CN114887924A - Terahertz imaging-based coal gangue identification and sorting device and method - Google Patents
Terahertz imaging-based coal gangue identification and sorting device and method Download PDFInfo
- Publication number
- CN114887924A CN114887924A CN202210470859.3A CN202210470859A CN114887924A CN 114887924 A CN114887924 A CN 114887924A CN 202210470859 A CN202210470859 A CN 202210470859A CN 114887924 A CN114887924 A CN 114887924A
- Authority
- CN
- China
- Prior art keywords
- coal
- gangue
- terahertz
- image
- sorting
- 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
-
- 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
-
- B—PERFORMING OPERATIONS; TRANSPORTING
- B07—SEPARATING SOLIDS FROM SOLIDS; SORTING
- B07B—SEPARATING SOLIDS FROM SOLIDS BY SIEVING, SCREENING, SIFTING OR BY USING GAS CURRENTS; SEPARATING BY OTHER DRY METHODS APPLICABLE TO BULK MATERIAL, e.g. LOOSE ARTICLES FIT TO BE HANDLED LIKE BULK MATERIAL
- B07B1/00—Sieving, screening, sifting, or sorting solid materials using networks, gratings, grids, or the like
- B07B1/28—Moving screens not otherwise provided for, e.g. swinging, reciprocating, rocking, tilting or wobbling screens
-
- B—PERFORMING OPERATIONS; TRANSPORTING
- B07—SEPARATING SOLIDS FROM SOLIDS; SORTING
- B07B—SEPARATING SOLIDS FROM SOLIDS BY SIEVING, SCREENING, SIFTING OR BY USING GAS CURRENTS; SEPARATING BY OTHER DRY METHODS APPLICABLE TO BULK MATERIAL, e.g. LOOSE ARTICLES FIT TO BE HANDLED LIKE BULK MATERIAL
- B07B15/00—Combinations of apparatus for separating solids from solids by dry methods applicable to bulk material, e.g. loose articles fit to be handled like bulk material
-
- 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
-
- 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
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/20—Image preprocessing
- G06V10/30—Noise filtering
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/40—Extraction of image or video features
- G06V10/44—Local feature extraction by analysis of parts of the pattern, e.g. by detecting edges, contours, loops, corners, strokes or intersections; Connectivity analysis, e.g. of connected components
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/40—Extraction of image or video features
- G06V10/50—Extraction of image or video features by performing operations within image blocks; by using histograms, e.g. histogram of oriented gradients [HoG]; by summing image-intensity values; Projection analysis
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/40—Extraction of image or video features
- G06V10/54—Extraction of image or video features relating to texture
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/70—Arrangements for image or video recognition or understanding using pattern recognition or machine learning
- G06V10/764—Arrangements for image or video recognition or understanding using pattern recognition or machine learning using classification, e.g. of video objects
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/70—Arrangements for image or video recognition or understanding using pattern recognition or machine learning
- G06V10/77—Processing image or video features in feature spaces; using data integration or data reduction, e.g. principal component analysis [PCA] or independent component analysis [ICA] or self-organising maps [SOM]; Blind source separation
- G06V10/774—Generating sets of training patterns; Bootstrap methods, e.g. bagging or boosting
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Physics & Mathematics (AREA)
- Multimedia (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Evolutionary Computation (AREA)
- Artificial Intelligence (AREA)
- Computing Systems (AREA)
- General Health & Medical Sciences (AREA)
- Medical Informatics (AREA)
- Databases & Information Systems (AREA)
- Software Systems (AREA)
- Data Mining & Analysis (AREA)
- Health & Medical Sciences (AREA)
- Bioinformatics & Computational Biology (AREA)
- General Engineering & Computer Science (AREA)
- Evolutionary Biology (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Life Sciences & Earth Sciences (AREA)
- Investigating Or Analysing Materials By Optical Means (AREA)
Abstract
The invention discloses a coal gangue identifying and sorting device and a method based on terahertz imaging, wherein a material conveying device matched with a vibrating screening device is arranged on one side of the device through the vibrating screening device, a terahertz light source and a terahertz linear array camera which are matched with each other are respectively arranged at the upper position and the lower position of the material conveying device, a queuing device matched with the material conveying device is arranged at the input end of the material conveying device, a sorting device matched with the material conveying device is arranged at the output end of the material conveying device, the terahertz linear array camera and the sorting device are respectively connected with a computer in a linear way, and through an image processing technology in the field of computer machine vision, the invention provides an algorithm for extracting gray scale and texture characteristics of coal and gangue in an original terahertz image, achieves the aim of identifying and sorting coal gangue by utilizing machine learning of a support vector machine, and can safely, effectively and comprehensively identify and sort the coal gangue, the accuracy of coal and gangue identification and the efficiency of separation are further improved.
