CN114677546A - Real-time mobile coal rock sorting method and device - Google Patents
Real-time mobile coal rock sorting method and device Download PDFInfo
- Publication number
- CN114677546A CN114677546A CN202210347212.1A CN202210347212A CN114677546A CN 114677546 A CN114677546 A CN 114677546A CN 202210347212 A CN202210347212 A CN 202210347212A CN 114677546 A CN114677546 A CN 114677546A
- Authority
- CN
- China
- Prior art keywords
- coal
- rock
- block
- blocks
- identified
- 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
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
-
- B—PERFORMING OPERATIONS; TRANSPORTING
- B07—SEPARATING SOLIDS FROM SOLIDS; SORTING
- B07C—POSTAL SORTING; SORTING INDIVIDUAL ARTICLES, OR BULK MATERIAL FIT TO BE SORTED PIECE-MEAL, e.g. BY PICKING
- B07C5/00—Sorting according to a characteristic or feature of the articles or material being sorted, e.g. by control effected by devices which detect or measure such characteristic or feature; Sorting by manually actuated devices, e.g. switches
- B07C5/02—Measures preceding sorting, e.g. arranging articles in a stream orientating
-
- B—PERFORMING OPERATIONS; TRANSPORTING
- B07—SEPARATING SOLIDS FROM SOLIDS; SORTING
- B07C—POSTAL SORTING; SORTING INDIVIDUAL ARTICLES, OR BULK MATERIAL FIT TO BE SORTED PIECE-MEAL, e.g. BY PICKING
- B07C5/00—Sorting according to a characteristic or feature of the articles or material being sorted, e.g. by control effected by devices which detect or measure such characteristic or feature; Sorting by manually actuated devices, e.g. switches
- B07C5/34—Sorting according to other particular properties
- B07C5/342—Sorting according to other particular properties according to optical properties, e.g. colour
-
- B—PERFORMING OPERATIONS; TRANSPORTING
- B07—SEPARATING SOLIDS FROM SOLIDS; SORTING
- B07C—POSTAL SORTING; SORTING INDIVIDUAL ARTICLES, OR BULK MATERIAL FIT TO BE SORTED PIECE-MEAL, e.g. BY PICKING
- B07C5/00—Sorting according to a characteristic or feature of the articles or material being sorted, e.g. by control effected by devices which detect or measure such characteristic or feature; Sorting by manually actuated devices, e.g. switches
- B07C5/36—Sorting apparatus characterised by the means used for distribution
- B07C5/361—Processing or control devices therefor, e.g. escort memory
- B07C5/362—Separating or distributor mechanisms
-
- 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
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T1/00—General purpose image data processing
- G06T1/0014—Image feed-back for automatic industrial control, e.g. robot with camera
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/0002—Inspection of images, e.g. flaw detection
- G06T7/0004—Industrial image inspection
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/60—Analysis of geometric attributes
- G06T7/66—Analysis of geometric attributes of image moments or centre of gravity
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/70—Determining position or orientation of objects or cameras
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/20—Special algorithmic details
- G06T2207/20081—Training; Learning
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/30—Subject of image; Context of image processing
- G06T2207/30004—Biomedical image processing
- G06T2207/30101—Blood vessel; Artery; Vein; Vascular
- G06T2207/30104—Vascular flow; Blood flow; Perfusion
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Data Mining & Analysis (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Evolutionary Biology (AREA)
- Evolutionary Computation (AREA)
- Bioinformatics & Computational Biology (AREA)
- General Engineering & Computer Science (AREA)
- Artificial Intelligence (AREA)
- Life Sciences & Earth Sciences (AREA)
- Robotics (AREA)
- Quality & Reliability (AREA)
- Geometry (AREA)
- Sorting Of Articles (AREA)
Abstract
The invention relates to a mobile coal rock real-time sorting method and a device, belonging to the technical field of coal rock sorting and comprising the following steps: acquiring a hyperspectral image to be identified; performing category identification on each pixel in the hyperspectral image to be identified by using the trained coal-rock classification model to obtain a coal-rock classification result; identifying the coal blocks and the rock blocks in the hyperspectral image to be identified according to the coal rock classification result to obtain a coal block-rock block identification result; the coal block-rock block identification result comprises position information of each coal block and each rock block in the image; and controlling a sorting execution unit to sort the coal blocks and the rock blocks on the conveyor belt according to the coal block-rock block recognition result. The accuracy of classification and identification of the coal rock is improved; the whole process is completely processed by a computer, manual intervention is not needed, the labor cost is saved, and the personal safety is guaranteed.
