CN105944976A - Method and device for sorting massive gangue by using digital image processing technology - Google Patents

Abstract

A method of sorting large pieces of coal gangue using digital image processing technology. First, the original image of the raw coal to be tested on the conveyor belt is obtained through a CCD camera, and the image is preprocessed using digital image processing technology; secondly, the gray value of the image is used to fill The height value of the image is reconstructed three-dimensionally to calculate the pixel volume of raw coal, and the actual volume of raw coal is calculated according to the corresponding pixel equivalent, and a large piece of raw coal is obtained according to the volume threshold; then determined according to the gray histogram of the image coal gangue, and use industrial robots to sort the coal gangue to designated areas; the invention has the characteristics of improving the accuracy and speed of sorting large pieces of coal gangue.

Description

A method and device for sorting large pieces of coal gangue using digital image processing technology

technical field

The invention relates to the technical field of sorting large coal gangue, in particular to a method and device for sorting large coal gangue by using digital image processing technology.

Background technique

my country has a wide distribution of coal fields and a full coal-forming age. It is the country with the largest coal mining volume in the world. Improving the quality and output of coal is an important measure to improve the economic benefits of coal mines and enhance the international competitiveness of the coal mining industry.

Coal gangue is a rock mixed with organic and inorganic compounds co-deposited with coal during the coal-forming process. It is a low-quality fuel with low calorific value, low carbon content, high hardness, high mineral content, and low organic matter content. If the content of coal gangue in the coal mine is too much, it will affect the output and quality of coal. Generally, coal gangue is stacked as a kind of industrial waste in the open air, which not only wastes valuable mineral resources, but also occupies farmland. The gangue leaching water will pollute the surrounding soil and groundwater, and coal gangue contains certain combustibles. Spontaneous combustion occurs under suitable conditions, and harmful gases such as sulfur dioxide, nitrogen oxides, carbon oxides, and smoke are emitted to pollute the atmospheric environment, affect the health of residents in mining areas, and bring very serious social, environmental, and economic problems. However, if coal gangue is used reasonably, or useful minerals and chemical products are recovered from coal gangue, or used in the construction industry as filling materials, or made into fertilizers, it will have great commercial value.

For many years, large pieces of coal gangue have been pushed from the hand-selected belt conveyor to the gangue discharge chute by manpower with a shovel. This method is extremely labor-intensive, and the efficiency of gangue selection and discharge is low, especially in underground mines. During fault passing, the manual sorting and discharge of large gangue seriously restricts the production of the entire mine and the lifting speed of the mine. The use of image processing technology and industrial robots to sort coal gangue can reduce the complexity of traditional image processing algorithms, improve the accuracy and speed of coal gangue sorting, and ensure the quality and output of coal, which has great commercial value.

Contents of the invention

In order to overcome the shortcomings of the above-mentioned prior art, the object of the present invention is to provide a method and device for sorting large pieces of coal gangue using digital image processing technology, using the method of filling the image height to obtain the volume of raw coal and using industrial robots to realize automatic sorting , which has the characteristics of reducing the complexity of image processing and improving the accuracy and speed of sorting large pieces of coal gangue.

In order to achieve the above object, the technical scheme adopted in the present invention is as follows:

A method for sorting large pieces of coal gangue using digital image processing technology, characterized in that it comprises the following steps:

Step 1: Obtain the left view, right view and front view of the measured raw coal on the conveyor belt through the CCD camera, and transmit the images to the computer;

Step 2: Calculate the pixel equivalent ε between the image size and the actual size, and obtain the actual size represented by the unit pixel;

Step 3: filter the left view, right view, and main view, and perform grayscale conversion to convert the image into a grayscale image;

Step 4: Segment the raw coal image and the background to obtain an effective coal block area; calculate the average gray value G _mean of the coal block area, and if G _mean is greater than 120, it is coal gangue;

Step 5: Binarize the main view using the adaptive threshold method, and calculate the positions of the upper, lower, left, and right vertices;