Description
Technical Field
The invention relates to the technical field of coal gangue sorting, in particular to a coal gangue identifying and sorting device and method based on terahertz imaging.
Background
Coal is the most abundant and widely distributed energy source in the world. In recent years, the demand of China for coal is greatly increased. The coal gangue is a black and gray rock with low carbon content associated with a coal bed in the coal forming process, and is solid waste discharged in the coal mining and coal washing processes. Therefore, before coal is utilized, the coal gangue must be accurately sorted. The traditional coal and gangue separation method comprises manual separation and mechanical separation. The manual sorting method has poor operating environment and low production efficiency; the mechanical separation equipment is huge, the process is complex, and the water resource waste is serious. With the rapid development of computer technology and machine vision, the photoelectric sorting technology is gradually popularized and applied, and mainly has two technical means, one is to adopt an industrial CCD camera to acquire images of coal and gangue, and the other is based on an X-ray transmission imaging technology, but the two methods have the problems of weak penetrating power and high radiation to massive gangue and can not ensure the health of a person who sends and sorts the gangue.
At present, terahertz imaging technology is more and more mature, terahertz imaging is a novel nondestructive testing technology, terahertz waves are non-ionizing radiation electromagnetic waves with the frequency of 0.1 to 10THz, and the terahertz imaging device has the advantages of being strong in penetrating capability, free of radiation to a human body and the like, so that the terahertz imaging device has better and safer effect than X-rays. The terahertz imaging technology can effectively solve the problems of high radiation and high cost in the existing coal and gangue sorting technology. However, the technology of identifying coal gangue by using terahertz waves in the current research is few, because the principle of terahertz imaging is greatly different from that of ordinary optical imaging, the existing terahertz image processing and identifying method adopts a general image processing and identifying method, optical images generated by ordinary optical photographing equipment have rich color and texture characteristics, but terahertz imaging is different, terahertz images often have no rich information, information which can be fully utilized as characteristics in the images often only has gray information and contour information, and the current method for researching gray scale and contour characteristics is more common. How to extract the characteristics of coal gangue from the terahertz image becomes important to realize coal preparation automation.
Disclosure of Invention
The invention aims to provide a coal gangue identification and sorting device and method based on terahertz imaging, an algorithm is provided to extract the gray level and texture characteristics of coal and gangue in an original terahertz image through an image processing technology in the field of computer machine vision, and the aim of identifying and sorting the coal gangue is fulfilled by utilizing machine learning of a support vector machine so as to solve the problems provided in the background technology.
In order to achieve the purpose, the invention is realized by the following technical means:
the utility model provides a gangue discernment sorting unit based on terahertz is imaged now, is including vibration screening device, vibration screening device one side is provided with vibration screening device complex material conveying device, material conveying device's upper and lower position is provided with terahertz light source and terahertz linear array camera of mutually supporting respectively, material conveying device's input is provided with the device of lining up with material conveying device matched with, material conveying device's output is provided with rather than complex sorting unit, terahertz linear array camera and sorting unit respectively with computer linear connection.
Furthermore, a pulverized coal belt matched with the vibrating screening device is arranged below the vibrating screening device, raw coal is crushed to form small blocks (the thickness is less than or equal to 10cm) of coal or coal gangue, the vibrating screening device receives a coal-gangue mixed material formed by crushing the raw coal, the crushed coal with small thickness is screened, and the pulverized coal belt is arranged below the vibrating screening device to receive screened pulverized coal.
Further, dust collector is all being connected to vibration screening plant, material transfer device, sorting unit, dust collector one side is provided with the protector with vibration screening plant complex, the dust collector lower part is equipped with waste rock belt, the clean coal belt that cooperatees with material transfer device and set up respectively, and clean coal belt, waste rock belt are located the material blanking point below of material transfer device coal, waste rock, receive the material and convey.
The invention discloses a method for identifying and sorting the mixed materials of coal and waste rock, which comprises the steps that a material conveying device receives clean coal and waste rock screened by a vibration screening device, a queuing device is fixed on the material conveying device, the mixed coal and waste rock on the material conveying device are orderly arranged without overlapping, the distance interval of the coal and the waste rock is controlled, the sorting is convenient later, the mixed materials of coal and waste rock are ensured to uniformly and equidistantly pass through a terahertz wave imaging area, a terahertz light source and a terahertz linear array camera are respectively arranged above and below the material conveying device, the terahertz linear camera is connected with a computer, the terahertz light source is used for polishing the coal or waste rock of the material conveying device, the terahertz light source is adopted, the penetrability is strong, no radiation is generated to the human body, the health of sorting personnel is ensured, terahertz images of the coal or waste rock collected by the terahertz linear camera are transmitted to the computer in real time, and the sorting device identifies and sorts the mixed materials of coal and waste rock on the material conveying device through data information sent by an image processing module of the computer, the clean coal belt and the gangue belt are positioned below the material blanking points of the coal and the gangue on the material conveying device, and are used for receiving and conveying the sorted materials.
