/or the gangue in the target image, so as to realize automatic separation of the coal blocks and the gangue.
In an embodiment of the present application, the above-mentioned preset recognition model includes: the first recognition model, the second recognition model and the third recognition model are taken as examples to specifically explain:
fig. 4 is a flowchart of a method of inputting a target image into a preset recognition model to obtain a recognition result of the target image, where the recognition result of the target image includes a specific implementation manner of categories of coal blocks and/or gangue in the target image.
Referring to fig. 4, in the embodiment of the present application, the inputting of the target image into the preset identification model to obtain the identification result of the target image, where the identification result of the target image includes the category of the coal block and/or the gangue in the target image specifically includes the following steps:
s401: and inputting the first image into a first recognition model, and performing coal and gangue recognition on the coal blocks and/or the gangue to obtain a first recognition result, wherein the first recognition result comprises the categories of the coal blocks and/or the gangue in the target image.
In the embodiment of the application, the first identification model is obtained by modeling according to the characteristics of the coal blocks and/or the gangue on the physical structure, which are obtained by the characteristic extraction sensor. As shown in fig. 5, for the dual-energy X-ray image of the coal block and the gangue acquired by the X-ray, since the coal block and the gangue are attenuated differently under the irradiation of the X-ray, finally, the imaging result has a certain statistical difference in gray scale. In order to better represent the difference and distinguish the coal blocks from the gangue, the embodiment of the application acquires a large number of X-ray images of the coal blocks and the gangue continuously, and obtains the gray level distribution histogram of the coal blocks and the gangue through the processing of the correlation function relationship. Theoretically, the coal blocks and the gangue show obvious statistical difference on the gray level distribution histogram, the coal blocks and/or the gangue can be effectively classified and recognized through setting a reasonable threshold value, a first recognition result is obtained, the first recognition result comprises the categories of the coal blocks and/or the gangue in the target image, and the categories of the coal blocks and/or the gangue in the target image are used for identifying the categories of the coal blocks and/or the gangue in the target image.
S402: and inputting the second image into a second recognition model, and performing coal and gangue recognition on the coal blocks and/or the gangue to obtain a second recognition result, wherein the second recognition result comprises the categories of the coal blocks and/or the gangue in the target image.
In the embodiment of the application, the classification and identification of the coal briquettes and the gangue mainly utilizes the difference of the coal briquettes and the gangue in physical structures by using the X-ray, but the accuracy of the identification is difficult to meet the actual application requirements by simply relying on the information acquired by the sensor, so that the embodiment of the application designs a second identification model, that is, the embodiment of the application utilizes the visual difference of the coal briquettes and the gangue to perform classification and identification to obtain a second identification result, wherein the second identification result comprises the category of the coal briquettes and/or the gangue in a target image, and the category of the coal briquettes and/or the gangue in the target image is used for identifying the category of the coal briquettes and/or the gangue in the target image.
It should be noted that the second identification model mainly uses the gray scale and texture features of the coal blocks and the gangue for classification and identification. In the actual identification process, the coal block is usually darker in color, while the gangue is lighter in color, so that the gray distribution and peak value of the coal block and the gangue are different, and therefore, the gray scale and texture analysis can be helpful for identifying the coal block and the gangue.
S403: and inputting the second image into a third recognition model, and performing coal and gangue recognition on the coal blocks and/or the gangue to obtain a third recognition result, wherein the third recognition result comprises the categories of the coal blocks and/or the gangue in the target image.
The differential description of the coal block and the gangue is limited due to visual features such as X-ray, gray scale and texture. However, deep learning has powerful feature representation and model fitting capabilities, and shows powerful performance in many tasks. Therefore, in order to fully mine the characteristics of the coal block and the gangue, in the embodiment of the present application, a deep learning method may be used for recognition, and a third recognition result is obtained, where the third recognition result includes the category of the coal block and/or the gangue in the target image, and the category of the coal block and/or the gangue in the target image is used to identify the category of the coal block and/or gangue in the target image.
In the embodiment of the application, a large-scale and diversified coal block and gangue database needs to be established, and a large number of labels are marked for model training. Secondly, training is carried out by using target detection algorithms such as Faster R-CNN, YOLO and the like to obtain a third recognition model.
In an embodiment of the present application, the above-mentioned preset recognition model includes: the first recognition model, the second recognition model, and the third recognition model are taken as examples, and the following description is specifically made on the construction process of the first recognition model, the second recognition model, and the third recognition model:
fig. 6 is a flowchart illustrating a method for constructing a specific implementation of the first recognition model according to an embodiment of the present application.
Referring to fig. 6, in the embodiment of the present application, the method for constructing the first recognition model specifically includes the following steps:
s601: dual energy X-ray images are acquired that include at least coal and/or gangue.