Description
Technical Field
The invention relates to the technical field of coal rock sorting, in particular to a mobile coal rock real-time sorting method and device.
Background
In the field of comprehensive exploitation, a coal rock identification technology is a key technology for improving safety and resource exploitation efficiency. The existing coal and rock identification method mainly comprises an artificial gamma ray identification method, a radar detection identification method and a coal and rock impact vibration measurement method. The manual gamma ray identification method cannot guarantee the personal safety of technicians due to radioactivity; the radar detection and identification method is based on electromagnetic wave propagation, and the contradiction that the measurement range and the precision cannot be obtained simultaneously is difficult to thoroughly solve; the measuring method of the coal rock impact vibration is easy to realize, but the accuracy is not high.
Disclosure of Invention
The invention aims to provide a method and a device for sorting mobile coal rocks in real time, which are combined with a hyperspectral imaging technology, improve the precision of sorting the coal rocks and realize the automation of sorting the coal rocks.
In order to achieve the purpose, the invention provides the following scheme:
a real-time mobile coal rock sorting method comprises the following steps:
acquiring a hyperspectral image to be identified; the hyperspectral image to be identified comprises a coal block-rock block mixed flow to be sorted, which is conveyed on a conveyor belt in real time;
performing category identification on each pixel in the hyperspectral image to be identified by using the trained coal-rock classification model to obtain a coal-rock classification result; the coal and rock classification result comprises a category identification result of each pixel in the hyperspectral image to be identified, and the category identification result comprises coal categories, rock categories and categories which cannot be identified;
identifying the coal blocks and the rock blocks in the hyperspectral image to be identified according to the coal rock classification result to obtain a coal block-rock block identification result; the coal block-rock block identification result comprises position information of each coal block and each rock block in the hyperspectral image to be identified;
and controlling a sorting execution unit to sort the coal blocks and the rock blocks on the conveyor belt according to the coal block-rock block recognition result.
Optionally, the method further comprises:
constructing a training data set; the training data set comprises training hyperspectral images, and coal blocks and rock blocks in the training hyperspectral images are marked with corresponding coal labels and rock labels;
constructing the coal-rock classification model by adopting a support vector machine algorithm;
and training the coal-rock classification model by using the training data set to obtain the trained coal-rock classification model.
Optionally, the identifying the coal blocks and the rock blocks in the hyperspectral image to be identified according to the coal and rock classification result to obtain a coal block-rock block identification result specifically includes:
detecting the object outline in the hyperspectral image to be identified by utilizing an edge detection algorithm;
determining the classes of all pixels in the object contour according to the coal rock classification result;
determining the type of the object contour according to the types of all pixels in the object contour;
and determining the centroid of the object contour, and determining the coordinates of the centroid as the position information of the object contour.
Optionally, an object contour in the hyperspectral image to be identified is detected by using a Robert edge detection algorithm.
Optionally, the controlling, according to the coal-rock recognition result, a sorting execution unit to sort the coal and the rock on the conveyor belt specifically includes:
determining the time for the coal or rock to reach a picking point according to the coal-rock recognition result and the running speed of the conveyor belt;
and controlling the sorting execution unit to sort the coal blocks and the rock blocks according to the time required by the sorting execution unit to reach the picking point and the time required by the coal blocks or the rock blocks to reach the picking point.
Optionally, the method further comprises: and performing smoothing processing and normalized transformation processing on the hyperspectral image to be identified.