A coal gangue identification and sorting method based on terahertz imaging comprises the following steps:
s1: selecting a plurality of coal blocks and gangue in the crushed raw coal, obtaining 7 characteristic quantities of the coal and gangue to form a training set of the coal and gangue according to the characteristic extraction step, selecting an optimal characteristic combination according to the linear dispersion of each characteristic, and training a classification model according to the optimal characteristic combination;
s2: establishing a terahertz image database of coal and gangue, labeling the terahertz images of the coal and the gangue, selecting a Gaussian kernel function to train on a training set, and testing the trained classification model on a testing set for identifying actual coal gangue;
s3: the sorting device positions and identifies and sorts the coal and the gangue on the material conveying device through data information sent by the computer image processing module.
Further, the step S3, in which the sorting device locates the coal and the gangue, is to locate the coal and the gangue by obtaining image centroid positions of the coal and the gangue, wherein a centroid position coordinate extraction formula is as follows:
wherein I (x, y) represents the pixel value of image I at x, y,is the centroid coordinate and S is the target area.
Further, the sorting device in S3 can perform secondary sorting in the sorting process, so that the coal can be further identified and sorted from the gangue, and the aim of recovering clean coal is fulfilled.
Further, the feature extraction step in the step S1 includes the following steps:
a1: acquiring terahertz images of original coal and gangue from a terahertz imaging system, and preprocessing the original terahertz images to obtain denoised terahertz images;
a2: carrying out edge detection on the denoised terahertz images of the coal and the gangue so as to facilitate feature extraction and classification identification;
a3: and extracting features according to the difference of the terahertz image gray level histograms of the denoised coal and gangue.
Further, in the step a1, median filtering is adopted to perform denoising processing on the original terahertz image, so as to eliminate the influence of pulse noise on the image and improve the definition of the terahertz image, and the median filtering denoising processing formula is as follows:
f(x,y)=median[g(k,l)](k,l)∈S xy
wherein g (x, y) is the image after median filtering denoising, f (x, y) is the original image before denoising, and S XY A template filter window.
Further, in the step a2, a Sobel operator is adopted to perform edge detection on the denoised terahertz image of the coal and the gangue, the Sobel operator is a discrete differential operator mainly used for edge detection, and performs weighting smoothing processing on the image according to gray level weighting differences of upper, lower, left and right adjacent points of a pixel point, and then performs differential operation.
Further, the step of extracting features through the difference of the gray histogram in the step a3 is as follows: intercepting the terahertz image of each coal and gangue into an image with 256 x 256 pixels, processing and analyzing the image by utilizing Matlab software, and statistically analyzing the gray characteristic quantity of the coal and gangue 2 To represent the gray level characteristics of coal and gangue, the gray level characteristic extraction formula is as follows:
wherein g (x, y) is the gray scale of the terahertz image, and the length and the width of the image are both N.
Furthermore, in order to further effectively and comprehensively perform image processing analysis on the coal and the gangue, the method increases the extraction of texture characteristic parameters of the terahertz images of the denoised coal and gangue, wherein the texture is the attribute which is formed by the repeated and alternate change of gray level distribution on a spatial position and reflects the gray level distribution of a certain area in the images, the method adopts the theory based on a gray level co-occurrence matrix to research the extraction of the texture characteristics of the coal and the gangue, the energy, contrast, correlation, uniformity and entropy can be extracted through the gray level co-occurrence matrix of the coal and the gangue as the texture characteristic parameters, and the gray level co-occurrence matrix P (i, j | d, theta) refers to the probability that a point of a gray level i reaches a gray level j at a specific position d (delta x, delta y);
wherein the content of the first and second substances,(x, y) is the coordinate of the pixel point, and the range of the delta x and the delta y is determined by two parameters of the pixel distance d and the angle theta; i and j are respectively used for expressing the gray scale of the pixel; d refers to the direction and distance between two different pixels; theta refers to a corresponding direction of the gray level co-occurrence matrix generation. According to the characteristics of coal and gangue, the angle theta is selected from four angles of 0 degree, 45 degrees, 90 degrees and 135 degrees, the characteristic values of the four directions are averaged, and the distance d is 1.