Step 6: Calculate the horizontal distances from the upper vertex, the lower vertex to the left vertex as N _left1 and N _left2 , and take the larger value N _left = max(N _left1 , N _left2 ) as the maximum number of pixels on the left side;

Step 7: Draw the gray histogram of the left view, the height corresponding to the minimum gray value G _lmin is 0, the height corresponding to the maximum gray value G _lmax is N _left , and the pixel heights of other gray values are

${\mathrm{<span\; class="notranslate">\; h</span>}}_{\mathrm{<span\; class="notranslate">\; l</span>}\mathrm{<span\; class="notranslate">\; x</span>}}<span\; class="notranslate">\; =</span>\frac{{\mathrm{<span\; class="notranslate">\; N</span>}}_{\mathrm{<span\; class="notranslate">\; l</span>}\mathrm{<span\; class="notranslate">\; e</span>}\mathrm{<span\; class="notranslate">\; f</span>}\mathrm{<span\; class="notranslate">\; t</span>}}}{{\mathrm{<span\; class="notranslate">\; G</span>}}_{\mathrm{<span\; class="notranslate">\; l</span>}\mathrm{<span\; class="notranslate">\; max</span>}}<span\; class="notranslate">\; -</span>{\mathrm{<span\; class="notranslate">\; G</span>}}_{\mathrm{<span\; class="notranslate">\; l</span>}\mathrm{<span\; class="notranslate">\; min</span>}}}<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; G</span>}}_{\mathrm{<span\; class="notranslate">\; l</span>}\mathrm{<span\; class="notranslate">\; x</span>}}<span\; class="notranslate">\; -</span>{\mathrm{<span\; class="notranslate">\; G</span>}}_{\mathrm{<span\; class="notranslate">\; l</span>}\mathrm{<span\; class="notranslate">\; m</span>}\mathrm{<span\; class="notranslate">\; i</span>}\mathrm{<span\; class="notranslate">\; no</span>}}<span\; class="notranslate">\; )</span>$

The three-dimensional model is established through the height value of each point, and the computer obtains the actual voxel block number s ₁ contained in the three-dimensional model space of the left view through statistics;

Step 8: Calculate the horizontal distances from the upper vertex, the lower vertex to the right vertex as N _right1 and N _right2 , and take the larger value N _right = max(N _right1 , N _right2 ) as the maximum number of pixels on the right side;

Step 9: Draw the grayscale histogram of the _right view, the height value corresponding to the minimum grayscale value G _rmin is 0, the height value corresponding to the maximum grayscale value G rmax is _Nright , and the pixel heights of other grayscale values are

${\mathrm{<span\; class="notranslate">\; h</span>}}_{\mathrm{<span\; class="notranslate">\; r</span>}\mathrm{<span\; class="notranslate">\; x</span>}}<span\; class="notranslate">\; =</span>\frac{{\mathrm{<span\; class="notranslate">\; N</span>}}_{\mathrm{<span\; class="notranslate">\; r</span>}\mathrm{<span\; class="notranslate">\; i</span>}\mathrm{<span\; class="notranslate">\; g</span>}\mathrm{<span\; class="notranslate">\; h</span>}\mathrm{<span\; class="notranslate">\; t</span>}}}{{\mathrm{<span\; class="notranslate">\; G</span>}}_{\mathrm{<span\; class="notranslate">\; r</span>}\mathrm{<span\; class="notranslate">\; max</span>}}<span\; class="notranslate">\; -</span>{\mathrm{<span\; class="notranslate">\; G</span>}}_{\mathrm{<span\; class="notranslate">\; r</span>}\mathrm{<span\; class="notranslate">\; min</span>}}}<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; G</span>}}_{\mathrm{<span\; class="notranslate">\; r</span>}\mathrm{<span\; class="notranslate">\; x</span>}}<span\; class="notranslate">\; -</span>{\mathrm{<span\; class="notranslate">\; G</span>}}_{\mathrm{<span\; class="notranslate">\; r</span>}\mathrm{<span\; class="notranslate">\; m</span>}\mathrm{<span\; class="notranslate">\; i</span>}\mathrm{<span\; class="notranslate">\; no</span>}}<span\; class="notranslate">\; )</span>$