On the other hand, the invention also provides a mobile coal rock real-time sorting device, which comprises:
a conveyor belt for conveying a coal-rock mixed stream to be sorted;
the hyperspectral imaging module is fixedly arranged above the conveyor belt and used for acquiring a hyperspectral image to be identified; the hyperspectral image to be identified comprises a coal block-rock block mixed flow to be sorted, which is conveyed on the conveyor belt in real time;
the computer is in signal connection with the hyperspectral imaging module and is used for receiving the hyperspectral image to be recognized and performing type recognition on each pixel in the hyperspectral image to be recognized by using the trained coal-rock classification model to obtain a coal-rock classification result; the coal and rock classification result comprises a category identification result of each pixel in the hyperspectral image to be identified, and the category identification result comprises coal categories, rock categories and categories which cannot be identified; the hyperspectral image to be identified is also used for identifying the coal blocks and the rock blocks in the hyperspectral image to be identified according to the coal and rock classification result to obtain a coal block-rock block identification result; the coal block-rock block identification result comprises position information of each coal block and each rock block in the hyperspectral image to be identified;
and the coal and rock sorting module is in signal connection with the computer and is used for acquiring the coal block-rock block identification result and sorting the coal blocks and the rock blocks on the conveyor belt according to the coal block-rock block identification result.
Optionally, the hyperspectral imaging module comprises:
a halogen lamp for generating a light source;
and the hyperspectral camera is used for acquiring a hyperspectral image to be identified.
Optionally, the apparatus further comprises: the dust fall gondola water faucet sets up the conveyer belt top is connected with computer electricity, is used for right dust on the conveyer belt suppresses.
Optionally, the coal-rock sorting module comprises:
the sorting control unit is used for determining the time when the coal blocks or the rock blocks reach the picking point according to the coal block-rock block recognition result and the running speed of the conveyor belt; the sorting control device is also used for generating a sorting control instruction according to the time required by the sorting execution unit to reach the picking point and the time required by the coal or rock to reach the picking point;
and the sorting execution unit is in control connection with the sorting control unit and is used for sorting the coal blocks and the rock blocks on the conveying belt according to the sorting control instruction.
According to the specific embodiment provided by the invention, the invention discloses the following technical effects:
the invention provides a real-time mobile coal rock sorting method and a real-time mobile coal rock sorting device, which comprise the following steps of: acquiring a hyperspectral image to be identified; performing category identification on each pixel in the hyperspectral image to be identified by using the trained coal-rock classification model to obtain a coal-rock classification result; identifying the coal blocks and the rock blocks in the hyperspectral image to be identified according to the coal rock classification result to obtain a coal block-rock block identification result; the coal block-rock block identification result comprises position information of each coal block and each rock block in the image; and controlling a sorting execution unit to sort the coal blocks and the rock blocks on the conveyor belt according to the coal block-rock block recognition result. The invention fully considers the principle that different coal rock components have different spectral absorption, combines the hyperspectral imaging technology, and compared with a radar detection identification method and a coal rock impact vibration measurement method, the hyperspectral imaging technology can fully reflect the difference of the physical structure and the chemical components of the coal rock, thereby improving the accuracy of classification and identification of the coal rock; and the process of real-time coal rock sorting is completely processed by a computer, and human intervention is not needed, so that compared with the method for identifying coal rocks by using artificial gamma rays in the prior art, the labor cost is saved, and the personal safety is guaranteed.
Drawings
In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings needed in the embodiments will be briefly described below, and it is obvious that the drawings in the following description are only some embodiments of the present invention, and it is obvious for those skilled in the art to obtain other drawings without creative efforts.
Fig. 1 is a flowchart of a mobile coal-rock real-time sorting method according to embodiment 1 of the present invention;
FIG. 2 is a flowchart of steps B1-B3 in the method provided in example 1 of the present invention;
FIG. 3 is a flowchart of step A3 in the method according to embodiment 1 of the present invention;
fig. 4 is a schematic structural diagram of a mobile coal rock real-time sorting device provided in embodiment 2 of the present invention.
Symbol interpretation:
1: a conveyor belt; 2: a hyperspectral imaging module; 3: a computer; 4: a coal rock sorting module; 5: a dust fall shower head; 21: a halogen lamp; 22: a hyperspectral camera; 41: a sorting control unit; 42: and a sorting execution unit.
Detailed Description
The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
The invention aims to provide a method and a device for sorting moving coal rocks in real time, which combine with a hyperspectral imaging technology, improve the sorting precision of the coal rocks and realize the automation of the sorting of the coal rocks.
In order to make the aforementioned objects, features and advantages of the present invention comprehensible, embodiments accompanied with figures are described in further detail below.