Further, energy: a measure of the uniformity of the gray level distribution of the image, and the thickness of the texture, reflects the uniformity of the gray level distribution in the image area,
further, contrast: a measure of the local variation present in the image, reflecting the sharpness of the image and the depth of the grooves of the texture,
further, the correlation: a measure of how similar the grey levels of an image are in the row or column direction, reflects how similar the image is,
further, uniformity: a measure of smoothness of the image distribution reflects the sharpness and regularity of the texture in the image
Further, entropy: a measure of the randomness of the amount of information contained in an image reflects the complexity of the gray level distribution of the image, with larger entropy values, more complex images,
further, the invention establishes a training set and a test set of coal and gangue identification characteristics and a classification identification model thereof through characteristic extraction, and specifically adopts a classification model of a Support Vector Machine (SVM) classifier, wherein the SVM classifier is a generalized linear classifier which carries out binary classification on data according to a supervised learning mode, the essence of the SVM is a method for quantifying the difference of two types of data, and the decision hyperplane and kernel function calculation formulas of the SVM model adopted by the invention are respectively as follows:
compared with the prior art, the invention has the following beneficial effects:
1. according to the invention, the terahertz image of coal and gangue is obtained based on the terahertz imaging technology and the gangue is identified, so that the defects of the existing gangue identification technology are effectively overcome, and the terahertz light source is non-radiative to human bodies, so that effective guarantee is provided for the health of sorting personnel;
2. the method comprises the steps of extracting corresponding characteristics of coal and gangue, establishing a training set and a test set of coal and gangue identification characteristics and a classification identification model of the coal and gangue, and processing corresponding data of the coal and the gangue through the obtained corresponding training set and the trained classification model by matching a corresponding sorting device with a computer to realize identification and sorting;
3. the coal and gangue identifying and sorting method can also effectively identify and sort small-block coal and gangue;
4. the sorting device can be combined with a computer image processing module to carry out secondary sorting, and further identifies and sorts out coal from sorted gangue, so that secondary recycling of coal is realized.
Description of the drawings:
FIG. 1 is a schematic diagram of the composition of a coal gangue identifying and sorting device of the invention;
FIG. 2 is a flow chart of a coal gangue identification method of the present invention;
FIG. 3 is a flow chart of a terahertz imaging recognition algorithm of the present invention;
fig. 4 is a decision plane diagram of the SVM of the present invention.
Wherein: 1-raw coal inlet, 2-gangue, 3-coal, 4-protective device, 5-queuing device, 6-terahertz light source, 7-dust removing device, 8-gangue belt, 9-clean coal belt, 10-sorting device, 11-terahertz linear array camera, 12-computer, 13-material conveying device, 14-vibration screening device and 15-pulverized coal belt.
The specific implementation mode is as follows:
in order to make the technical solutions of the present invention better understood by those skilled in the art, the present invention is further described below with reference to the following examples:
DETAILED DESCRIPTION OF EMBODIMENT (S) OF INVENTION
The invention relates to a method for preparing a composite material, which comprises the following steps: see the drawings.
In this embodiment, a coal gangue identification sorting device based on terahertz imaging, including vibration screening device 14, vibration screening device 14 one side is provided with material conveying device 13 with vibration screening device 14 complex, the upper and lower position of material conveying device 13 is provided with terahertz light source 6 and terahertz linear array camera 11 of mutually supporting respectively, the input of material conveying device 13 is provided with the device of lining up 5 with material conveying device 13 matched with, the output of material conveying device 13 is provided with rather than complex sorting unit 10, terahertz linear array camera 11 and sorting unit 10 respectively with computer 12 linear connection.
In the embodiment of the invention, a pulverized coal belt 15 matched with the vibrating screening device 14 is arranged below the vibrating screening device 14, raw coal is crushed to form small blocks (the thickness is less than or equal to 10cm) of coal or coal gangue, a mixed material of coal 3 and coal gangue 2 formed by crushing the raw coal is received through a raw coal inlet 1 of the vibrating screening device 14 for screening, and crushed coal with smaller thickness is removed by screening, and the pulverized coal belt 15 is arranged below the vibrating screening device 14 for receiving screened pulverized coal.
In the embodiment of the invention, the vibrating screening device 14, the material conveying device 13 and the sorting device 10 are all connected with the dust removing device 7, one side of the dust removing device 7 is provided with the protective device 4 matched with the vibrating screening device 14, the lower part of the dust removing device 7 is respectively provided with the gangue belt 8 and the clean coal belt 9 matched with the material conveying device 13, and the clean coal belt 9 and the gangue belt 8 are positioned below the material dropping point of coal and gangue on the material conveying device 13, receive and convey the material.