The three-dimensional model is established through the height value of each point, and the computer obtains the actual voxel block number s2 contained in the three _- dimensional model space of the right view through statistics;

Sum the number of actual voxel blocks contained in the left and right view three-dimensional model space, sum=s ₁ +s ₂ ;

Step ten: use the pixel equivalent ε to convert the pixel volume in the image coordinate system into the actual measurement volume, V=sum×ε ³ ;

Comparing the calculated volume V with the threshold V _max , those larger than V _max are large pieces of coal gangue;

Step 11: Obtain the coordinates of the gangue on the conveyor belt through target positioning, and transmit the coordinate information to the industrial robot through Ethernet; the industrial robot grabs the gangue to the designated area, and other raw coal is transported to the designated coal area by the conveyor belt.

Said V _max is 100 cm ³ .

A device for sorting large pieces of coal gangue using digital image processing technology, including a conveyor belt for transporting coal gangue and raw coal, an upper camera is installed above the conveyor belt, and a left camera and a right camera are installed at the left and right ends of the conveyor belt. The industrial robot on one side of the conveyor belt; the left camera and the right camera lens are placed facing each other, both fixed on the edge of the conveyor belt and on the same vertical plane as the upper camera, and the industrial robot is on the same side as the left camera and along the conveyor belt The direction is set in front of the left camera; the output ends of the left camera, the right camera and the upper camera are connected to the input end of the industrial robot.

Beneficial effects of the present invention:

The large coal gangue sorting industrial robot involved in the present invention can realize the automatic sorting process of the large coal gangue, and has the characteristics of automatic, non-destructive, high accuracy and fast speed. If the present invention is applied to the field of coal mining, large pieces of coal gangue mixed in coal mines can be better sorted out to ensure the quality and output of coal. At the same time, the coal gangue can also be sorted out for comprehensive utilization and reduce environmental pollution. Promote economic development and have great market potential.

Description of drawings

Fig. 1 is the flowchart of the calculation of large coal gangue in the present invention.

Fig. 2 is the flow chart of the sorting of bulk coal gangue in the present invention.

Fig. 3 is a structural schematic diagram of the device of the present invention.

detailed description

The implementation of the present invention will be described in detail below in conjunction with the drawings and examples.

As shown in Figure 1: a kind of method utilizing digital image processing technology to sort bulk coal gangue is characterized in that, comprises the following steps:

Step 1: Obtain the left view, right view and front view of the measured raw coal on the conveyor belt through the CCD camera, and transmit the images to the computer;

Step 2: Calculate the pixel equivalent ε between the image size and the actual size, and obtain the actual size represented by the unit pixel;

Step 3: filter the left view, right view, and main view, and perform grayscale conversion to convert the image into a grayscale image;

Step 4: Segment the raw coal image and the background to obtain an effective coal block area; calculate the average gray value G _mean of the coal block area, and if G _mean is greater than 120, it is coal gangue;

Step 5: Binarize the main view using the adaptive threshold method, and calculate the positions of the upper, lower, left, and right vertices;

Step 6: Calculate the horizontal distances from the upper vertex, the lower vertex to the left vertex as N _left1 and N _left2 , and take the larger value N _left = max(N _left1 , N _left2 ) as the maximum number of pixels on the left side;

Step 7: Draw the gray histogram of the left view, the height corresponding to the minimum gray value G _lmin is 0, the height corresponding to the maximum gray value G _lmax is N _left , and the pixel heights of other gray values are