Example 1:
as shown in fig. 1, the invention provides a real-time mobile coal rock sorting method, which comprises the following steps:
a1, acquiring a hyperspectral image to be identified; the hyperspectral image to be identified comprises a coal block-rock block mixed flow to be sorted, which is conveyed on a conveyor belt in real time;
a2, performing category identification on each pixel in the hyperspectral image to be identified by using the trained coal-rock classification model to obtain a coal-rock classification result; the coal and rock classification result comprises a category identification result of each pixel in the hyperspectral image to be identified, and the category identification result comprises coal categories, rock categories and categories which cannot be identified;
a3, identifying the coal blocks and the rock blocks in the hyperspectral image to be identified according to the coal and rock classification result to obtain a coal block-rock block identification result; the coal block-rock block identification result comprises position information of each coal block and each rock block in the hyperspectral image to be identified;
and A4, controlling a sorting execution unit to sort the coal blocks and the rock blocks on the conveyor belt according to the coal block-rock block recognition result.
As shown in fig. 2, the method may further include:
b1, constructing a training data set; the training data set comprises training hyperspectral images, and coal blocks and rock blocks in the training hyperspectral images are marked with corresponding coal labels and rock labels;
b2, constructing the coal-rock classification model by adopting a support vector machine algorithm;
and B3, training the coal-rock classification model by using the training data set to obtain the trained coal-rock classification model.
As shown in fig. 3, in this embodiment, step a3 specifically includes:
a31, detecting the object contour in the hyperspectral image to be identified by using an edge detection algorithm;
a32, determining the classes of all pixels in the object contour according to the coal rock classification result;
a33, determining the type of the object contour according to the types of all pixels in the object contour;
and A34, determining the centroid of the object contour, and determining the centroid coordinate of the object contour as the position information of the object contour.
In a specific implementation manner, the object contour in the hyperspectral image to be identified can be detected by using a Robert edge detection algorithm.
In this embodiment, step a4 specifically includes:
a41, determining the time when the coal or rock block reaches a picking point according to the coal-rock block identification result and the running speed of the conveyor belt;
and A42, controlling the sorting execution unit to sort the coal blocks and the rock blocks according to the time required by the sorting execution unit to reach the picking point and the time required by the coal blocks or the rock blocks to reach the picking point.
In order to enable the hyperspectral image to obtain a better display effect, smoothing processing and normalization transformation processing can be carried out on the hyperspectral image to be identified.
The method proposed by the present invention is described below with reference to an example:
101. and imaging the coal blocks and the rocks to obtain a hyperspectral image.
In this example, the hyperspectral image is obtained by imaging the coal rock flow with a hyperspectral camera. The hyperspectral camera may be: near infrared and short wave infrared cameras. The coal rock stream may be a coal block, rock block, or a coal block-rock mixture stream transported on a conveyor belt.
In some embodiments of the present invention, after imaging the coal and rock with the hyperspectral camera to obtain the hyperspectral image, the hyperspectral image may be processed with viewspec software and MATLAB software, and the specific processing may be preprocessing of hyperspectral data of each pixel point of the image, for example, smoothing and normalization transform processing of the hyperspectral data. The specific steps can be as follows: and opening an image file corresponding to the hyperspectral image by using the viewspec software, converting the image file into an ASCII code form, importing the ASCII code form into an EXCEL form, and processing the EXCEL form data by using MATLAB to obtain a processed hyperspectral image.
Wherein, MATLAB is a commercial mathematical software produced by MathWorks corporation in usa, MATLAB is a combination of two words of matrix & laboratory, which means a matrix factory (matrix laboratory), and the software mainly faces a high-tech computing environment of scientific computing, visualization and interactive programming.
In some embodiments of the invention, the scheme further comprises: the hyperspectral data is preprocessed in MATLAB software by using a preset preprocessing method, and the preprocessing method can comprise smoothing processing and normalization transformation processing. It should be understood that the hyper-spectral image is subjected to smoothing processing and normalized transformation processing, so that a better display effect of the hyper-spectral image can be obtained.
The hyperspectral camera is equipment which can image a target on hundreds of imaging channels through a hyperspectral technology, wherein one imaging channel corresponds to one wave band; the hyperspectral camera has the advantages of small influence by the surface moisture of an object, small volume and convenient installation and debugging, so that the hyperspectral camera can be conveniently operated under or on a coal mine.