In the embodiment of the application, the first identification model is constructed by modeling according to the characteristics of the coal block and/or the gangue on the physical structure, which are obtained by the characteristic extraction sensor, as shown in fig. 5, the dual-energy X-ray image of the coal block and the gangue is acquired by the X-ray, and as the attenuation degrees of the coal block and the gangue under the irradiation of the X-ray are different, the final imaging result has certain statistical difference in gray scale. In order to better represent the difference, the coal blocks and the gangue are distinguished, a large number of dual-energy X-ray images of the coal blocks and the gangue are continuously acquired, and a gray level distribution histogram of the coal blocks and the gangue is obtained through processing of a correlation function relation. Theoretically, the coal blocks and the gangue show obvious statistical difference on the gray level distribution histogram, and the coal blocks and the gangue can be effectively classified and identified by setting a reasonable threshold value. It should be noted that the dual-energy X-ray used in the embodiments of the present application to classify the coal and the gangue mainly utilizes the difference in physical structure.
S602: and calculating the mass absorption coefficient ratio corresponding to the dual-energy X-ray image according to a preset calculation formula, wherein the mass absorption coefficient ratio is the characteristic of the coal blocks and/or the gangue on the physical structure.
When X-rays are irradiated, a large amount of energy is absorbed, and the intensity of the X-rays is also sharply attenuated, which is called X-ray attenuation. The attenuation follows a certain regularity, the determining factor of which is the characteristics and properties of the substance itself. For this reason, a mathematical formula can be established by the attenuation law and the characteristics of the substance itself, which is the theoretical basis for X-ray identification of substances.
Sorting principle based on dual-energy X-ray:
with regard to such rays, the beam is formed by the lambert-beer principle:
I＝I0e-μhρ
wherein I is the intensity of the ray before incidence, I0The intensity of the radiation after passing through the detected substance; mu is a mass absorption coefficient; h thickness of the object to be measured; and rho is the density of the measured object.
Therefore, the penetration force of the X-ray is related to the density of the substance, and the substance can be distinguished by utilizing the different attenuation intensities of the X-ray after the X-ray penetrates through the substances with different densities.
In the embodiment of the application, the density of the coal block and the density of the gangue are greatly different, so that the intensity attenuation is different after the high-energy X-ray and the low-energy X-ray respectively irradiate the coal block and the gangue, and the feasibility of the method for detecting the coal gangue identification based on the X-ray is also explained.
After the high-energy X-ray and the low-energy X-ray respectively irradiate the coal block and the gangue, the intensity attenuation of the coal block and the gangue is respectively as follows:
Figure BDA0002602756190000121
Figure BDA0002602756190000122
combining the two formulas and finishing to obtain:
Figure BDA0002602756190000123
since the high-energy X-ray and the low-energy X-ray penetrate through the same substance, the density of the substance and the thickness of the penetrated substance along the incident path of the X-ray are the same in the calculation process and can be mutually offset. Therefore, the calculated K value is a ratio of the mass absorption coefficients, and is only related to the atomic number of the substance, and the K value is a mark quantity representing the attribute of the substance, which is also a basis for identifying the coal gangue by the first identification model in the embodiment of the application.
Under the irradiation of dual-energy X-rays, acquiring dual-energy X-ray images of a large amount of coal and gangue, respectively calculating K values to obtain statistical information of the K values of the coal blocks and the gangue, and distinguishing the coal blocks and the gangue according to the difference of the K values. For example: the high-quality absorption coefficient and the low-quality absorption coefficient of the coal briquette are respectively 0.1589 and 0.1989 through statistics, and then the K value is calculated to be about 1.251; and the high-mass absorption coefficient and the low-mass absorption coefficient of the gangue are 0.1791 and 0.4099 respectively, so that the K value is calculated to be about 2.29. The coal blocks and the gangue can be distinguished according to the difference of K values.
S603: and according to the mass absorption coefficient, marking the coal blocks and/or the gangue in the dual-energy X-ray image to obtain a marked dual-energy X-ray image, wherein the marked dual-energy X-ray image comprises the categories of the coal blocks and/or the gangue.
S604: and taking the marked dual-energy X-ray image as sample data, and performing model training according to the mass absorption coefficient ratio to obtain a first recognition model.
In the embodiment of the application, coal blocks and/or gangue in the dual-energy X-ray image are/is marked according to the mass absorption coefficient ratio to obtain the marked dual-energy X-ray image, and model training is performed according to the mass absorption coefficient ratio by using the marked dual-energy X-ray image as sample data to obtain a first recognition model.
Fig. 7 is a flowchart illustrating a method for constructing a specific implementation of the second recognition model according to an embodiment of the present application.
Referring to fig. 7, in the embodiment of the present application, the method for constructing the second recognition model specifically includes the following steps:
s701: and intercepting a video image from the monitoring video, wherein the video image at least comprises coal blocks and/or gangue.
In the embodiment of the application, in the running scene of the belt, the coal blocks and the gangue in the coal gangue mixture borne by the running belt are always monitored in real time by the monitoring equipment, and when the second identification model is constructed, the video images are intercepted from the monitoring videos by acquiring the monitoring videos corresponding to the running of the monitoring belt.
S702: and carrying out gray level analysis and texture analysis on the video image, and extracting gray level characteristics and texture characteristics of the video image, wherein the gray level characteristics and the texture characteristics of the video image are characteristics of coal blocks and/or gangue on texture and/or color.