In the invention, a material conveying device 13 receives coal 3 and waste rock 2 screened by a vibrating screening device 14, a queuing device 5 is fixed on the material conveying device 13, the disordered coal 3 and waste rock 2 on the material conveying device 13 are orderly arranged without overlapping, the distance interval between the coal 3 and the waste rock 2 is controlled, the subsequent sorting is convenient, the mixed coal and waste rock material is ensured to pass through a terahertz wave imaging area at a constant speed and an equal distance, a terahertz light source 6 and a terahertz linear array camera 11 are respectively arranged above and below the material conveying device 13, the terahertz linear camera 11 is connected with a computer 12, the terahertz light source 6 polishes the coal or the waste rock of the material conveying device, the terahertz light source is adopted, the penetrability is strong, no radiation is generated to a human body, the health of sorting personnel is ensured, the terahertz image of the coal or the waste rock collected by the terahertz linear camera 11 is transmitted to the computer 12 in real time, the sorting device 10 identifies and sorts the coal-gangue mixture on the material conveying device 13 through data information sent by the computer image processing module, and the clean coal belt 9 and the gangue belt 8 are positioned below material dropping points of coal and gangue on the material conveying device 13, receive and convey the sorted materials.
A coal gangue identification and sorting method based on terahertz imaging comprises the following steps:
s1: selecting a plurality of coal blocks and gangue in the crushed raw coal, obtaining 7 characteristic quantities of the coal and gangue to form a training set of the coal and gangue according to the characteristic extraction step, selecting an optimal characteristic combination according to the linear dispersion of each characteristic, and training a classification model according to the optimal characteristic combination;
s2: establishing a terahertz image database of coal and gangue, labeling the terahertz images of the coal and the gangue, selecting a Gaussian kernel function to train on a training set, and testing the trained classification model on a testing set for identifying actual coal gangue;
s3: the sorting device positions and identifies and sorts the coal and the gangue on the material conveying device through data information sent by the computer image processing module.
Further, the step S3, in which the sorting device locates the coal and the gangue, is to locate the coal and the gangue by obtaining image centroid positions of the coal and the gangue, wherein a centroid position coordinate extraction formula is as follows:
wherein I (x, y) represents the pixel value of the image I at x, y,is the centroid coordinate and S is the target area.
Further, the sorting device in S3 can perform secondary sorting in the sorting process, so that the coal can be further identified and sorted from the gangue, and the aim of recovering clean coal is fulfilled.
Further, the feature extraction step in the step S1 includes the following steps:
a1: acquiring terahertz images of original coal and gangue from a terahertz imaging system, and preprocessing the original terahertz images to obtain denoised terahertz images;
a2: carrying out edge detection on the denoised terahertz images of the coal and the gangue so as to facilitate feature extraction and classification identification;
a3: and extracting features according to the difference of the terahertz image gray level histograms of the denoised coal and gangue.
Further, in the step a1, median filtering is adopted to perform denoising processing on the original terahertz image, so as to eliminate the influence of pulse noise on the image and improve the definition of the terahertz image, and the median filtering denoising processing formula is as follows:
f(x,y)=median[g(k,l)](k,l)∈S xy
wherein g (x, y) is the image after median filtering denoising treatment, f (x, y) is the original image before denoising, S XY Is a template filter window.
Further, in the step a2, a Sobel operator is adopted to perform edge detection on the denoised terahertz image of the coal and the gangue, the Sobel operator is a discrete differential operator mainly used for edge detection, and performs weighting smoothing processing on the image according to gray level weighting differences of upper, lower, left and right adjacent points of a pixel point, and then performs differential operation.
Further, the step of extracting features through the difference of the gray histogram in the step a3 is as follows: intercepting the terahertz image of each coal and gangue into an image with 256 x 256 pixels, processing and analyzing the image by utilizing Matlab software, and statistically analyzing the gray characteristic quantity of the coal and gangue 2 To represent the gray level characteristics of coal and gangue, the gray level characteristic extraction formula is as follows:
wherein g (x, y) is the gray scale of the terahertz image, and the length and the width of the image are both N.
Furthermore, in order to further effectively and comprehensively perform image processing analysis on the coal and the gangue, the method increases the extraction of texture characteristic parameters of the terahertz images of the denoised coal and gangue, wherein the texture is the attribute which is formed by the repeated and alternate change of gray level distribution on a spatial position and reflects the gray level distribution of a certain area in the images, the method adopts the theory based on a gray level co-occurrence matrix to research the extraction of the texture characteristic of the coal and the gangue, the energy, contrast, correlation, uniformity and entropy can be extracted through the gray level co-occurrence matrix of the coal and the gangue as the texture characteristic parameters, and the gray level co-occurrence matrix P (i, jd, theta) refers to the probability that a point of a gray level i reaches a gray level j at a specific position d (delta x, delta y);
wherein the content of the first and second substances,(x, y) is the coordinate of the pixel point, and the range of the delta x and the delta y is determined by two parameters of the pixel distance d and the angle theta; i and j are respectively used for expressing the gray scale of the pixel; d refers to the direction and distance between two different pixels; theta refers to a corresponding direction of the gray level co-occurrence matrix generation. According to the characteristics of coal and gangue, the angle theta of the invention is selected from four angles of 0 degree, 45 degrees, 90 degrees and 135 degrees, the characteristic values of the four directions are averaged, and the distance d is 1.
