CN203335082U - Coal recognizing system based on image antisymmetric bi-orthogonal wavelet transform - Google Patents

Abstract

The utility model discloses a coal recognizing system based on image antisymmetric bi-orthogonal wavelet transform. The system comprises a central processing unit, a light source, a camera, an anti-explosion perspective window, a perspective window hairbrush, a hairbrush controller, a refrigerating module, an anti-explosion outer shell, a data initializing interface and a coal cutter controller. According to the system, the refrigerating module is additionally arranged in the anti-explosion outer shell and made of phase-changing constant temperature material, heat generated by heat dissipating devices can be absorbed by the refrigerating module, therefore, temperature in the anti-explosion outer shell can be kept constant, the recognizing system can work within a constant temperature range, the situation that components are damaged because the temperature in the anti-explosion outer shell is too high is avoided, the service life of the system is prolonged, and working reliability is improved.

Description

A kind of coal and rock identify system based on the conversion of image anti-symmetrical bi-orthogonal wavelet

Technical field

The utility model relates to a kind of coal and rock identify system based on the conversion of image anti-symmetrical bi-orthogonal wavelet, is a kind of device that is used for identifying coal-face coal seam and roch layer interface, belongs to image and processes and mode identification technology.

Background technology

The coal resources in China distributional region is vast, the condition of coal seam occurrence complexity.Due to the restriction of the geological conditionss such as Coal Seam Thickness Change, tomography, make the exploitation work plane in coal seam arrange difficulty relatively.For efficient, safety and the environmental protection exploitation that realizes coal resources, differentiate coal seam and play vital effect with the interface of rock stratum in progress of coal mining, and then also be conducive to adjust pit mining and dispose, formulation development, driving and back production plan.

In progress of coal mining, the recognition result of coal-rock interface can be used as the exploitation position that important evidence is regulated coal-winning machine, development machine etc., for example regulates the height of coal mining machine roller rocking arm or the cut scope of control cutting head of roadheader.The coal production coal and rock identify is mainly completed by the sense of hearing and vision by the staff at present.Due to the impact of dust in working environment and noise, make mining environment become complicated, be unfavorable for staff's subjective judgement, improved False Rate.Therefore coal-winning machine is inevitably understood cut to rock stratum or is leaked mining coal seam.Following problems mainly can be caused in the cut rock stratum: the rock that mistake is adopted is sneaked in raw coal and is caused the ature of coal amount to descend; Shorten gear life and the bearing life of coal-winning machine; Increase staff's the amount of labour; Increase the incidence of mine accident.And leak the coefficient of mining that mining coal seam has mainly reduced coal, the non-renewable resources that reduced productivity effect serious waste.

At present, be applied to the coal and rock identify system in actual production and install of a great variety, but recognition effect and the device compliance all undesirable.For example, scheme one, take sensor as basic coal and rock identify system, mainly utilizes a plurality of sensors to monitor every coefficient of coal-winning machine coal cutting simultaneously, and data are carried out to real-time analysis.This system need to be improved coal-winning machine equipment, and construction cost is high, and sensor line is fragile, poor reliability; Scheme two, high-pressure water shot coal and rock identify system, the different active forces that utilize contrajet to form nozzle are identified coal-rock interface.This system can cause on-the-spot draining difficulty, increases staff's the amount of labour, and device is complicated; Scheme three, the coal and rock identify system of the stressed effect of judgement cutting machine, according to the force-bearing situation judgement coal lithotypes of cut machine cut head.Due to the diversity of coal petrography kind, this system identification weak effect, and the device rapid wear changes.

In one piece of patent documentation that application number is 201220568471.9, a kind of coal and rock identify system based on the discrete many wavelet transformations of image is disclosed, this system architecture is simple, reliability is high, discrimination is high, but the existence due to the internal system radiating element, make the temperature of internal system constantly raise, can cause the damage of internal component.

The utility model content

The purpose of this utility model is to provide a kind of coal and rock identify system based on image anti-symmetrical bi-orthogonal wavelet conversion, in order to solve existing coal and rock identify system because the internal system excess Temperature damages the problem of components and parts.