${\mathrm{<span\; class="notranslate">\; h</span>}}_{\mathrm{<span\; class="notranslate">\; l</span>}\mathrm{<span\; class="notranslate">\; x</span>}}<span\; class="notranslate">\; =</span>\frac{{\mathrm{<span\; class="notranslate">\; N</span>}}_{\mathrm{<span\; class="notranslate">\; l</span>}\mathrm{<span\; class="notranslate">\; e</span>}\mathrm{<span\; class="notranslate">\; f</span>}\mathrm{<span\; class="notranslate">\; t</span>}}}{{\mathrm{<span\; class="notranslate">\; G</span>}}_{\mathrm{<span\; class="notranslate">\; l</span>}\mathrm{<span\; class="notranslate">\; max</span>}}<span\; class="notranslate">\; -</span>{\mathrm{<span\; class="notranslate">\; G</span>}}_{\mathrm{<span\; class="notranslate">\; l</span>}\mathrm{<span\; class="notranslate">\; min</span>}}}<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; G</span>}}_{\mathrm{<span\; class="notranslate">\; l</span>}\mathrm{<span\; class="notranslate">\; x</span>}}<span\; class="notranslate">\; -</span>{\mathrm{<span\; class="notranslate">\; G</span>}}_{\mathrm{<span\; class="notranslate">\; l</span>}\mathrm{<span\; class="notranslate">\; m</span>}\mathrm{<span\; class="notranslate">\; i</span>}\mathrm{<span\; class="notranslate">\; no</span>}}<span\; class="notranslate">\; )</span>$

The three-dimensional model is established through the height value of each point, and the computer obtains the actual voxel block number s ₁ contained in the three-dimensional model space of the left view through statistics;

Step 8: Calculate the horizontal distances from the upper vertex, the lower vertex to the right vertex as N _right1 and N _right2 , and take the larger value N _right = max(N _right1 , N _right2 ) as the maximum number of pixels on the right side;

Step 9: Draw the grayscale histogram of the _right view, the height value corresponding to the minimum grayscale value G _rmin is 0, the height value corresponding to the maximum grayscale value G rmax is _Nright , and the pixel heights of other grayscale values are

${\mathrm{<span\; class="notranslate">\; h</span>}}_{\mathrm{<span\; class="notranslate">\; r</span>}\mathrm{<span\; class="notranslate">\; x</span>}}<span\; class="notranslate">\; =</span>\frac{{\mathrm{<span\; class="notranslate">\; N</span>}}_{\mathrm{<span\; class="notranslate">\; r</span>}\mathrm{<span\; class="notranslate">\; i</span>}\mathrm{<span\; class="notranslate">\; g</span>}\mathrm{<span\; class="notranslate">\; h</span>}\mathrm{<span\; class="notranslate">\; t</span>}}}{{\mathrm{<span\; class="notranslate">\; G</span>}}_{\mathrm{<span\; class="notranslate">\; r</span>}\mathrm{<span\; class="notranslate">\; max</span>}}<span\; class="notranslate">\; -</span>{\mathrm{<span\; class="notranslate">\; G</span>}}_{\mathrm{<span\; class="notranslate">\; r</span>}\mathrm{<span\; class="notranslate">\; min</span>}}}<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; G</span>}}_{\mathrm{<span\; class="notranslate">\; r</span>}\mathrm{<span\; class="notranslate">\; x</span>}}<span\; class="notranslate">\; -</span>{\mathrm{<span\; class="notranslate">\; G</span>}}_{\mathrm{<span\; class="notranslate">\; r</span>}\mathrm{<span\; class="notranslate">\; m</span>}\mathrm{<span\; class="notranslate">\; i</span>}\mathrm{<span\; class="notranslate">\; no</span>}}<span\; class="notranslate">\; )</span>$

The three-dimensional model is established through the height value of each point, and the computer obtains the actual voxel block number s2 contained in the three _- dimensional model space of the right view through statistics;

Sum the number of actual voxel blocks contained in the left and right view three-dimensional model space, sum=s ₁ +s ₂ ;

Step ten: use the pixel equivalent ε to convert the pixel volume in the image coordinate system into the actual measurement volume, V=sum×ε ³ ;

Comparing the calculated volume V with the threshold V _max , those larger than V _max are large pieces of coal gangue;

Step 11: Obtain the coordinates of the gangue on the conveyor belt through target positioning, and transmit the coordinate information to the industrial robot through Ethernet; the industrial robot grabs the gangue to the designated area, and other raw coal is transported to the designated coal area by the conveyor belt.