Further, energy: a measure of the uniformity of the gray level distribution of the image, and the thickness of the texture, reflects the uniformity of the gray level distribution in the image area,
further, contrast: a measure of the local variation present in the image, reflecting the sharpness of the image and the depth of the grooves of the texture,
further, the correlation: a measure of how similar the grey levels of an image are in the row or column direction, reflects how similar the image is,
further, uniformity: a measure of smoothness of the image distribution reflects the sharpness and regularity of the texture in the image
Further, entropy: a measure of the randomness of the amount of information contained in an image reflects the complexity of the gray level distribution of the image, with larger entropy values, more complex images,
further, the invention establishes a training set and a test set of coal and gangue identification characteristics and a classification identification model thereof through characteristic extraction, and specifically adopts a classification model of a Support Vector Machine (SVM) classifier, wherein the SVM classifier is a generalized linear classifier which carries out binary classification on data according to a supervised learning mode, the essence of the SVM is a method for quantifying the difference of two types of data, and the decision hyperplane and kernel function calculation formulas of the SVM model adopted by the invention are respectively as follows:
the embodiments disclosed in the present invention are within the scope of the claims, and the specific embodiments are only for describing the specific embodiments of the present invention, and the scope of the present invention is not limited to the specific embodiments, and the specific embodiments should not be construed as limiting the scope of the claims.
In addition, the present invention does not disclose relevant components in the specification and the drawings, which do not hinder the understanding of the present invention by those skilled in the art, and does not disclose other conventional components of the present invention, which do not hinder the understanding of the present invention by those skilled in the art.
The product structure connection relation falling within the protection scope of the invention falls within the protection content of the invention; it is within the spirit of the present invention that conventional technical modifications to the structure of product parts, such as those made in the specific embodiments of the present invention, may be made without departing from the spirit of the present invention.
While certain exemplary embodiments of the invention have been described above by way of illustration only, it will be apparent to those skilled in the art that the described embodiments may be modified in various different ways without departing from the scope of the invention. Accordingly, the foregoing description is illustrative in nature and is not to be construed as limiting the scope of the invention as claimed.
Unless defined otherwise, all academic and scientific terms used herein have the same meaning as is understood by one of ordinary skill in the art to which this invention belongs.
In case of conflict, the present specification, including definitions, will control.
All percentages, parts, ratios, etc., are by weight unless otherwise indicated.
When a value or range of values, preferred range or list of lower preferable values and upper preferable values is given, it should be understood that it specifically discloses any range formed by any pair of values of any lower range limit or preferred value and any upper range limit or preferred value, regardless of whether ranges are separately disclosed. Where a range of numerical values is described herein, unless otherwise stated, the range is intended to include the endpoints of the range and all integers and fractions within the range.
When the term "about" or "approximately" is used to describe a numerical value or an end of a range, the disclosure should be interpreted to include the specific numerical value or end points referred to.
The use of "a" and "an" are merely for convenience and to provide a general context for the invention. Unless expressly stated otherwise, this description should be read to include one or at least one.
Claims (10)
1. The utility model provides a gangue discernment sorting unit based on terahertz formation of image which characterized in that: including the vibration screening device, vibration screening device one side is provided with the material conveyer with vibration screening device complex, material conveyer's upper and lower position is provided with the terahertz light source and the terahertz linear array camera of mutually supporting respectively, material conveyer's input is provided with the device of lining up with material conveyer matched with, material conveyer's output is provided with rather than complex sorting unit, terahertz linear array camera and sorting unit respectively with computer linear connection.
2. The coal gangue identification and sorting device based on terahertz imaging is characterized in that: the dust removing device is characterized in that a pulverized coal belt matched with the vibration screening device is installed below the vibration screening device, the material conveying device and the sorting device are all connected with a dust removing device, a protective device matched with the vibration screening device is arranged on one side of the dust removing device, and a gangue belt and a clean coal belt matched with the material conveying device are arranged on the lower portion of the dust removing device respectively.
3. A coal gangue identification and sorting method based on terahertz imaging is characterized by comprising the following steps: comprises the following steps:
s1: selecting a plurality of coal blocks and gangue in the crushed raw coal, obtaining 7 characteristic quantities of the coal and gangue to form a training set of the coal and gangue according to the characteristic extraction step, selecting an optimal characteristic combination according to the linear dispersion of each characteristic, and training a classification model according to the optimal characteristic combination;
s2: establishing a terahertz image database of coal and gangue, labeling the terahertz images of the coal and the gangue, selecting a Gaussian kernel function to train on a training set, and testing the trained classification model on a testing set for identifying actual coal gangue;
s3: the sorting device positions and identifies and sorts the coal and the gangue on the material conveying device through data information sent by the computer image processing module.