For achieving the above object, scheme of the present utility model is: a kind of coal and rock identify system based on the conversion of image anti-symmetrical bi-orthogonal wavelet, comprise CPU, light source, camera, explosion-proof window, the window hairbrush, the hairbrush controller, refrigeration module, explosion-resistant enclosure, data initialization interface and coal-winning machine controller, the respective input mouth of described CPU respectively with described camera, the corresponding output port of coal-winning machine controller is connected with the data initialization interface, the corresponding output port of described CPU respectively with described camera, the respective input mouth of coal-winning machine controller, light source is connected with the hairbrush controller, described hairbrush controller control connection window hairbrush, described coal-winning machine controller is for the control connection coal-winning machine, described camera is a charge coupled device camera, described CPU, light source, camera, explosion-proof window, the hairbrush controller, refrigeration module is arranged in explosion-resistant enclosure.

Described CPU is comprised of with recognition unit and communication unit control module, IMAQ and memory cell, image processing, described image is processed with the recognition unit input and is connected described IMAQ and memory cell and control module, output connection control unit, communication unit and IMAQ and memory cell.

Described light source installs the circular white-light illuminating lamp of diffuse reflector high brightness symmetrically additional by many groups LED array and forms.

Described refrigeration module is made by phase-change constant-temperature material, for controlling the temperature of coal and rock identify internal system.

The beneficial effect that the utility model reaches: the utility model has been set up a module for refrigeration in explosion-resistant enclosure, this refrigeration module is comprised of phase-change constant-temperature material, can absorb the heat that radiating element produces, the temperature constant that keeps explosion-resistant enclosure inside, make recognition system be operated in a stationary temperature scope, avoid damaging because internal temperature is too high components and parts, extended the application life of system, improved the reliability of work.

The accompanying drawing explanation

Fig. 1 is the coal and rock identify system construction drawing based on the conversion of image anti-symmetrical bi-orthogonal wavelet of the present utility model;

Fig. 2 is CPU each several part workflow diagram of the present utility model;

Fig. 3 is the coal and rock identify flow chart based on the conversion of image anti-symmetrical bi-orthogonal wavelet of the present utility model;

Fig. 4 is the flow chart of the mine image enchancing method based on the anti-symmetrical bi-orthogonal wavelet conversion of the present utility model;

Fig. 5 is the schematic diagram based on anti-symmetrical bi-orthogonal wavelet conversion decomposition part of the present utility model;

Fig. 6 is the schematic diagram based on anti-symmetrical bi-orthogonal wavelet conversion reconstruct part of the present utility model.

The specific embodiment

Below in conjunction with accompanying drawing, the utility model is described in further detail.Coal and rock identify system of the present utility model comprises CPU, light source, camera, explosion-proof window, the window hairbrush, the hairbrush controller, refrigeration module, explosion-resistant enclosure, data initialization interface and coal-winning machine controller, the respective input mouth of described CPU respectively with described camera, the corresponding output port of coal-winning machine controller is connected with the data initialization interface, the corresponding output port of described CPU respectively with described camera, the respective input mouth of coal-winning machine controller, light source is connected with the hairbrush controller, described CPU, light source, camera, explosion-proof window, the hairbrush controller, refrigeration module is arranged in explosion-resistant enclosure.

Fig. 1 is the coal and rock identify system construction drawing based on the conversion of image anti-symmetrical bi-orthogonal wavelet of the present utility model.

As shown in the figure, 1 is CPU, and 2 is light source, and 3 is camera, 4 is explosion-proof window, and 5 is the hairbrush controller, and 6 is the window hairbrush, and 7 is refrigeration module, 8 is explosion-resistant enclosure, and 9 is the data initialization interface, and 10 is the coal-winning machine controller, and 11 is coal mining machine roller.

The light source of the present embodiment installs the circular white-light illuminating lamp of high brightness symmetrically of diffuse reflector additional by many groups LED array and forms, according to the demand of environment and camera daylighting, angle, the quantity of being lighted by central processing unit controls LED; Camera is a charge coupled device camera, the function of have automatic focusing, automatically regulating exposure, rotating up and down; Explosion-proof window gathers coal petrography image and down-hole electrical explosion proof for camera by window; The window hairbrush is for the coal dust on clean explosion-proof window, and the image that makes camera collect is not disturbed by the coal dust on explosion-proof window, to obtain coal petrography image clearly; The hairbrush controller is for receiving the electric impulse signal of CPU, drives the window hairbrush and swings on window, cleans the coal dust on it; Refrigeration module is a kind of device be comprised of phase-change constant-temperature material, for controlling the temperature of coal and rock identify device inside, prevents that LED illuminating lamp, camera and other heat radiation elements from producing amount of heat and making internal temperature too high and damage components and parts; Explosion-resistant enclosure is for the down-hole electrical explosion proof; The data initialization interface, for system initialization, is about to the local coal of oneself acquisition and the average multi-dimension texture Energy distribution of rock and writes CPU; The coal-winning machine controller is used for receiving the instruction of CPU, to control the coal mining track of coal-winning machine, avoids the coal mining machine roller cutting rock or leaks mining coal seam.