In the embodiment of the application, the visual difference of the coal blocks and the gangue is utilized for classification and identification, and the gray level and the texture characteristics of the coal blocks and the gangue are mainly used for classification and identification. Generally, the coal block has a relatively black color, and the gangue has a relatively light color, so that the gray distribution and peak value of the coal block and the gangue are different. It can be seen that the gray scale and texture analysis of the coal blocks and the gangue can help to identify the coal blocks and the gangue.
It should be noted that, for gray analysis, in an actually acquired video image, a matte coal block looks relatively black and has a relatively low gray value, while a glossy coal block reflects light under light irradiation, so that a video image has a luminous point, and the gray value of the luminous point is relatively high. Therefore, the gray level histograms of the coal blocks and the gangue can visually reflect the gray level range and the frequency distribution of the coal blocks and the gangue, wherein the mean value and the variance are two common characteristic parameters related to the gray level. Therefore, in the embodiment of the application, the mean and variance of the coal blocks and the gangue are mainly extracted as the gray features for analysis.
For the texture analysis, because the chemical substances constituting the coal block and the gangue are different, the reflected physical properties are different, and the coal block and the gangue have certain difference in texture, so that the texture features of the coal block and the gangue image are used for identifying the coal block and the gangue image in the embodiment of the application. Therefore, in the embodiment of the application, the characteristic parameter values such as the second moment, the contrast, the correlation, the entropy, the inverse moment difference and the like can be extracted from the gray level co-occurrence matrix of the coal blocks and the gangue to quantitatively describe the texture characteristics of the image for analysis.
S703: and marking the coal blocks and/or the gangue in the video image according to the gray features and the texture features of the video image to obtain a marked video image, wherein the marked video image comprises the categories of the coal blocks and/or the gangue.
S704: and taking the marked video image as sample data, and performing model training by using an SVM classifier as a classified neural network to obtain a second recognition model.
In the embodiment of the application, the coal blocks and/or the gangue in the video image are labeled according to the gray feature and the texture feature of the video image to obtain a labeled video image, the labeled video image is used as sample data, and after the gray feature and the texture feature of the coal blocks and the gangue are obtained, an SVM classifier is used as a classified neural network for training.
The SVM is a classic classifier and has good effects in a plurality of classification and recognition tasks. The SVM is suitable for learning and classifying characteristics of a system with small samples, high dimensionality and nonlinearity. Meanwhile, the SVM has strong generalization capability and self-learning capability.
However, in the actual training process, the parameter setting in the SVM is mainly set by manual experience, and the quality of the parameter has a large influence on the classification effect. Therefore, in the embodiment of the present application, a particle swarm algorithm is selected to search for the optimal value of the parameter in the SVM. However, the particle swarm algorithm is also prone to the problems of local optimization, low search accuracy and the like, and therefore, the particle swarm algorithm needs to be improved first, and then the improved particle swarm algorithm is used for optimizing parameter values in the SVM.
The improvement scheme of the particle swarm optimization is as follows:
the conventional particle swarm algorithm has the following formula:
Figure BDA0002602756190000141
Figure BDA0002602756190000142
where ω denotes the inertial weight, c1Denotes a self-learning factor, c2Represents a social learning factor, r1，r2Represents [0, 1 ]]In betweenA random number.
Figure BDA0002602756190000151
And
Figure BDA0002602756190000152
respectively representing the position and the speed of a jth variable of the ith particle in the nth iteration;
Figure BDA0002602756190000153
and
Figure BDA0002602756190000154
and j variable values respectively representing the ith particle history optimal position and the whole particle history optimal position.
In order to solve the problems in the conventional particle swarm optimization, a new inertia weight adjustment is introduced, and the new nonlinear dynamic inertia weight adjustment is as follows:
Figure BDA0002602756190000155
Figure BDA0002602756190000156
in the formula: h is a [0, 1 ]]A random number in between; n current iteration times; n maximum number of iterations; omegamaxIs the maximum value of the inertial weight; omegaminIs the minimum value of the inertial weight.
It should be noted that, in the embodiment of the present application, the gray scale and the texture feature may be directly extracted from the original image of the coal block and the gangue, or the original image may be preprocessed first, and this preprocessing step will eliminate the interference of the belt on the foreground object (the coal block and the gangue), and has an important role in improving the performance of the algorithm.
Further, in the embodiment of the present application, before performing gray scale analysis and texture analysis on the video image, it may also need to include: specifically, fig. 8 is a schematic flow diagram of a specific implementation of the preprocessing process shown in the embodiment of the present application.
As shown in fig. 8, the pretreatment process specifically includes the following steps:
s801: and carrying out image graying processing on the video image to obtain a grayscale image of the target image.
The image graying processing in the preprocessing step is performed because a grayscale image is mainly used for identifying the coal gangue, and therefore, the image is grayed first.
S802: and carrying out image smoothing on the gray level image of the target image to obtain an image after smoothing.
The image smoothing in the preprocessing step is because the image is interfered by noise due to the environment and shooting equipment during the acquisition, conversion and transmission processes, and the information contained in the image is affected, so that the image is smoothed to prepare for the subsequent extraction and identification of the coal block and gangue image features and eliminate the noise generated during the digitization of the image.
S803: and carrying out image sharpening on the image after the smoothing processing to obtain an image after the sharpening processing.