Because the hyperspectral camera is provided with hundreds of imaging channels, one imaging channel corresponds to one wave band, all the channels cannot be used when imaging the coal blocks and the rocks, and partial channels can be selected for imaging. In some embodiments of the present application, the partial imaging channel in the hyperspectral image includes: imaging channels corresponding to a plurality of wave bands of the coal and the rock can be distinguished, and the wave bands correspond to the imaging channels one to one. In this kind of embodiment, the imaging channel is the imaging channel that a plurality of wave bands that can distinguish coal and rock correspond, explains that the imaging channel has the effect of discernment coal and rock, and the high spectral image that obtains based on this kind of imaging channel reuses the target classifier and classifies coal and rock, can further improve the discernment accuracy, promotes coal petrography separation efficiency.
Specifically, one way to determine the imaging channel may include the following operations: step 1, imaging the coal briquette by using all imaging channels in a hyperspectral camera to obtain a hyperspectral image of the coal briquette; step 2, imaging the rock by using all imaging channels in the hyperspectral camera to obtain a hyperspectral image of the rock; step 3, comparing and classifying the hyperspectral images of the coal blocks of the same imaging channel with the hyperspectral images of the rocks to obtain a classification precision table; and 4, determining the imaging channel with the highest precision in the classification precision table as a target imaging channel.
102. And performing coal-rock classification on the hyperspectral image based on a coal-rock classification model obtained by pre-training to obtain the type of each pixel in the hyperspectral image.
The coal-rock classification model is obtained based on training of a pre-labeled training sample, the training sample comprises coal flow images obtained by imaging of a plurality of hyperspectral cameras in a mining area, and the types of the coal-rock classification model comprise coal blocks, rocks and belts. As a better two-classification model at present, a coal-rock classification model adopts a Support Vector Machine (SVM) for classification. Taking the Viewspec software and the MATLAB software as examples, ASCII code conversion is performed on a target image in the Viewspec software, SVM recognition is performed on the ASCII code of the target image by using the MATLAB software, three categories of coal blocks, rocks and belts are established, and an SVM is used for recognition according to the divided categories.
It should be understood that each pixel point in the hyperspectral image is classified by using the coal-rock classification model, and the category to which each pixel point belongs is determined to be coal, rock or belt. Of course, if there is a pixel that is not in the above three categories, the pixel is marked as a category that cannot be identified.
In addition, it is easy to understand that the coal-rock classification model can only identify the position information of one of the coal block and the rock block or can identify both the coal block and the rock block at the same time according to the requirements of different application scenarios. For example, in a coal preparation plant, the coal blocks need to be continuously conveyed to the coal bunker on the coal flow conveyor belt, and at the moment, the purpose of sorting the rock blocks from the coal flow conveyor belt can be achieved only by identifying the rock blocks and the position information of the rock blocks.
103. And acquiring the position information of the object profile based on the object profile and the type information of the object to be sorted.
After the classification of each pixel point is obtained by using the coal-rock classification model, all the pixel points of the classification of the coal and all the pixel points of the classification of the rock in the hyperspectral image are obtained. Therefore, the coal-rock classification model can obtain the classification result, namely the distribution of the coal blocks and the rock blocks in the hyperspectral image.
And performing edge detection on the hyperspectral image and the outline of the object to be identified by adopting a Robert edge detection algorithm to obtain the outline of the object to be identified and determine the position areas of the coal blocks and the rock blocks in the hyperspectral image.
And obtaining the mass centers of the coal blocks and the rock blocks by using a mass center method, determining mass center coordinates, and taking the mass center coordinates as the position information of the coal blocks and the rock blocks in the target image.
After the identification result of the coal and/or rock is obtained, the coal and/or rock needs to be positioned so as to facilitate the subsequent sorting step. Firstly, the outline of the coal block and the rock is obtained by Robert edge detection, and the position area of the coal block and the rock 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 the rock blocks in the target image. This allows for the positioning of the coal and/or rock in the target image.
104. And controlling the mechanical arm to perform sorting operation according to the position information of the coal blocks and/or the rock blocks so as to sort the coal blocks and the rock blocks.