Fig. 2 is CPU each several part workflow diagram of the present utility model, and as shown in the figure, CPU comprises control module, IMAQ and memory cell, image processing and recognition unit and communication unit.The control module Main Function is to control camera, light source and hairbrush controller, and it is worked on request; IMAQ and memory cell be mainly used in contrasting collected by camera to the coal petrography image screened and stored; Image is processed with recognition unit and is mainly used in utilizing anti-symmetrical bi-orthogonal wavelet conversion and threshold function table processing strengthen image texture information and identify coal lithotypes to the object images of storage; Communication unit transfers to for image is processed to the result obtained with recognition unit processes the track that the coal-winning machine controller is mined to adjust coal-winning machine.

The specific works flow process of the described coal and rock identify system based on the conversion of image anti-symmetrical bi-orthogonal wavelet is:

(1) need to gather coal, this image of rock sample of local coal-face before the system routine work and utilize native system to test its sample image texture information after anti-symmetrical bi-orthogonal wavelet conversion and threshold function table processing, calculate the average texture Energy distribution under its different scale, and write image processing and recognition unit by the data initialization interface;

(2) initialize running environment by CPU: 1. put bright light source; 2. the hairbrush controller is controlled window hairbrush cleaning window surface once; 3. initialize camera apparatus;

(3) control module is adjusted brightness and camera sampling time interval and the angle of illuminating lamp according to environmental demand, takes coal petrography object picture, and deposits the coal petrography picture of shooting in IMAQ and memory cell;

(4) image is processed with recognition unit from IMAQ and memory cell reads image data, and adopt anti-symmetrical bi-orthogonal wavelet conversion and threshold function table to process to image, then and be stored in image and process with the coal of recognition unit, the average multi-dimension texture Energy distribution value of this image of rock sample and compare, identifying iconic element is coal or rock;

(5) communication unit is processed image with the result of recognition unit and is transferred to the coal-winning machine controller, to control the coal mining track of coal-winning machine, avoids the coal mining machine roller cutting rock or leaks mining coal seam.

In the system running, refrigeration module successively absorbs heat, with the holding device internal temperature in steady temperature; According to camera daylighting and angle demand, the quantity of central processing unit controls LED illumination lamp ignition; The hairbrush controller is subject to the control of CPU, and interval, control the window hairbrush and clear up explosion-proof window surface once at regular intervals.

Fig. 3 is the coal and rock identify flow chart based on the conversion of image anti-symmetrical bi-orthogonal wavelet of the present utility model.

Coal and rock identify process based on the anti-symmetrical bi-orthogonal wavelet conversion of the present utility model is divided into two stages: the first stage is the sample extraction stage, and second stage is the coal and rock identify stage.The sample extraction stage carries out before routine work, under identical image-forming condition, gathers the coal sample image set { f of one group of local coal-face ₁, f ₂f _mand this image set of rock sample { g ₁, g ₂g _m; and obtain the sample image texture information through anti-symmetrical bi-orthogonal wavelet conversion and threshold function table processing; calculate the high frequency subgraph reconstruct average texture Energy distribution under its different scale, and it is write to image processing and recognition unit, be convenient to the coal and rock identify of second stage.For sake of convenience, we narrated in conjunction with two stages simultaneously, and concrete steps are as follows:

A. under identical image-forming condition, gather one group of coal sample image set { f ₁, f ₂f _mand one group of this image set of rock sample { g ₁, g ₂g _m; Gather sample image f to be identified _x;

B. coloured image is converted into to gray level image;

C. coal, this image set of rock sample and sample image to be identified are done to the anti-symmetrical bi-orthogonal wavelet conversion, obtain each image wavelet territory low frequency component and high fdrequency component, and process through threshold function table the texture information that strengthens image;

D. calculate based on above-mentioned image set { f ₁, f ₂f _mand { g ₁, g ₂g _mthe average multi-dimension texture Energy distribution of wavelet field reconstructed high frequency component, be designated as respectively with

; Computed image f _xthe multi-dimension texture Energy distribution of wavelet field reconstructed high frequency component, be designated as

;

E. according to sample image multi-dimension texture Energy distribution to be identified

with coal, the average multi-dimension texture Energy distribution of rock

with

between the type of relation judgement coal petrography object.