The image sharpening in the preprocessing step is to sharpen the smoothed image in order to enhance the detail information in the image and enhance the detail information blurred after the image smoothing.
S804: and performing image segmentation processing on the sharpened image by adopting an Otsu algorithm to obtain an image to be identified.
The image segmentation in the preprocessing step refers to that in an actual coal and gangue separation environment, coal blocks and gangue are transported on a conveyor belt together, a shot image contains the conveyor belt and the coal blocks and gangue, and due to the fact that the conveyor belt causes difficulty in feature extraction and identification of subsequent coal blocks and gangue images, the coal blocks and gangue images are processed through image segmentation.
Fig. 10 is a flowchart illustrating a method for constructing a specific implementation of the third recognition model according to an embodiment of the present application.
Referring to fig. 10, in the embodiment of the present application, the method for constructing the third recognition model specifically includes the following steps:
s1001: under fixed illumination, coal gangue images of different places and different types are collected, and the coal gangue images at least comprise coal blocks and/or gangue.
In recent years, a large amount of research and application are carried out on a target identification technology based on deep learning, and simultaneous identification of multiple types of targets can be realized by utilizing strong feature extraction capability and discrimination capability of a neural network. Therefore, the technology of deep learning is applied to the identification of the coal gangue in the embodiment of the application.
The data set of the target detection based on the deep learning has certain requirements, and in the embodiment of the application, the coal gangue image acquisition needs to follow the following principle:
(1) the diversity of coal block and gangue samples needs to be ensured
A large number of different samples are needed for convolutional neural network training, so that coal blocks and gangue sample images of different types need to be collected in different places for a long time, the order of magnitude is about 10000, and the diversity of samples is guaranteed.
(2) Sample balancing
If the data of each sample in the training set sample is not balanced, the generalization of the model on the test set is not good, and the classifier cannot meet the classification requirement. If the sample balance deviation is too large, the sample data needs to be supplemented or deleted. However, in consideration of the particularity of coal block and gangue identification, gangue is mainly selected from the coal gangue mixture, and therefore the number of gangue samples is required to be more than that of coal block samples.
(3) Fixed illumination
Under the same illumination condition, the stability of coal block and gangue image samples is ensured, and image noise caused by different illumination is reduced.
S1002: and performing multi-scale and multi-semantic feature extraction on the coal gangue image to obtain multi-fusion features, wherein the multi-fusion features are features of the coal blocks and/or the gangue on multiple scales and multiple dimensions.
S1003: and marking the coal gangue image according to the multi-fusion characteristics to obtain a marked coal gangue image, wherein the marked coal gangue image comprises the categories of coal blocks and/or gangue.
After the data acquisition is completed, the data set needs to be labeled, in the embodiment of the application, label of the data set is completed by using LabelImage software, and the main labeled content is the category of the coal blocks and the gangue.
S1004: and taking the marked coal gangue image as sample data, and performing model training by using a convolutional neural network to obtain a third recognition model.
In the embodiment of the application, model training of the convolutional neural network can be realized based on a target detection algorithm of the Faster R-CNN, however, the fast R-CNN is a classic target detection algorithm which only uses the last convolutional layer of the backbone network (VGG16) to extract features, so that the network cannot effectively adapt to the scale change of an object. However, in actual production processes, the dimensional changes of the coal blocks and the gangue are very extensive. Therefore, in order to improve the recognition rate of the coal blocks and the gangue, in the embodiment of the application, the features with different resolutions in the Faster R-CNN are fused based on the attention mechanism, and the recognition rate of the object is improved by using the multi-scale features. The high-resolution characteristic receptive field is small, so that small target characteristics can be better extracted; the characteristic receptive field of low resolution is larger, and the large target characteristic can be better extracted, therefore, the attention mechanism is introduced into FasterR-CNN, and the target detection is carried out by utilizing the multi-scale characteristic, so that the precision of the target detection can be effectively improved. The modified frame is shown in fig. 11.
As shown in fig. 11, we use different levels of features in VGG16, where the features at the bottom level contain more location information and the features at the top level have better semantic information. In order to fuse features of different levels, firstly down-sampling is carried out on features of high resolution, up-sampling is carried out on features of low resolution to obtain features of the same scale and different levels, then concat operation is used for the features to obtain features F after fusion, weights of the features of different channels are obtained by utilizing an Attention mechanism between the channels, so that new features G are obtained, and element levels of the features F and the features G are added to obtain final features H. In this case, H is a feature that fuses multiple dimensions and multiple meanings, and subsequent features are performed on H.
Corresponding to the embodiment of the application function implementation method, the application also provides a coal and gangue identification device, electronic equipment and a corresponding embodiment.
Fig. 12 is a schematic structural diagram of a gangue identification device according to an embodiment of the present application.
Referring to fig. 12, an embodiment of the present application provides a coal gangue identification device, which specifically includes:
the acquiring unit 1201 is configured to acquire a target image of the coal and gangue to be identified, where the coal and gangue to be identified at least includes coal blocks and/or gangue.