Said V _max is 100 cm ³ .

As shown in Figure 3: a device that uses digital image processing technology to sort large pieces of coal gangue, including a conveyor belt for transporting coal gangue and raw coal, an upper camera is installed above the conveyor belt, and a left camera is installed at the left and right ends of the conveyor belt The right camera and the industrial robot fixed on one side of the conveyor belt; the left camera and the right camera lens are placed facing each other, both fixed on the edge of the conveyor belt and on the same vertical plane as the upper camera, and the industrial robot and the left camera are on the same vertical plane. side, and arranged in front of the left camera along the direction of the conveyor belt; the output ends of the left camera, right camera, and upper camera are connected to the input end of the industrial robot.

Working principle of the present invention:

Firstly, the original image of the measured raw coal on the conveyor belt is obtained through the CCD camera, and the image is preprocessed by digital image processing technology; secondly, the height value of the image is filled with the gray value of the image and three-dimensional reconstruction is performed to calculate the pixel volume of the raw coal. And calculate the actual volume of raw coal according to the corresponding pixel equivalent, and judge the large piece of raw coal according to the volume threshold; then determine the gangue according to the gray histogram of the image, and use the industrial robot to sort the gangue to the designated area.

Claims (3)

1. a method utilizing digital image processing technology to sort bulk coal gangue, is characterized in that, comprises the following steps: Step 1: Obtain the left view, right view and front view of the measured raw coal on the conveyor belt through the CCD camera, and transmit the images to the computer; Step 2: Calculate the pixel equivalent ε between the image size and the actual size, and obtain the actual size represented by the unit pixel; Step 3: filter the left view, right view, and main view, and perform grayscale conversion to convert the image into a grayscale image; Step 4: Segment the raw coal image and the background to obtain an effective coal block area; calculate the average gray value G _mean of the coal block area, and if G _mean is greater than 120, it is coal gangue; Step 5: Binarize the main view using the adaptive threshold method, and calculate the positions of the upper, lower, left, and right vertices; Step 6: Calculate the horizontal distances from the upper vertex, the lower vertex to the left vertex as N _left1 and N _left2 , and take the larger value N _left = max(N _left1 , N _left2 ) as the maximum number of pixels on the left side; Step 7: Draw the gray histogram of the left view, the height corresponding to the minimum gray value G _lmin is 0, the height corresponding to the maximum gray value G _lmax is N _left , and the pixel heights of other gray values are ${\mathrm{<span\; class="notranslate">\; h</span>}}_{\mathrm{<span\; class="notranslate">\; l</span>}\mathrm{<span\; class="notranslate">\; x</span>}}<span\; class="notranslate">\; =</span>\frac{{\mathrm{<span\; class="notranslate">\; N</span>}}_{\mathrm{<span\; class="notranslate">\; l</span>}\mathrm{<span\; class="notranslate">\; e</span>}\mathrm{<span\; class="notranslate">\; f</span>}\mathrm{<span\; class="notranslate">\; t</span>}}}{{\mathrm{<span\; class="notranslate">\; G</span>}}_{\mathrm{<span\; class="notranslate">\; l</span>}\mathrm{<span\; class="notranslate">\; max</span>}}<span\; class="notranslate">\; -</span>{\mathrm{<span\; class="notranslate">\; G</span>}}_{\mathrm{<span\; class="notranslate">\; l</span>}\mathrm{<span\; class="notranslate">\; min</span>}}}<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; G</span>}}_{\mathrm{<span\; class="notranslate">\; l</span>}\mathrm{<span\; class="notranslate">\; x</span>}}<span\; class="notranslate">\; -</span>{\mathrm{<span\; class="notranslate">\; G</span>}}_{\mathrm{<span\; class="notranslate">\; l</span>}\mathrm{<span\; class="notranslate">\; min</span>}}<span\; class="notranslate">\; )</span>$ The