In steps A, for the ease of processing, the coal of collection, this image set of rock sample and sample image to be identified should carry out under identical image-forming condition, for example illumination, camera parameter and lighting angle etc.The image size should equate without background, for the value that improves discrimination m can not be too little.

In step B, convert coloured image to gray level image.If the color digital image gathered is f, its R, G, the B component is respectively f _r, f _g, f _b.The data mode of f is three-dimensional array, f _r, f _g, f _bdata mode be two-dimensional array, the coordinate of pixel in position (x, the y) correspondence image of array element, the value of array element is the gray value that image (x, y) is located pixel, and gray value is got interval [0,255] integer in, 0 corresponding black wherein, 255 corresponding whites.

The formula that coloured image is converted to gray level image is:

F _gray(x，y)＝0.299f _R(x，y)+0.587f _G(x，y)+0.114f _B(x，y)

In step C, coal, this image set of rock sample and sample image to be identified are done to anti-symmetrical bi-orthogonal wavelet conversion and threshold process, as shown in Figure 4, concrete steps are as follows:

C1 decomposes the multiple dimensioned anti-symmetrical bi-orthogonal wavelet of mine original images by using A ' trous algorithm, obtains a low frequency component and three high fdrequency components under each yardstick;

C2 adopts the single threshold function to carry out the figure image intensifying to the low frequency component decomposed, and high fdrequency component adopts the dual threshold function to carry out the figure image intensifying;

C3 carries out anti-symmetrical bi-orthogonal wavelet reconstruct to low frequency, high fdrequency component after strengthening, and adopts the soft-threshold function to carry out the figure image intensifying;

The method that adopts A ' trous algorithm to carry out the anti-symmetrical bi-orthogonal wavelet decomposition to the mine image in described C1 step is: at first the row of signal is got to definite value, each row is regarded one-dimensional signal as and is decomposed, then the row of the signal after decomposing are got to definite value, every a line is regarded one-dimensional signal as and is remake once decomposition, the low frequency component obtained is the approximation coefficient of anti-symmetrical bi-orthogonal wavelet, and high fdrequency component is respectively level detail component, vertical detail component and the diagonal details component of anti-symmetrical bi-orthogonal wavelet.

The single threshold function of described C2 is: $w_o=\left\{\begin{array}{c}w+T_d(G-1),w>T_d\\ w-T_d(G-1),w<-T_d\\ \mathrm{Gw},\left|w\right|\&le;T_d\end{array}\right.$ , the dual threshold function is: $w_o=\left\{\begin{array}{c}w,w>T_{s2}\\ \frac{T_{s2}-G{T}_{s1}}{T_{s2}-T_{s1}}(w-T_{s2})+T_{s2},{\mathrm{<figure-callout\; id="1"\; label="T\; s"\; filenames="HSA0000092214540000031.png,HSA0000092214540000032.png"\; state="\{\{state\}\}">T</figure-callout>}}_s<w\&le;T_{s2}\\ \mathrm{Gw},-{\mathrm{<figure-callout\; id="1"\; label="T\; s"\; filenames="HSA0000092214540000031.png,HSA0000092214540000032.png"\; state="\{\{state\}\}">T</figure-callout>}}_s<w\&le;T_{s1}\\ \frac{T_{s2}-G{T}_{s1}}{T_{s2}-T_{s1}}(w+T_{s2})-T_{s2},-T_{s2}<w\&le;-T_{s1}\\ w,w\&le;-T_{s2}\end{array}\right.$ , wherein, G is for strengthening coefficient, and w is the wavelet coefficient before strengthening, w _ofor the wavelet coefficient after strengthening, T _dfor single threshold, T _s1, T _s2for dual threshold; T _d, T _s1, T _s2, the value of G adopts man-machine interaction to be selected.

The soft-threshold function of described C3 is: $w_o=\left\{\begin{array}{c}\left[\mathrm{sgn}\left(w\right)\right]\left(\right|w|-m{T}_r),\left|w\right|\&GreaterEqual;T_r\\ \mathrm{nw},\left|w\right|<T_r\end{array}\right.$ , wherein w is the wavelet coefficient before strengthening, w _othe wavelet coefficient after strengthening, m, n is regulatory factor, and m, n ∈ (0,1), T _rbe soft-threshold, soft-threshold adopts the method for estimating to obtain.