In this step, it may be selected to sort the coal blocks or the rock blocks individually, that is, to decide whether the robot arm grabs the coal blocks or the rock blocks, for example, if we let the robot arm grab the rock blocks, the coordinate position of the identified rock block needs to be sent to the robot arm, and then the robot arm receives the coordinate position and converts it into the corresponding coordinate position in the coordinate system of the robot arm itself. And then, giving an advanced coordinate deviation delta X to the mechanical arm according to the moving speed of the belt and the time required by the mechanical arm to move to the designated grabbing position, and controlling the mechanical arm to grab according to the coordinate position and the coordinate deviation delta X. For example: if a rock block is identified, the pixel coordinate position of the rock block in the hyperspectral image is (X1, Y1), then the mechanical arm converts the pixel coordinate position into the coordinate (X1, Y1) of the coordinate system of the mechanical arm, then the moving time of the belt and the arm is considered comprehensively, and finally the mechanical arm needs to move from the initial position to the coordinate (X1+ delta X, Y1) to execute the grabbing action, wherein the abscissa X of the coordinate represents the moving direction of the belt, and the ordinate Y represents the vertical direction of the moving direction of the belt. The mechanical arm grabs the rock to the belt of the running rock, and the coal is still left on the coal flow belt, so that the task of sorting the coal and the rock by using the mechanical arm once is completed.
In the embodiment, different spectral characteristic information is shown due to different components of the coal briquette and the rock and different absorption of different components to the spectrum, so that the coal briquette and the rock can be distinguished according to the target imaging channel determined according to the spectral image characteristic information of the coal briquette and the rock; secondly, recognizing the hyperspectral images of the coal flow by using a target recognition model, wherein the target recognition model is obtained by performing data training on the hyperspectral images of the coal blocks and the rocks, and recognizing the position information of the coal blocks and the rocks by using a big data mode; consequently, use the different characteristics of spectral image characteristic that utilizes both to demonstrate of coal cinder and rock in this application to the mode of training with a large amount of data is discerned coal cinder and rock, can effectually avoid under the coal stream environment the influence of environmental factor such as dust and moisture, thereby accurately discerning coal cinder and rock, improves the degree of accuracy that coal cinder and rock were selected separately, further improves the efficiency that coal cinder and rock were selected separately.
Example 2:
as shown in fig. 4, the present invention further provides a mobile coal rock real-time sorting apparatus, including:
a conveyor belt 1 for conveying a coal-rock mixed flow to be sorted;
the hyperspectral imaging module 2 is fixedly arranged above the conveyor belt 1 and used for acquiring a hyperspectral image to be identified; the hyperspectral image to be identified comprises a coal block-rock block mixed flow to be sorted, which is conveyed on the conveyor belt 1 in real time;
the computer 3 is in signal connection with the hyperspectral imaging module 2 and is used for receiving the hyperspectral image to be identified and identifying the type of each pixel in the hyperspectral image to be identified by using a trained coal-rock classification model to obtain a coal-rock classification result; the coal and rock classification result comprises a category identification result of each pixel in the hyperspectral image to be identified, and the category identification result comprises coal categories, rock categories and categories which cannot be identified; the hyperspectral image to be identified is also used for identifying the coal blocks and the rock blocks in the hyperspectral image to be identified according to the coal and rock classification result to obtain a coal block-rock block identification result; the coal block-rock block identification result comprises position information of each coal block and each rock block in the hyperspectral image to be identified;
and the coal and rock sorting module 4 is in signal connection with the computer 3 and is used for acquiring the coal and rock recognition result and sorting the coal and rock on the conveyor belt 1 according to the coal and rock recognition result.
In the present embodiment, the hyperspectral imaging module 2 includes:
a halogen lamp 21 for generating a light source;
and the hyperspectral camera 22 is used for acquiring a hyperspectral image to be identified.
In this embodiment, the apparatus further includes: dust fall gondola water faucet 5 sets up 1 top of conveyer belt is connected with the 3 electricity of computer, is used for right dust on the conveyer belt 1 suppresses.
In this embodiment, the coal rock sorting module 4 includes:
a sorting control unit 41, configured to determine, according to the coal block-rock block identification result and the running speed of the conveyor belt 1, a time when the coal block or rock block reaches a picking point; and is also used for generating a sorting control instruction according to the time required by the sorting execution unit 42 to reach the picking point and the time of the coal or rock block to reach the picking point;
and the sorting execution unit 42 is in control connection with the sorting control unit 41 and is used for sorting the coal blocks and the rock blocks on the conveyor belt 1 according to the sorting control instruction.