4. The coal gangue identification and sorting method based on terahertz imaging according to claim 3, characterized in that: the step S3, in which the sorting device positions the coal and the gangue by obtaining the image centroid positions of the coal and the gangue, wherein the centroid position coordinate extraction formula is as follows:
5. The coal gangue identification and sorting method based on terahertz imaging according to claim 3, characterized in that: in the step S3, the sorting device performs secondary sorting during the sorting process, and further identifies and sorts the coal from the gangue.
6. The coal gangue identification and sorting method based on terahertz imaging according to claim 3, characterized in that: the feature extraction step in the step S1 includes the following steps:
a1: acquiring terahertz images of original coal and gangue from a terahertz imaging system, and preprocessing the original terahertz images to obtain denoised terahertz images;
a2: carrying out edge detection on the denoised terahertz images of the coal and the gangue so as to facilitate feature extraction and classification identification;
a3: and extracting features according to the difference of the terahertz image gray level histograms of the denoised coal and gangue.
7. The coal gangue identification and sorting method based on terahertz imaging according to claim 6, characterized in that: in the step a1, the median filtering is adopted to perform denoising processing on the original terahertz image, so as to eliminate the influence of pulse noise on the image and improve the definition of the terahertz image, and the median filtering denoising processing formula is as follows:
f(x,y)=median[g(k,l)](k,l)∈S xy
wherein g (x, y) is the image after median filtering denoising, f (x, y) is the original image before denoising, and S XY Filtering the window for the template; in the step A2, the de-noised terahertz images of the coal and the gangue are subjected to edge detection by using a Sobel operator, the Sobel operator is a discrete differential operator mainly used for edge detection, and the images are subjected to weighted smoothing treatment according to gray level weighting differences of upper and lower adjacent points and left and right adjacent points of a pixel point and then subjected to differential operation.
8. The coal gangue identification and sorting method based on terahertz imaging according to claim 6, characterized in that: the step of extracting features through the difference of the gray level histograms in the step a3 includes: intercepting the terahertz image of each coal and gangue into an image with 256 x 256 pixels, processing and analyzing the image by utilizing Matlab software, and statistically analyzing the gray characteristic quantity of the coal and gangue 2 To represent the gray level characteristics of coal and gangue, the gray level characteristic extraction formula is as follows:
wherein g (x, y) is the gray scale of the terahertz image, and the length and the width of the image are both N.
9. The coal gangue identification and sorting method based on terahertz imaging according to claim 6, characterized in that: the method also comprises the step of extracting the texture characteristic parameters of the terahertz images of the denoised coal and gangue, the method adopts the theory based on the gray level co-occurrence matrix to research the texture characteristic extraction of the coal and gangue, the energy, the contrast, the correlation, the uniformity and the entropy can be extracted through the gray level co-occurrence matrix of the coal and the gangue to be used as the texture characteristic parameters, and the gray level co-occurrence matrix P (i, j | d, theta) refers to the probability that the point of a gray level i departs from a certain specific position d ═ delta x, delta y and reaches a gray level j;
wherein the content of the first and second substances,(x, y) is the coordinate of the pixel point, and the range of the delta x and the delta y is determined by two parameters of the pixel distance d and the angle theta; i and j are respectively used for expressing the gray level of the pixel; d refers to the direction and distance between two different pixels; theta refers to a corresponding direction of the gray level co-occurrence matrix generation.