The recognition unit 1202 is configured to input the target image into a preset recognition model, so as to obtain at least two recognition results of the target image.
And a fusing unit 1203, configured to perform fusion processing on at least two recognition results of the target image according to a preset rule to obtain a recognition result of the coal and gangue to be recognized, where the recognition result of the coal and gangue to be recognized includes the category of the coal block and/or the gangue in the target image.
With regard to the apparatus in the above-described embodiment, the specific manner in which each module performs the operation has been described in detail in the embodiment related to the method, and will not be elaborated here.
Fig. 13 is a schematic structural diagram of an electronic device shown in an embodiment of the present application.
Referring to fig. 13, an electronic device 1300 includes a memory 1310 and a processor 1320.
Processor 1320 may be a Central Processing Unit (CPU), other general purpose Processor, a Digital Signal Processor (DSP), an Application Specific Integrated Circuit (ASIC), a Field Programmable Gate Array (FPGA) or other Programmable logic device, discrete Gate or transistor logic, discrete hardware components, etc. A general purpose processor may be a microprocessor or the processor may be any conventional processor or the like.
The memory 1310 may include various types of storage units, such as system memory, Read Only Memory (ROM), and permanent storage. The ROM may store, among other things, static data or instructions for the processor 1320 or other modules of the computer. The persistent storage device may be a read-write storage device. The persistent storage may be a non-volatile storage device that does not lose stored instructions and data even after the computer is powered off. In some embodiments, the persistent storage device employs a mass storage device (e.g., magnetic or optical disk, flash memory) as the persistent storage device. In other embodiments, the permanent storage may be a removable storage device (e.g., floppy disk, optical drive). The system memory may be a read-write memory device or a volatile read-write memory device, such as a dynamic random access memory. The system memory may store instructions and data that some or all of the processors require at runtime. Further, the memory 1310 may include any combination of computer-readable storage media, including various types of semiconductor memory chips (DRAM, SRAM, SDRAM, flash memory, programmable read-only memory), magnetic and/or optical disks, may also be employed. In some embodiments, memory 1310 may include a removable storage device that is readable and/or writable, such as a Compact Disc (CD), a digital versatile disc read only (e.g., DVD-ROM, dual layer DVD-ROM), a Blu-ray disc read only, an ultra-dense disc, a flash memory card (e.g., SD card, min SD card, Micro-SD card, etc.), a magnetic floppy disk, or the like. Computer-readable storage media do not contain carrier waves or transitory electronic signals transmitted by wireless or wired means.
The memory 1310 has stored thereon executable code that, when processed by the processor 1320, may cause the processor 1320 to perform some or all of the methods described above.
The aspects of the present application have been described in detail hereinabove with reference to the accompanying drawings. In the above embodiments, the descriptions of the respective embodiments have respective emphasis, and for parts that are not described in detail in a certain embodiment, reference may be made to related descriptions of other embodiments. Those skilled in the art should also appreciate that the acts and modules referred to in the specification are not necessarily required in the present application. In addition, it can be understood that the steps in the method of the embodiment of the present application may be sequentially adjusted, combined, and deleted according to actual needs, and the modules in the device of the embodiment of the present application may be combined, divided, and deleted according to actual needs.
Furthermore, the method according to the present application may also be implemented as a computer program or computer program product comprising computer program code instructions for performing some or all of the steps of the above-described method of the present application.
Alternatively, the present application may also be embodied as a non-transitory machine-readable storage medium (or computer-readable storage medium, or machine-readable storage medium) having stored thereon executable code (or a computer program, or computer instruction code) which, when executed by a processor of an electronic device (or electronic device, server, etc.), causes the processor to perform part or all of the various steps of the above-described method according to the present application.
Those of skill would further appreciate that the various illustrative logical blocks, modules, circuits, and algorithm steps described in connection with the applications disclosed herein may be implemented as electronic hardware, computer software, or combinations of both.
The flowchart and block diagrams in the figures illustrate the architecture, functionality, and operation of possible implementations of systems and methods according to various embodiments of the present application. In this regard, each block in the flowchart or block diagrams may represent a module, segment, or portion of code, which comprises one or more executable instructions for implementing the specified logical function(s). It should also be noted that, in some alternative implementations, the functions noted in the block may occur out of the order noted in the figures. For example, two blocks shown in succession may, in fact, be executed substantially concurrently, or the blocks may sometimes be executed in the reverse order, depending upon the functionality involved. It will also be noted that each block of the block diagrams and/or flowchart illustration, and combinations of blocks in the block diagrams and/or flowchart illustration, can be implemented by special purpose hardware-based systems which perform the specified functions or acts, or combinations of special purpose hardware and computer instructions.
Having described embodiments of the present application, the foregoing description is intended to be exemplary, not exhaustive, and not limited to the disclosed embodiments. Many modifications and variations will be apparent to those of ordinary skill in the art without departing from the scope and spirit of the described embodiments. The terminology used herein is chosen in order to best explain the principles of the embodiments, the practical application, or improvements made to the technology in the marketplace, or to enable others of ordinary skill in the art to understand the embodiments disclosed herein.