Described soft-threshold adopts Donoho and Johnstone to unify threshold value , wherein, the size that N is signal or length, σ is that noise criteria is poor, adopts at image and carries out the high frequency diagonal line details component in the first order wavelet coefficient of wavelet decomposition, with its estimation values sigma substitution of standard variance, unifies the threshold value formula

calculate soft-threshold.

Strengthen the texture information of mine image while in described step C3, carrying out anti-symmetrical bi-orthogonal wavelet reconstruct: first pairing approximation coefficient, level detail component, vertical detail component and diagonal detail coefficients column weight structure, then to its line reconstruction; Low frequency high fdrequency component during to small echo reconstruct at different levels adopts soft-threshold to process.

In step D, calculating chart image set { f ₁, f ₂f _m, { g ₁, g ₂g _mthe average multi-dimension texture Energy distribution of wavelet field reconstructed high frequency component

,

with image f _xthe multi-dimension texture Energy distribution of wavelet field reconstructed high frequency component .Piece image obtains equal-sized 4 width subgraphs through 1 wavelet decomposition, and a low frequency component subgraph and three high fdrequency component subgraphs are respectively approximate details LL, level detail LH, vertical detail HL and diagonal details HH image.While again carrying out wavelet decomposition, only the low-frequency approximation details LL of front one deck carried out to decomposition transform, generate 4 width subgraph (LL ₁, LH ₁, HL ₁and HH ₁), simultaneously, the 3 panel heights frequency detail pictures (LH, HL and HH) of one deck before retaining.By that analogy, each decomposition is all to front first approximation details LL _j+1carry out wavelet transformation, form 4 new width subgraph (LL _j, LH _j, HL _jand HH _j), after each the decomposition, the high frequency detail pictures increases by 3 width (LH _j, HL _jand HH _jeach increases by 1 width), and low-frequency approximation detail pictures number is constant., when after the j layer decomposes, approximate detail pictures becomes 1 width, and level, vertical and diagonal detail pictures is respectively the j width.Restructuring procedure is just contrary with decomposable process, and each reconstructed high frequency detail pictures reduces by 3 width, and low frequency detail pictures quantity is constant.

Fig. 5 is the schematic diagram based on anti-symmetrical bi-orthogonal wavelet conversion decomposition part of the present utility model.Approximation coefficient to original image

decompose through row, then decompose through space, obtain a low frequency component of wavelet decomposition

with three high fdrequency component W _ψ(j, m, n) _ LH, W _ψ(j, m, n) _ HL, W _ψ(j, m, n) _ HH.Wherein, h ' decomposes low pass filter, and g ' decomposes high-pass filter.

Fig. 6 is the schematic diagram based on anti-symmetrical bi-orthogonal wavelet conversion reconstruct part of the present utility model.The low frequency component by wavelet coefficient

with three high fdrequency component W _ψ(j, m, n) _ LH, W _ψ(j, m, n) _ HL, W _ψ(j, m, n) _ HH process column weight structure, then pass through line reconstruction, obtain the image approximate coefficient .Wherein, h is the reconstruct low pass filter, and g is the reconstruct high-pass filter.

High frequency detail pictures information while extracting reconstruct after soft-threshold is processed, by formula

$E_2^j=\underset{x}{\mathrm{\&Sigma;}}\underset{y}{\mathrm{\&Sigma;}}{\left[f_{\mathrm{HL}}^j(x,y)\right]}^2,$ $E_3^j=\underset{x}{\mathrm{\&Sigma;}}\underset{y}{\mathrm{\&Sigma;}}{\left[f_{\mathrm{HH}}^j(x,y)\right]}^2,$ Calculate the level of j yardstick, the energy of vertical and diagonal detail pictures, j=1 wherein, 2,3 ... J, f (x, y) is the gray value that (x, y) locates pixel,

energy value for the level of calculating the j yardstick, vertical and diagonal detail pictures;

By formula

, calculate total high frequency details energy of piece image j yardstick.