10. The coal gangue identification and sorting method based on terahertz imaging according to claim 6, characterized in that: the invention also comprises a training set and a test set of coal and gangue identification characteristics and a classification identification model thereof which are established by characteristic extraction, in particular to a classification model adopting a Support Vector Machine (SVM) classifier, and the calculation formulas of a decision hyperplane and a kernel function of the SVM model adopted by the invention are respectively as follows:
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202210470859.3A CN114887924A (en) | 2022-04-28 | 2022-04-28 | Terahertz imaging-based coal gangue identification and sorting device and method |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202210470859.3A CN114887924A (en) | 2022-04-28 | 2022-04-28 | Terahertz imaging-based coal gangue identification and sorting device and method |
Publications (1)
Publication Number | Publication Date |
---|---|
CN114887924A true CN114887924A (en) | 2022-08-12 |
Family
ID=82719694
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202210470859.3A Pending CN114887924A (en) | 2022-04-28 | 2022-04-28 | Terahertz imaging-based coal gangue identification and sorting device and method |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN114887924A (en) |
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN116871177A (en) * | 2023-09-05 | 2023-10-13 | 国擎(山东)信息科技有限公司 | Method and system for separating kaolin crude ore based on multispectral technology |
Citations (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN102521572A (en) * | 2011-12-09 | 2012-06-27 | 中国矿业大学 | Image recognition method of coal and gangue |
CN106733684A (en) * | 2017-01-22 | 2017-05-31 | 杨宝祥 | Mineral dry separation system and method |
CN108311410A (en) * | 2018-01-27 | 2018-07-24 | 华北科技学院 | A kind of dry method lump material sorting unit based on terahertz imaging |
CN109834055A (en) * | 2019-04-09 | 2019-06-04 | 安徽理工大学 | A kind of bastard coal separation system and its method for separating based on dual-energy x-ray |
CN110245635A (en) * | 2019-06-20 | 2019-09-17 | 安徽理工大学 | A kind of Infrared image recognition of coal and spoil |
CN112222015A (en) * | 2020-10-23 | 2021-01-15 | 安徽理工大学 | Double-view-angle imaging raw coal gangue pre-discharging device and method |
-
2022
- 2022-04-28 CN CN202210470859.3A patent/CN114887924A/en active Pending
Patent Citations (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN102521572A (en) * | 2011-12-09 | 2012-06-27 | 中国矿业大学 | Image recognition method of coal and gangue |
CN106733684A (en) * | 2017-01-22 | 2017-05-31 | 杨宝祥 | Mineral dry separation system and method |
CN108311410A (en) * | 2018-01-27 | 2018-07-24 | 华北科技学院 | A kind of dry method lump material sorting unit based on terahertz imaging |
CN109834055A (en) * | 2019-04-09 | 2019-06-04 | 安徽理工大学 | A kind of bastard coal separation system and its method for separating based on dual-energy x-ray |
CN110245635A (en) * | 2019-06-20 | 2019-09-17 | 安徽理工大学 | A kind of Infrared image recognition of coal and spoil |
CN112222015A (en) * | 2020-10-23 | 2021-01-15 | 安徽理工大学 | Double-view-angle imaging raw coal gangue pre-discharging device and method |
Non-Patent Citations (2)
Title |
---|
牛清娜;: "基于边缘检测算法的煤矸石自动分选技术研究", 河北工程大学学报(自然科学版), no. 04 * |
米强: "基于机器视觉的煤与矸石自动识别研究", 《中国优秀硕士学位论文全文数据库 工程科技I辑》, no. 3, pages 24 - 26 * |
Cited By (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN116871177A (en) * | 2023-09-05 | 2023-10-13 | 国擎(山东)信息科技有限公司 | Method and system for separating kaolin crude ore based on multispectral technology |
CN116871177B (en) * | 2023-09-05 | 2023-11-17 | 国擎(山东)信息科技有限公司 | Method and system for separating kaolin crude ore based on multispectral technology |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN108971190B (en) | Machine vision-based household garbage sorting method | |
Bai et al. | Image segmentation method for coal particle size distribution analysis | |
CN101059425A (en) | Method and device for identifying different variety green tea based on multiple spectrum image texture analysis | |
CN105160654A (en) | Towel label defect detecting method based on feature point extraction | |
CN108765412A (en) | A kind of steel strip surface defect sorting technique | |
CN104850854A (en) | Talc ore product sorting processing method and talc ore product sorting system | |
CN112893159B (en) | Coal gangue sorting method based on image recognition | |
Ji et al. | Classification and identification of foreign fibers in cotton on the basis of a support vector machine | |
CN109087286A (en) | A kind of detection method and application based on Computer Image Processing and pattern-recognition | |
Rahamathunnisa et al. | Vegetable disease detection using k-means clustering and svm | |
CN107145884A (en) | Gangue near-infrared image identification technology | |
CN110335233A (en) | Express-way guard-rail plates defect detecting system and method based on image processing techniques | |
Rabidas et al. | A study of different texture features based on local operator for benign-malignant mass classification | |
CN104952754B (en) | Silicon chip method for separating after plated film based on machine vision | |
CN114887924A (en) | Terahertz imaging-based coal gangue identification and sorting device and method | |
Shen et al. | Development of a new machine vision algorithm to estimate potato's shape and size based on support vector machine | |
JP2004045356A (en) | Surface defect detection method | |
CN201041547Y (en) | Device for analyzing and recognizing different types of green teas based on multi-spectrum image texture | |
Wang et al. | Automatic separation system of coal gangue based on DSP and digital image processing | |
CN108596249B (en) | Image feature extraction and classification method and device | |
Yahya et al. | Image enhancement background for high damage Malay manuscripts using adaptive Threshold Binarization | |
Bai et al. | Recognition of bovine milk somatic cells based on multi-feature extraction and a GBDT-AdaBoost fusion model | |
Lu et al. | Crystal morphology monitoring based on in-situ image analysis of L-glutamic acid crystallization | |
Sethy et al. | Quantification of rice chalkiness using image processing | |
Murray et al. | Multiscale AM-FM analysis of pneumoconiosis x-ray images |
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 |