Claims (10)

1. A coal gangue identification method is characterized by comprising the following steps:
acquiring a target image of coal and gangue to be identified, wherein the coal and gangue to be identified at least comprises coal blocks and/or gangue;
inputting the target image into a preset identification model to obtain at least two identification results of the target image;
and fusing at least two recognition results of the target image according to a preset rule to obtain a recognition result of the coal and gangue to be recognized, wherein the recognition result of the coal and gangue to be recognized comprises the category of the coal block and/or the gangue in the target image.
2. The method according to claim 1, wherein the fusion processing is performed on at least two recognition results of the target image according to a preset rule to obtain the recognition result of the gangue to be recognized, specifically:
and performing linear weighting processing on at least two recognition results of the target image to obtain the recognition result of the coal gangue to be recognized.
3. The method of claim 1 or 2, wherein the target image comprises: the first image of the coal gangue to be identified under the dual-energy X-ray and the second image obtained through the image acquisition device, wherein the preset identification model comprises: the first recognition model and the second recognition model input the target image into a preset recognition model to obtain at least two recognition results of the target image, specifically:
inputting the first image into the first recognition model to obtain a first recognition result of the target image, wherein the first recognition model is obtained by modeling according to the characteristics of the coal block and/or the gangue on the physical structure, which are obtained by the characteristic extraction sensor;
and inputting the second image into the second recognition model to obtain a second recognition result of the target image, wherein the second recognition model is obtained by modeling according to the characteristics of the coal block and/or the gangue on texture and/or color, which are obtained by the characteristic extraction sensor.
4. The method of claim 3, wherein the pre-set recognition model further comprises: and the third identification model is used for inputting the target image into a preset identification model to obtain an identification result of the target image, and specifically comprises the following steps:
and inputting the second image into the third identification model to obtain a third identification result of the target image, wherein the third identification model is obtained by modeling the characteristics of the coal block and/or the gangue on multiple scales and multiple dimensions, which are obtained by the characteristic extraction sensor.
5. The method of claim 4, further comprising: and determining the position information of the coal block and/or the gangue in the target image.
6. The method of claim 5, further comprising:
and sending the identification result of the coal and gangue to be identified and the position information of the coal block and/or the gangue in the target image to a coal and gangue separation system, so that the coal and gangue separation system performs coal and gangue separation on the coal and gangue to be separated according to the identification result of the coal and gangue to be identified and the position information of the coal block and/or the gangue in the target image.
7. The method according to claim 5, wherein the determining of the position information of the coal block and/or the gangue in the target image is specifically:
detecting the outline of the coal block and/or the gangue by adopting a Canny edge detection method, and determining the position area of the coal block and/or the gangue in the target image;
obtaining the centroid of the position area by using a centroid method, and determining a centroid coordinate;
and taking the centroid coordinate as the position information of the coal block and/or the gangue in the target image.
8. A coal gangue identification device is characterized by comprising:
the device comprises an acquisition unit, a processing unit and a processing unit, wherein the acquisition unit is used for acquiring a target image of the coal gangue to be identified, and the coal gangue to be identified at least comprises coal blocks and/or gangue;
the identification unit is used for inputting the target image into a preset identification model to obtain at least two identification results of the target image;
and the fusion unit is used for performing fusion processing on at least two recognition results of the target image according to a preset rule to obtain a recognition result of the coal and gangue to be recognized, wherein the recognition result of the coal and gangue to be recognized comprises the type of the coal block and/or the gangue in the target image.
9. An electronic device, comprising:
a processor; and
a memory having executable code stored thereon, which when executed by the processor, causes the processor to perform the method of coal refuse separation according to any one of claims 1-7.
10. A non-transitory machine-readable storage medium having stored thereon executable code that, when executed by a processor of an electronic device, causes the processor to perform the method of coal gangue separation of any one of claims 1-7.
CN202010729919.XA 2020-07-27 2020-07-27 Coal and gangue identification method and device, electronic equipment and storage medium Pending CN111881909A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202010729919.XA CN111881909A (en) 2020-07-27 2020-07-27 Coal and gangue identification method and device, electronic equipment and storage medium