The principles and embodiments of the present invention have been described herein using specific examples, which are provided only to help understand the method and the core concept of the present invention; meanwhile, for a person skilled in the art, according to the idea of the present invention, the specific embodiments and the application range may be changed. In view of the above, the present disclosure should not be construed as limiting the invention.
Claims (10)
1. A real-time mobile coal rock sorting method is characterized by comprising the following steps:
acquiring a hyperspectral image to be identified; the hyperspectral image to be identified comprises a coal block-rock block mixed flow to be sorted, which is conveyed on a conveyor belt in real time;
performing category identification on each pixel in the hyperspectral image to be identified by using the trained coal-rock classification model to obtain a coal-rock classification result; the coal and rock classification result comprises a category identification result of each pixel in the hyperspectral image to be identified, and the category identification result comprises coal categories, rock categories and categories which cannot be identified;
identifying the coal blocks and the rock blocks in the hyperspectral image to be identified according to the coal rock classification result to obtain a coal block-rock block identification result; the coal block-rock block identification result comprises position information of each coal block and each rock block in the hyperspectral image to be identified;
and controlling a sorting execution unit to sort the coal blocks and the rock blocks on the conveyor belt according to the coal block-rock block recognition result.
2. The method of claim 1, further comprising:
constructing a training data set; the training data set comprises training hyperspectral images, and coal blocks and rock blocks in the training hyperspectral images are marked with corresponding coal labels and rock labels;
constructing the coal-rock classification model by adopting a support vector machine algorithm;
and training the coal-rock classification model by using the training data set to obtain the trained coal-rock classification model.
3. The method according to claim 1, wherein the identifying the coal blocks and the rock blocks in the hyperspectral image to be identified according to the coal rock classification result to obtain a coal block-rock block identification result specifically comprises:
detecting the object outline in the hyperspectral image to be identified by utilizing an edge detection algorithm;
determining the categories of all pixels in the object contour according to the coal rock classification result;
determining the type of the object contour according to the types of all pixels in the object contour;
and determining the centroid of the object contour, and determining the coordinates of the centroid as the position information of the object contour.
4. The method according to claim 3, characterized in that the object contour in the hyperspectral image to be identified is detected using a Robert edge detection algorithm.
5. The method according to claim 1, wherein the controlling the sorting execution unit to sort the coal blocks and the rock blocks on the conveyor belt according to the coal block-rock block recognition result specifically comprises:
determining the time when the coal or rock reaches a picking point according to the coal-rock recognition result and the running speed of the conveyor belt;
and controlling the sorting execution unit to sort the coal blocks and the rock blocks according to the time required by the sorting execution unit to reach the picking point and the time required by the coal blocks or the rock blocks to reach the picking point.
6. The method of claim 1, further comprising: and performing smoothing processing and normalized transformation processing on the hyperspectral image to be identified.
7. The utility model provides a remove real-time sorting device of coal petrography, its characterized in that the device includes:
a conveyor belt for conveying a coal-rock mixed stream to be sorted;
the hyperspectral imaging module is fixedly arranged above the conveyor belt and used for acquiring a hyperspectral image to be identified; the hyperspectral image to be identified comprises a coal block-rock block mixed flow to be sorted, which is conveyed on the conveyor belt in real time;
the computer is in signal connection with the hyperspectral imaging module and is used for receiving the hyperspectral image to be recognized and performing type recognition on each pixel in the hyperspectral image to be recognized by using the trained coal-rock classification model to obtain a coal-rock classification result; the coal and rock classification result comprises a category identification result of each pixel in the hyperspectral image to be identified, and the category identification result comprises coal categories, rock categories and categories which cannot be identified; the hyperspectral image to be identified is also used for identifying the coal blocks and the rock blocks in the hyperspectral image to be identified according to the coal and rock classification result to obtain a coal block-rock block identification result; the coal block-rock block identification result comprises position information of each coal block and each rock block in the hyperspectral image to be identified;
and the coal and rock sorting module is in signal connection with the computer and is used for acquiring the coal block-rock block identification result and sorting the coal blocks and the rock blocks on the conveyor belt according to the coal block-rock block identification result.