Further, in order to improve identification accuracy, we choose m known coal, rock image, utilize formula

calculate the average of the reconstructed high frequency multi-dimension texture Energy distribution under its same scale j, obtain the average multi-dimension texture Energy distribution of coal, rock image with

In step e, because the difference existed between the image texture that is all coal or rock is little, otherwise, image texture difference between coal image and rock image is large, the distance identification coal petrography object type in the time of therefore can converting reconstruct based on anti-symmetrical bi-orthogonal wavelet according to image between the multi-dimension texture Energy distribution of high fdrequency component.Concrete steps are as follows:

E1 gets the multi-dimension texture Energy distribution of coal petrography sample image to be identified

form a j=1,2,3 ... the J dimensional feature vector $E_x^{}$ ${}_1^{},E_x^2\&CenterDot;\&CenterDot;\&CenterDot;\&CenterDot;\&CenterDot;\&CenterDot;E_x^J;$

E2 reads coal and rock and sample Energy distribution to be identified from image processing and recognition unit with the j=1 of sub-band, 2,3 ... average texture Energy distribution vector under the J yardstick

with

;

E3 is compute vector respectively

with

between distance, be designated as d _coal,, d _rock, range formula is selected Euclidean distance:

$d_{\mathrm{coal}}=|E_x-{\stackrel{\&OverBar;}{E}}_{\mathrm{coal}}|=\sqrt{{|E_x^1-{\stackrel{\&OverBar;}{E}}_{\mathrm{<figure-callout\; id="1"\; label="coal"\; filenames="HSA0000092214540000031.png,HSA0000092214540000032.png"\; state="\{\{state\}\}">coal</figure-callout>}}^1|}^2+{|E_x^2-{\stackrel{\&OverBar;}{E}}_{\mathrm{coal}}^2|}^2+\&CenterDot;\&CenterDot;\&CenterDot;{|E_x^J-{\stackrel{\&OverBar;}{E}}_{\mathrm{coal}}^J|}^2}$

$d_{\mathrm{rock}}=|E_x-{\stackrel{\&OverBar;}{E}}_{\mathrm{rock}}|=\sqrt{{|E_x^1-{\stackrel{\&OverBar;}{E}}_{\mathrm{<figure-callout\; id="1"\; label="rock"\; filenames="HSA0000092214540000031.png,HSA0000092214540000032.png"\; state="\{\{state\}\}">rock</figure-callout>}}^1|}^2+{|E_x^2-{\stackrel{\&OverBar;}{E}}_{\mathrm{rock}}^2|}^2+\&CenterDot;\&CenterDot;\&CenterDot;{|E_x^J-{\stackrel{\&OverBar;}{E}}_{\mathrm{rock}}^J|}^2}$

Wherein, E ^jfor the energy value under the j yardstick.

E4 is d relatively _coal,, d _rocksize, sample image to be identified is included into to the less class of distance.

Claims (4)

1. the coal and rock identify system based on image anti-symmetrical bi-orthogonal wavelet conversion, comprise CPU, light source, camera, explosion-proof window, the window hairbrush, the hairbrush controller, refrigeration module, explosion-resistant enclosure, data initialization interface and coal-winning machine controller, the respective input mouth of described CPU respectively with described camera, the corresponding output port of coal-winning machine controller is connected with the data initialization interface, the corresponding output port of described CPU respectively with described camera, the respective input mouth of coal-winning machine controller, light source is connected with the hairbrush controller, described hairbrush controller control connection window hairbrush, described coal-winning machine controller is for the control connection coal-winning machine, described camera is a charge coupled device camera, described central processing unit, light source, camera, explosion-proof window, the hairbrush controller, refrigeration module is arranged in explosion-resistant enclosure.

2. a kind of coal and rock identify system based on image anti-symmetrical bi-orthogonal wavelet conversion according to claim 1, it is characterized in that, described CPU is comprised of with recognition unit and communication unit control module, IMAQ and memory cell, image processing, described image is processed with the recognition unit input and is connected described IMAQ and memory cell and control module, output connection control unit, communication unit and IMAQ and memory cell.

3. a kind of coal and rock identify system based on the conversion of image anti-symmetrical bi-orthogonal wavelet according to claim 1, is characterized in that, described light source installs the circular white-light illuminating lamp of diffuse reflector high brightness symmetrically additional by many groups LED array and forms.

4. a kind of coal and rock identify system based on the conversion of image anti-symmetrical bi-orthogonal wavelet according to claim 1, is characterized in that, described refrigeration module is made by phase-change constant-temperature material, for controlling the temperature of coal and rock identify internal system.

Images