three-dimensional model is established through the height value of each point, and the computer obtains the actual voxel block number s ₁ contained in the three-dimensional model space of the left view through statistics; Step 8: Calculate the horizontal distances from the upper vertex, the lower vertex to the right vertex as N _right1 and N _right2 , and take the larger value N _right = max(N _right1 , N _right2 ) as the maximum number of pixels on the right side; Step 9: Draw the grayscale histogram of the _right view, the height value corresponding to the minimum grayscale value G _rmin is 0, the height value corresponding to the maximum grayscale value G rmax is _Nright , and the pixel heights of other grayscale values are ${\mathrm{<span\; class="notranslate">\; h</span>}}_{\mathrm{<span\; class="notranslate">\; r</span>}\mathrm{<span\; class="notranslate">\; x</span>}}<span\; class="notranslate">\; =</span>\frac{{\mathrm{<span\; class="notranslate">\; N</span>}}_{\mathrm{<span\; class="notranslate">\; r</span>}\mathrm{<span\; class="notranslate">\; i</span>}\mathrm{<span\; class="notranslate">\; g</span>}\mathrm{<span\; class="notranslate">\; h</span>}\mathrm{<span\; class="notranslate">\; t</span>}}}{{\mathrm{<span\; class="notranslate">\; G</span>}}_{\mathrm{<span\; class="notranslate">\; r</span>}\mathrm{<span\; class="notranslate">\; m</span>}\mathrm{<span\; class="notranslate">\; a</span>}\mathrm{<span\; class="notranslate">\; x</span>}}<span\; class="notranslate">\; -</span>{\mathrm{<span\; class="notranslate">\; G</span>}}_{\mathrm{<span\; class="notranslate">\; r</span>}\mathrm{<span\; class="notranslate">\; m</span>}\mathrm{<span\; class="notranslate">\; i</span>}\mathrm{<span\; class="notranslate">\; no</span>}}}<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; G</span>}}_{\mathrm{<span\; class="notranslate">\; r</span>}\mathrm{<span\; class="notranslate">\; x</span>}}<span\; class="notranslate">\; -</span>{\mathrm{<span\; class="notranslate">\; G</span>}}_{\mathrm{<span\; class="notranslate">\; r</span>}\mathrm{<span\; class="notranslate">\; m</span>}\mathrm{<span\; class="notranslate">\; i</span>}\mathrm{<span\; class="notranslate">\; no</span>}}<span\; class="notranslate">\; )</span>$ The three-dimensional model is established through the height value of each point, and the computer obtains the actual voxel block number s2 contained in the three _- dimensional model space of the right view through statistics; Sum the number of actual voxel blocks contained in the left and right view three-dimensional model space, sum=s ₁ +s ₂ ; Step ten: use the pixel equivalent ε to convert the pixel volume in the image coordinate system into the actual measurement volume, V=sum×ε ³ ; Comparing the calculated volume V with the threshold V _max , those larger than V _max are large pieces of coal gangue; Step 11: Obtain the coordinates of the gangue on the conveyor belt through target positioning, and transmit the coordinate information to the industrial robot through Ethernet; the industrial robot grabs the gangue to the designated area, and other raw coal is transported to the designated coal area by the conveyor belt. 2. A method for sorting large pieces of coal gangue using digital image processing technology according to claim 1, characterized in that said V _max is 100cm ³ . 3. a kind of device utilizing digital image processing technology to sort large pieces of coal gangue based on the method described in claim 1, comprising a conveyor belt for transporting coal gangue and raw coal, an upper camera is installed above the conveyor belt, and the left and right ends of the conveyor belt are installed An industrial robot with a left camera and a right camera fixed on one side of the conveyor belt; the left camera and the right camera lens are placed facing each other, both of which are fixed on the edge of the conveyor belt and on the same vertical plane as the upper camera. The side camera is on the same side, and is arranged in front of the left camera along the direction of the conveyor belt; the output ends of the left camera, the right camera, and the upper camera are connected to the input end of the industrial robot.