8. The apparatus of claim 7, wherein the hyperspectral imaging module comprises:
a halogen lamp for generating a light source;
and the hyperspectral camera is used for acquiring a hyperspectral image to be identified.
9. The apparatus of claim 7, further comprising: the dust fall gondola water faucet sets up the conveyer belt top is connected with computer electricity, is used for right dust on the conveyer belt suppresses.
10. The apparatus of claim 7, wherein the coal-rock sorting module comprises:
the sorting control unit is used for determining the time when the coal blocks or the rock blocks reach the picking point according to the coal block-rock block recognition result and the running speed of the conveyor belt; the coal or rock mass picking device is also used for generating a sorting control instruction according to the time required by the sorting execution unit to reach the picking point and the time required by the coal or rock mass to reach the picking point;
and the sorting execution unit is in control connection with the sorting control unit and is used for sorting the coal blocks and the rock blocks on the conveying belt according to the sorting control instruction.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210347212.1A CN114677546A (en) | 2022-04-01 | 2022-04-01 | Real-time mobile coal rock sorting method and device |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210347212.1A CN114677546A (en) | 2022-04-01 | 2022-04-01 | Real-time mobile coal rock sorting method and device |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN114677546A true CN114677546A (en) | 2022-06-28 |
Family
ID=82075264
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202210347212.1A Pending CN114677546A (en) | 2022-04-01 | 2022-04-01 | Real-time mobile coal rock sorting method and device |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN114677546A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115646848A (en) * | 2022-10-21 | 2023-01-31 | 微山金源煤矿 | Intelligent classification method and system for coal mine |
-
2022
- 2022-04-01 CN CN202210347212.1A patent/CN114677546A/en active Pending
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115646848A (en) * | 2022-10-21 | 2023-01-31 | 微山金源煤矿 | Intelligent classification method and system for coal mine |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Li et al. | Coal gangue detection and recognition algorithm based on deformable convolution YOLOv3 | |
| CN106548182A (en) | Based on deep learning and the causal analytic pavement distress survey method and device of master | |
| CN112893159B (en) | Coal gangue sorting method based on image recognition | |
| CN107891015A (en) | Bastard coal infrared identification device in a kind of bastard coal sorting | |
| CN113145492A (en) | Visual grading method and grading production line for pear appearance quality | |
| CN103785622A (en) | Machine-vision-based part sorting device and method | |
| CN114677546A (en) | Real-time mobile coal rock sorting method and device | |
| CN111161292A (en) | Ore size measurement method and application system | |
| CN116665011A (en) | Coal flow foreign matter identification method for coal mine belt conveyor based on machine vision | |
| CN115376130A (en) | Method, device, equipment and medium for generating electric power construction site identification model | |
| CN114535133B (en) | Method, device and system for sorting coal and gangue based on dual-energy ray transmission imaging | |
| CN107891012B (en) | Pearl size and circularity sorting device based on equivalent algorithm | |
| CN116309191A (en) | Intelligent gas inspection display method based on image enhancement | |
| Cao et al. | Smaller target detection algorithms based on YOLOv5 in safety helmet wearing detection | |
| CN106944362B (en) | A kind of intelligent classification reminding method, apparatus and system based on video image identification | |
| Zhao et al. | Image processing based on gray information in sorting system of coal gangue | |
| CN116067911A (en) | Mineral multicomponent grade identification and separation method | |
| CN112183374A (en) | Automatic express sorting device and method based on raspberry group and deep learning | |
| CN116740499A (en) | Coal gangue identification method based on infrared thermal imaging technology | |
| Prasetio | An Approaching Machine Learning Model: Tile Inspection Case Study | |
| CN111199250A (en) | Transformer substation air switch state checking method and device based on machine learning | |
| CN111723861B (en) | Gangue percentage identification method for coal mine control baffle | |
| CN114092410A (en) | Mobile phone surface defect accurate identification method based on multi-scale convolutional neural network | |
| CN107330470A (en) | The method and apparatus for recognizing picture | |
| CN113569943A (en) | Deep neural network-based slag piece bulk early warning method, system and device |
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 |