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202010729919.XA CN111881909A (en) 2020-07-27 2020-07-27 Coal and gangue identification method and device, electronic equipment and storage medium

Publications (1)

Publication Number Publication Date
CN111881909A true CN111881909A (en) 2020-11-03

Family

ID=73201555

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202010729919.XA Pending CN111881909A (en) 2020-07-27 2020-07-27 Coal and gangue identification method and device, electronic equipment and storage medium

Country Status (1)

Country Link
CN (1) CN111881909A (en)

Cited By (8)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN112749636A (en) * 2020-12-29 2021-05-04 精英数智科技股份有限公司 Method, device and system for monitoring water exploration and drainage of coal mine and storage medium
CN112991438A (en) * 2021-04-12 2021-06-18 天津美腾科技股份有限公司 Coal and gangue detection and identification system, intelligent coal discharge system and model training method
CN113344912A (en) * 2021-07-06 2021-09-03 矿冶科技集团有限公司 Ore particle segmentation system, method and machine-readable storage medium
CN113458005A (en) * 2021-06-11 2021-10-01 安徽理工大学 Coal gangue X-ray identification system based on pure mineral simulation
CN113591583A (en) * 2021-06-30 2021-11-02 沈阳科来沃电气技术有限公司 Intelligent boron ore beneficiation system and method
CN114535133A (en) * 2022-01-12 2022-05-27 山东大学 Coal and gangue sorting method, device and system based on dual-energy ray transmission imaging
CN117011302A (en) * 2023-10-08 2023-11-07 山东济宁运河煤矿有限责任公司 Intelligent dry separation system based on coal gangue identification
CN117288783A (en) * 2023-11-24 2023-12-26 深圳翱翔锐影科技有限公司 X-ray-based gangue sorting method, computer equipment and storage medium

Cited By (12)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN112749636A (en) * 2020-12-29 2021-05-04 精英数智科技股份有限公司 Method, device and system for monitoring water exploration and drainage of coal mine and storage medium
CN112749636B (en) * 2020-12-29 2023-10-31 精英数智科技股份有限公司 Monitoring method, device and system for water drainage detection of coal mine and storage medium
CN112991438A (en) * 2021-04-12 2021-06-18 天津美腾科技股份有限公司 Coal and gangue detection and identification system, intelligent coal discharge system and model training method
CN113458005A (en) * 2021-06-11 2021-10-01 安徽理工大学 Coal gangue X-ray identification system based on pure mineral simulation
CN113591583A (en) * 2021-06-30 2021-11-02 沈阳科来沃电气技术有限公司 Intelligent boron ore beneficiation system and method
CN113344912A (en) * 2021-07-06 2021-09-03 矿冶科技集团有限公司 Ore particle segmentation system, method and machine-readable storage medium
CN113344912B (en) * 2021-07-06 2024-04-16 矿冶科技集团有限公司 Ore particle segmentation system, method and machine-readable storage medium
CN114535133A (en) * 2022-01-12 2022-05-27 山东大学 Coal and gangue sorting method, device and system based on dual-energy ray transmission imaging
CN117011302A (en) * 2023-10-08 2023-11-07 山东济宁运河煤矿有限责任公司 Intelligent dry separation system based on coal gangue identification
CN117011302B (en) * 2023-10-08 2024-01-09 山东济宁运河煤矿有限责任公司 Intelligent dry separation system based on coal gangue identification
CN117288783A (en) * 2023-11-24 2023-12-26 深圳翱翔锐影科技有限公司 X-ray-based gangue sorting method, computer equipment and storage medium
CN117288783B (en) * 2023-11-24 2024-02-20 深圳翱翔锐影科技有限公司 X-ray-based gangue sorting method, computer equipment and storage medium

Similar Documents

Publication Publication Date Title
CN111881909A (en) Coal and gangue identification method and device, electronic equipment and storage medium
Chatterjee Vision-based rock-type classification of limestone using multi-class support vector machine
EP3349048B1 (en) Inspection devices and methods for detecting a firearm in a luggage
CN109598287B (en) Appearance flaw detection method for resisting network sample generation based on deep convolution generation
CN111275688A (en) Small target detection method based on context feature fusion screening of attention mechanism
CN106093066A (en) A kind of magnetic tile surface defect detection method based on the machine vision attention mechanism improved
CN104834942B (en) Remote sensing image variation detection method and system based on mask classification
CN103258213A (en) Vehicle model dynamic identification method used in intelligent transportation system
CN111833322B (en) Garbage multi-target detection method based on improved YOLOv3
CN111127454A (en) Method and system for generating industrial defect sample based on deep learning
CN103679214A (en) Vehicle detection method based on online area estimation and multi-feature decision fusion
Zhao et al. Butterfly recognition based on faster R-CNN
CN107886066A (en) A kind of pedestrian detection method based on improvement HOG SSLBP
CN114627383A (en) Small sample defect detection method based on metric learning
CN107103608A (en) A kind of conspicuousness detection method based on region candidate samples selection
CN114882215A (en) Shape selection identification method for particle aggregate region of photoelectric coal gangue sorting image
Hu et al. Automatic detection of sound knots and loose knots on sugi using gray level co-occurrence matrix parameters
Alohali et al. Automated fault detection in the Arabian Basin
CN115953666B (en) Substation site progress identification method based on improved Mask-RCNN
CN117011274A (en) Automatic glass bottle detection system and method thereof
Li et al. Online rail fastener detection based on YOLO network
Guo et al. Automatic Rock Classification Algorithm Based on Ensemble Residual Network and Merged Region Extraction
Hubálovský et al. Evaluation of deepfake detection using YOLO with local binary pattern histogram
Deo et al. Vision-based size classification of iron ore pellets using ensembled convolutional neural network
CN109886340A (en) A kind of Remote Image Classification

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination