CN103578111B - Rotary kiln based on flame image structural similarity burns till state identification method - Google Patents

Abstract

The present invention relates to a kind of rotary kiln based on flame image structural similarity and burn till state identification method. Its technical scheme is, established standards flame gray level image y, normal flame gray scale picture library FN and abnormal flame gray scale picture library FA. Obtain flame image P to be measured, treat survey flame image P and carry out filtering and greyscale transformation, obtain flame gray level image x to be measured; Flame gray level image x to be measured and all standard flame source gray level image y are averaged to structural similarity coefficient calculations, obtain a+b average structure likeness coefficient MSSIM (x, y); Be averaged structural similarity property coefficient maximum MAXmssim, if the corresponding standard flame source gray level image of average structure likeness coefficient maximum MAXmssim y belongs to normal flame gray scale picture library FN, flame image P to be measured belongs to normal condition; Otherwise belong to abnormality. The present invention has the advantages that precision is high, computation complexity is low, processing procedure is short and can realize the variation of on-line real time monitoring flame status.

Description

Rotary kiln based on flame image structural similarity burns till state identification method

Technical field

The invention belongs to the technical field that rotary kiln burns till state recognition. Be particularly related to a kind of based on flame image structural similarityRotary kiln burns till state identification method.

Background technology

Rotary kiln is that calcining or roasting and other modes are processed Thermal Equipment used in various raw material of industry technique, for to inputMaterial carry out machinery, physical or chemical treatment, in the industries such as building materials, chemical industry, metallurgy, have a wide range of applications. Turned roundThe particularity of kiln structure and the impact of process complexity, gained clinker quality index is difficult to on-line measurement, and the state that burns till of grog is alsoBe difficult to accurate identification, add the factor such as multivariable close coupling characteristic and uncertain noises of rotary kiln process, make rotary kiln fortuneTurn over the still open loop operational phase in " manually seeing fire " of journey, be difficult to realize the automatic control of rotary kiln, long-termOperation easily causes that clinker quality index is unstable, production capacity is low, energy consumption is high and the problem such as hand labor intensity is large.

Flame image is the effecting reaction of combustion process, and monochrome information has reflected radiation temperature and the combustion efficiency in combustion process,Flame profile has reflected the shape of combustion reaction generation area. Therefore, utilize rational technology to judge the shape that burns till of flame imageState, thereby running status and the combustion stability of definite rotary kiln, tool is of great significance.

In recent years, to burn till one of study hotspot of status recognition technique be digital image processing techniques to rotary kiln. Locate in advance through imageReason, image are cut apart and rear image is carried out feature extraction and chosen, and the characteristics of image that extracts and choose is sent in pattern classifierCarry out the pattern-recognition of target, thereby judge that flame burns till state. Up to the present, each to Digital Image Processing of numerous scholarsIndividual module has all been done a large amount of and deep research, uses digital image processing techniques to judge that flame image burns till the recognition methods of stateThough have the advantage that accuracy of identification is high, algorithm is various and ripe, these class methods inevitably exist computation complexity high, locateThe defect that the reason time is long, cannot on-line real time monitoring, does not fundamentally change the pattern of " manually seeing fire ", and affects backThe safety and reliability of rotary kiln control system, causes that clinker quality index is unstable, production capacity is low and high in cost of production series of problems.

Summary of the invention

The present invention is intended to overcome prior art defect, object be to provide that a kind of precision is high, computation complexity is low, processing procedure is short andThe rotary kiln based on flame image structural similarity that can realize on-line real time monitoring burns till state identification method.

For completing above-mentioned task, the concrete steps of the technical solution used in the present invention are:

The first step, standard flame source image Q are demarcated by kiln operation expert; Standard flame source image Q composition standard flame source picture library L,Standard flame source picture library L is divided into normal flame picture library LN and abnormal flame picture library LA, and normal flame picture library LN is by the normal shape of a widthThe standard flame source image Q composition of state, abnormal flame picture library LA is made up of the standard flame source image Q of b width abnormality; To markAccurate flame image Q carries out filtering processing and greyscale transformation, obtains standard flame source gray level image y; All standard flame source gray level imagesY composition standard flame source gray scale picture library F, standard flame source gray scale picture library F is divided into normal flame gray scale picture library FN and abnormal flame gray scalePicture library FA.

Second step, from the flame video of rotary kiln clinkering zone gathering, obtain a flame image P to be measured; By flame figure to be measuredCarry out filtering processing and greyscale transformation as P, obtain flame gray level image x to be measured.

The computational methods of the 3rd step, employing average structure likeness coefficient, calculate a flame gray level image x to be measured and each widthAverage structure likeness coefficient MSSIM (x, y) between standard flame source gray level image y, obtains a+b average structure similitude systemNumber MSSIM (x, y).

The 4th step, a+b the average structure likeness coefficient MSSIM (x, y) that the 3rd step is obtained compare, and choose average knotStructure likeness coefficient maximum MAXmssim; If the corresponding standard of average structure likeness coefficient maximum MAXmssimFlame gray level image y belongs to normal flame gray scale picture library FN, and the state that burns till of judging flame image P to be measured is normal condition;If the corresponding standard flame source gray level image of average structure likeness coefficient maximum MAXmssim y belongs to abnormal flame gray scalePicture library FA, the state that burns till of judging flame image P to be measured is abnormality.

If it is normal condition that the 5th step is judged the state that burns till of flame image P to be measured, directly carries out next width flame image and burn tillThe identification of state;

If judge, the state that burns till of flame image P to be measured is abnormality, and system alarm, improves rotary kiln flame and burn till after state,Carry out again the identification that next width flame image burns till state.

The 6th step, repetition the second～five step, until finish.

Described a is 1～1000 natural number; Described b is 1～1000 natural number.

The computational methods of described average structure likeness coefficient are: by flame gray level image x to be measured and standard flame source gray level imageY respectively with the window of formed objects by pixel move to the lower right corner from the upper left corner, obtain M flame gray level image piece x to be measured_jWith M standard flame source gray level image piece y_j, calculate each flame gray level image piece x to be measured_jWith corresponding standard flame source gray-scale mapPicture piece y_jBetween structural similarity coefficient S SIM (x_j,y_j), obtain M structural similarity coefficient S SIM (x_j,y_j), then rightM structural similarity coefficient S SIM (x_j,y_j) carry out cumulative mean, obtain flame gray level image x to be measured and standard flame source gray-scale mapThe average structure likeness coefficient MSSIM (x, y) of picture y:

$\mathrm{MSSIM}(x,y)=\frac{1}{M}\underset{j=1}{\overset{M}{\mathrm{\Σ}}}\mathrm{SSIM}(x_j,y_j)---\left(1\right)$

In formula (1): SSIM (x_j,y_j) be flame gray level image piece x to be measured_jWith corresponding standard flame source gray level image piece y_jKnotStructure likeness coefficient:

$\mathrm{SSIM}(x_j,y_j)=\frac{{({2u}_{x,j}{u}_{y,j}+C_1)}^2}{{({u_{x,j}}^2+{u_{y,j}}^2+C_2)}^2}\·\frac{{2\mathrm{\σ}}_{\mathrm{xy},j}+C_2}{{{\mathrm{\σ}}_{x,j}}^2+{{\mathrm{\σ}}_{y,j}}^2+C_2}---\left(2\right)$

In formula (2): C₁＝6.5025，C₂＝58.5225；

u_x，jFor flame gray level image piece x to be measured_jAverage:

$u_{x,j}=\frac{1}{N}\underset{i=1}{\overset{N}{\mathrm{\Σ}}}{x}_{j,i}---\left(3\right)$

u_y,jFor standard flame source gray level image piece y_jAverage:

$u_{y,j}=\frac{1}{N}\underset{i=1}{\overset{N}{\mathrm{\Σ}}}{y}_{j,i}---\left(4\right)$

σ_x,jFor flame gray level image piece x to be measured_jStandard deviation:

${\mathrm{\σ}}_{x,j}={\left(\frac{1}{N-1}\underset{i=1}{\overset{N}{\mathrm{\Σ}}}{(x_{j,i}-u_{x,j})}^2\right)}^{1/2}---\left(5\right)$

σ_y,jFor standard flame source gray level image piece y_jStandard deviation:

${\mathrm{\σ}}_{y,j}={\left(\frac{1}{N-1}\underset{i=1}{\overset{N}{\mathrm{\Σ}}}{(y_{j,i}-u_{y,j})}^2\right)}^{1/2}---\left(6\right)$

σ_xy,jFor standard flame source gray level image piece y_jWith flame gray level image piece x to be measured_jBetween covariance:

${\mathrm{\σ}}_{\mathrm{xy},j}=\frac{1}{N-1}\underset{i=1}{\overset{N}{\mathrm{\Σ}}}(x_{j,i}-u_{x,j})(y_{j,i}-u_{y,j})---\left(7\right)$

In formula (3)～(7): N is flame gray level image piece x to be measured_jWith standard flame source gray level image piece y_jPixel number;

x_j,iFor flame gray level image piece x to be measured_jThe value of i pixel;

y_j,iFor standard flame source gray level image piece y_jThe value of i pixel.

Owing to adopting technique scheme, the present invention compared with prior art tool has the following advantages:

The present invention adopts structural similarity index first in the technical field of rotary kiln flame image identification, from brightness, contrastDifferentiate rotary kiln with three kinds of different angles of structure and burn till state, discrimination precision is higher; In addition, the method belongs to one and more meetsThe image quality evaluating method of human visual system's general principle, differentiates the result subjective assessment of more fitting.

The method of discrimination that the present invention uses, without flame image is carried out to special training and study, greatly reduces calculation of complexDegree, has dwindled handling duration.

The present invention can independently differentiate the state that burns till of single width flame image, meet on-line real time monitoring rotary kiln Flame changeCondition, can find within the extremely short time that rotary kiln flame burns till the variation of state, and make in time corresponding adjustment, thereby increasesThe safety and reliability of rotary kiln, makes the clinker quality of production more stable, can greatly reduce industrial productionCost, lays the foundation for realizing the closed-loop control of monitoring in real time to whole rotary system, for really realizing with " machine is seen fire "Replace " manually seeing fire " that basic guarantee is provided.

Therefore, the present invention has that precision is high, computation complexity is low, processing procedure is short and can realize on-line real time monitoring flame statusThe feature changing.

Brief description of the drawings

Below in conjunction with drawings and the embodiments, the present invention is described in further detail:

Fig. 1 is a kind of flow chart of the present invention;

Fig. 2 is the flame image P to be measured in embodiment 1;

Fig. 3 is the flame gray level image x to be measured in embodiment 1;

Fig. 4 is the flame image P to be measured in embodiment 2;

Fig. 5 is the flame gray level image x to be measured in embodiment 2;

Fig. 6 is the flame image P to be measured in embodiment 3;

Fig. 7 is the flame gray level image x to be measured in embodiment 3;

Fig. 8 is the flame image P to be measured in embodiment 4;

Fig. 9 is the flame gray level image x to be measured in embodiment 4.

Detailed description of the invention

Below in conjunction with the drawings and specific embodiments, the invention will be further described, not the restriction to its protection domain:

Embodiment 1

A kind of Alumina Rotary Kiln based on flame image structural similarity burns till state identification method. This rotary kiln burns till state to be knownThe concrete steps of other method are as shown in Figure 1:

The first step, standard flame source image Q operate expert by Alumina Rotary Kiln and demarcate, and standard flame source image Q is of a size of512 × 384 × 3. Standard flame source image Q composition standard flame source picture library L, standard flame source picture library L is divided into normal flame picture library LNWith abnormal flame picture library LA, normal flame picture library LN is made up of the standard flame source image Q of 20 width normal conditions, abnormal flamePicture library LA is made up of the standard flame source image Q of 20 width abnormalities; Standard flame source image Q is carried out to filtering and greyscale transformation,Obtain standard flame source gray level image y; All standard flame source gray level image y composition standard flame source gray scale picture library F, standard flame source gray scalePicture library F is divided into normal flame gray scale picture library FN and abnormal flame gray scale picture library FA.

The described wavelet decomposition that is filtered into is extracted low pass part.

Described standard flame source gray level image y is of a size of 256 × 192.

Second step, from gather Alumina Rotary Kiln clinkering zone flame video obtain width colour as shown in Figure 2 flame to be measuredImage P, flame image P to be measured is of a size of 512 × 384 × 3; Flame image P to be measured is carried out to wavelet decomposition and extract low pass partAnd greyscale transformation, obtain being of a size of as shown in Figure 3 256 × 192 flame gray level image x to be measured.

The computational methods of the 3rd step, employing average structure likeness coefficient, calculate a flame gray level image x to be measured and each widthAverage structure likeness coefficient MSSIM (x, y) between standard flame source gray level image y, obtains this flame gray level image x to be measuredAnd 20 average structure likeness coefficients between the normal flame gray scale of 20 width picture library FN Plays flame gray level image yMSSIM (x, y) obtains this flame gray level image x to be measured and the abnormal flame gray scale of 20 width picture library FA Plays flame ash simultaneously20 average structure likeness coefficient MSSIM (x, y) between degree image y.

40 average structure likeness coefficient MSSIM (x, y) are as shown in table 1.

Table 1

The 4th step, according to shown in table 1,40 average structure likeness coefficient MSSIM (x, y) that the 3rd step is obtained compare, choosing average structure likeness coefficient maximum MAXmssim is 0.6238. Due to average structure likeness coefficient maximumThe corresponding standard flame source gray level image of MAXmssim y belongs to normal flame gray scale picture library FN, judges flame image P to be measuredThe state that burns till be normal condition.

If it is normal condition that the 5th step is judged the state that burns till of flame image P to be measured, directly carries out next width flame image and burn tillThe identification of state;

If judge, the state that burns till of flame image P to be measured is abnormality, and system alarm, improves rotary kiln flame and burn till after state,Carry out again the identification that next width flame image burns till state.

Because the condition judgement that burns till of this flame image P to be measured is normal condition, burn till therefore directly carry out next width flame imageThe identification of state.

The 6th step, repetition the second～five step, until finish.

The computational methods of the average structure likeness coefficient described in the present embodiment are: by flame gray level image x to be measured and standard flame sourceGray level image y respectively with 4 × 4 window by pixel move to the lower right corner from the upper left corner, obtain 193929 flame gray scales to be measuredImage block x_jWith 193929 standard flame source gray level image piece y_j, calculate each flame gray level image piece x to be measured_jWith corresponding markAccurate flame gray level image piece y_jBetween structural similarity coefficient S SIM (x_j,y_j), obtain 193929 structural similarity property coefficientsSSIM(x_j,y_j), then to 193929 structural similarity coefficient S SIM (x_j,y_j) carry out cumulative mean, obtain flame ash to be measuredThe average structure likeness coefficient MSSIM (x, y) of degree image x and standard flame source gray level image y:

$\mathrm{MSSIM}(x,y)=\frac{1}{M}\underset{j=1}{\overset{M}{\mathrm{\Σ}}}\mathrm{SSIM}(x_j,y_j)---\left(1\right)$

In formula (1): M=193929;

SSIM(x_j,y_j) be flame gray level image piece x to be measured_jWith corresponding standard flame source gray level image piece y_j'sStructural similarity property coefficient:

$\mathrm{SSIM}(x_j,y_j)=\frac{{({2u}_{x,j}{u}_{y,j}+C_1)}^2}{{({u_{x,j}}^2+{u_{y,j}}^2+C_2)}^2}\·\frac{{2\mathrm{\σ}}_{\mathrm{xy},j}+C_2}{{{\mathrm{\σ}}_{x,j}}^2+{{\mathrm{\σ}}_{y,j}}^2+C_2}---\left(2\right)$

In formula (2): C₁＝6.5025，C₂＝58.5225；

u_x,jFor flame gray level image piece x to be measured_jAverage:

$u_{x,j}=\frac{1}{N}\underset{i=1}{\overset{N}{\mathrm{\Σ}}}{x}_{j,i}---\left(3\right)$

u_y,jFor standard flame source gray level image piece y_jAverage:

$u_{y,j}=\frac{1}{N}\underset{i=1}{\overset{N}{\mathrm{\Σ}}}{y}_{j,i}---\left(4\right)$

σ_x,jFor flame gray level image piece x to be measured_jStandard deviation:

${\mathrm{\σ}}_{x,j}={\left(\frac{1}{N-1}\underset{i=1}{\overset{N}{\mathrm{\Σ}}}{(x_{j,i}-u_{x,j})}^2\right)}^{1/2}---\left(5\right)$

σ_y,jFor standard flame source gray level image piece y_jStandard deviation:

${\mathrm{\σ}}_{y,j}={\left(\frac{1}{N-1}\underset{i=1}{\overset{N}{\mathrm{\Σ}}}{(y_{j,i}-u_{y,j})}^2\right)}^{1/2}---\left(6\right)$

σ_xy,jFor standard flame source gray level image piece y_jWith flame gray level image piece x to be measured_jBetween covariance:

${\mathrm{\σ}}_{\mathrm{xy},j}=\frac{1}{N-1}\underset{i=1}{\overset{N}{\mathrm{\Σ}}}(x_{j,i}-u_{x,j})(y_{j,i}-u_{y,j})---\left(7\right)$

In formula (3)～(7): N=16 is flame gray level image piece x to be measured_jWith standard flame source gray level image piece y_jPixelNumber;

x_j,iFor flame gray level image piece x to be measured_jThe value of i pixel;

y_j,iFor standard flame source gray level image piece y_jThe value of i pixel.

Embodiment 2

A kind of Alumina Rotary Kiln based on flame image structural similarity burns till state identification method. This rotary kiln burns till state to be knownThe concrete steps of other method are as shown in Figure 1:

Described in the first step, this step, be filtered into 20 rank Butterworth LPF filtering; Described standard flame source gray level image yBe of a size of 512 × 384. All the other are with embodiment 1 first step.

Second step, from gather Alumina Rotary Kiln clinkering zone flame video obtain width colour as shown in Figure 4 flame to be measuredImage P, flame image P to be measured is of a size of 512 × 384 × 3; By 20 rank Butterworth LPFs for flame image P to be measuredDevice filtering is also carried out greyscale transformation, obtains being of a size of as shown in Figure 5 512 × 384 flame gray level image x to be measured.

The 3rd step, this step except 40 average structure likeness coefficient MSSIM (x, y) as shown in table 2 outside, all the other same embodiment1 the 3rd step.

Table 2

The 4th step, according to shown in table 2,40 average structure likeness coefficient MSSIM (x, y) that the 3rd step is obtained compare, choosing average structure likeness coefficient maximum MAXmssim is 0.6439. Due to average structure likeness coefficient maximumThe corresponding standard flame source gray level image of MAXmssim y belongs to abnormal flame gray scale picture library FA, judges flame image P to be measuredThe state that burns till be abnormality.

If it is normal condition that the 5th step is judged the state that burns till of flame image P to be measured, directly carries out next width flame image and burn tillThe identification of state;

If judge, the state that burns till of flame image P to be measured is abnormality, and system alarm, improves rotary kiln flame and burn till after state,Carry out again the identification that next width flame image burns till state.

Because the condition judgement that burns till of this flame image P to be measured is abnormality, therefore system alarm improves rotary kiln flame and burnsAfter one-tenth state, then carry out the identification that next width flame image burns till state.

The 6th step, repetition the second～five step, until finish.

The computational methods of the average structure likeness coefficient described in the present embodiment are: by flame gray level image x to be measured and standard flame sourceGray level image y respectively with 8 × 8 window by pixel move to the lower right corner from the upper left corner, obtain 190385 flame gray scales to be measuredImage block x_jWith 190385 standard flame source gray level image piece y_j, calculate each flame gray level image piece x to be measured_jWith corresponding markAccurate flame gray level image piece y_jBetween structural similarity coefficient S SIM (x_j,y_j), obtain 190385 structural similarity property coefficientsSSIM(x_j,y_j), then to 190385 structural similarity coefficient S SIM (x_j,y_j) carry out cumulative mean, obtain flame ash to be measuredThe average structure likeness coefficient MSSIM (x, y) of degree image x and standard flame source gray level image y:

$\mathrm{MSSIM}(x,y)=\frac{1}{M}\underset{j=1}{\overset{M}{\mathrm{\Σ}}}\mathrm{SSIM}(x_j,y_j)---\left(1\right)$

In formula (1): M=190385;

SSIM(x_j,y_j) be flame gray level image piece x to be measured_jWith corresponding standard flame source gray level image piece y_j'sStructural similarity property coefficient:

$\mathrm{SSIM}(x_j,y_j)=\frac{{({2u}_{x,j}{u}_{y,j}+C_1)}^2}{{({u_{x,j}}^2+{u_{y,j}}^2+C_2)}^2}\·\frac{{2\mathrm{\σ}}_{\mathrm{xy},j}+C_2}{{{\mathrm{\σ}}_{x,j}}^2+{{\mathrm{\σ}}_{y,j}}^2+C_2}---\left(2\right)$

In formula (2): C₁＝6.5025，C₂＝58.5225；

u_x,jFor flame gray level image piece x to be measured_jAverage:

$u_{x,j}=\frac{1}{N}\underset{i=1}{\overset{N}{\mathrm{\Σ}}}{x}_{j,i}---\left(3\right)$

u_y,jFor standard flame source gray level image piece y_jAverage:

$u_{y,j}=\frac{1}{N}\underset{i=1}{\overset{N}{\mathrm{\Σ}}}{y}_{j,i}---\left(4\right)$

σ_x,jFor flame gray level image piece x to be measured_jStandard deviation:

${\mathrm{\σ}}_{x,j}={\left(\frac{1}{N-1}\underset{i=1}{\overset{N}{\mathrm{\Σ}}}{(x_{j,i}-u_{x,j})}^2\right)}^{1/2}---\left(5\right)$

σ_y，jFor standard flame source gray level image piece y_jStandard deviation:

${\mathrm{\σ}}_{y,j}={\left(\frac{1}{N-1}\underset{i=1}{\overset{N}{\mathrm{\Σ}}}{(y_{j,i}-u_{y,j})}^2\right)}^{1/2}---\left(6\right)$

σ_xy,jFor standard flame source gray level image piece y_jWith flame gray level image piece x to be measured_jBetween covariance:

${\mathrm{\σ}}_{\mathrm{xy},j}=\frac{1}{N-1}\underset{i=1}{\overset{N}{\mathrm{\Σ}}}(x_{j,i}-u_{x,j})(y_{j,i}-u_{y,j})---\left(7\right)$

In formula (3)～(7): N=64 is flame gray level image piece x to be measured_jWith standard flame source gray level image piece y_jPixelNumber;

x_j,iFor flame gray level image piece x to be measured_jThe value of i pixel;

y_j,iFor standard flame source gray level image piece y_jThe value of i pixel.

Embodiment 3

A kind of cement rotary kiln based on flame image structural similarity burns till state identification method. This rotary kiln burns till state recognitionThe concrete steps of method are as shown in Figure 1:

The first step, standard flame source image Q operate expert by cement rotary kiln and demarcate, and standard flame source image Q is of a size of352 × 288 × 3. Standard flame source image Q composition standard flame source picture library L, standard flame source picture library L is divided into normal flame picture library LNWith abnormal flame picture library LA, normal flame picture library LN is made up of the standard flame source image Q of 15 width normal conditions, abnormal flamePicture library LA is made up of the standard flame source image Q of 10 width abnormalities; Standard flame source image Q is carried out to filtering and greyscale transformation,Obtain standard flame source gray level image y; All standard flame source gray level image y composition standard flame source gray scale picture library F, standard flame source gray scalePicture library F is divided into normal flame gray scale picture library FN and abnormal flame gray scale picture library FA.

The described threshold filter that is filtered into.

Described standard flame source gray level image y is of a size of 352 × 288.

Second step, from gather cement rotary kiln clinkering zone flame video obtain width colour as shown in Figure 6 flame figure to be measuredPicture P, flame image P to be measured is of a size of 352 × 288 × 3; Flame image P to be measured is carried out to threshold filter and greyscale transformation,To the flame gray level image x to be measured that is of a size of as shown in Figure 7 352 × 288.

The computational methods of the 3rd step, employing average structure likeness coefficient, calculate a flame gray level image x to be measured and each widthAverage structure likeness coefficient MSSIM (x, y) between standard flame source gray level image y, obtains this flame gray level image x to be measuredAnd 15 average structure likeness coefficients between the normal flame gray scale of 15 width picture library FN Plays flame gray level image yMSSIM (x, y) obtains this flame gray level image x to be measured and the abnormal flame gray scale of 10 width picture library FA Plays flame ash simultaneously10 average structure likeness coefficient MSSIM (x, y) between degree image y.

25 average structure likeness coefficient MSSIM (x, y) are as shown in table 3.

Table 3

The 4th step, according to shown in table 3,25 average structure likeness coefficient MSSIM (x, y) that the 3rd step is obtained compare, choosing average structure likeness coefficient maximum MAXmssim is 0.7616. Due to average structure likeness coefficient maximumThe corresponding standard flame source gray level image of MAXmssim y belongs to normal flame gray scale picture library FN, judges flame image P to be measuredThe state that burns till be normal condition.

The 5th step, with the 5th step of embodiment 1.

The 6th step, with the 6th step of embodiment 1.

The computational methods of described average structure likeness coefficient are: by flame gray level image x to be measured and standard flame source gray level imageY respectively with 11 × 11 window by pixel move to the lower right corner from the upper left corner, obtain 95076 flame gray level image piece x to be measured_jWith 95076 standard flame source gray level image piece y_j, calculate each flame gray level image piece x to be measured_jWith corresponding standard flame source gray scaleImage block y_jBetween structural similarity coefficient S SIM (x_j,y_j), obtain 95076 structural similarity coefficient S SIM (x_j,y_j)，Then to 95076 structural similarity coefficient S SIM (x_j,y_j) carry out cumulative mean, obtain flame gray level image x to be measured and standardThe average structure likeness coefficient MSSIM (x, y) of flame gray level image y:

$\mathrm{MSSIM}(x,y)=\frac{1}{M}\underset{j=1}{\overset{M}{\mathrm{\Σ}}}\mathrm{SSIM}(x_j,y_j)---\left(1\right)$

In formula (1): M=95076;

SSIM(x_j,y_j) be flame gray level image piece x to be measured_jWith corresponding standard flame source gray level image piece y_j'sStructural similarity property coefficient:

$\mathrm{SSIM}(x_j,y_j)=\frac{{({2u}_{x,j}{u}_{y,j}+C_1)}^2}{{({u_{x,j}}^2+{u_{y,j}}^2+C_2)}^2}\·\frac{{2\mathrm{\σ}}_{\mathrm{xy},j}+C_2}{{{\mathrm{\σ}}_{x,j}}^2+{{\mathrm{\σ}}_{y,j}}^2+C_2}---\left(2\right)$

In formula (2): C₁＝6.5025，C₂＝58.5225；

u_x,jFor flame gray level image piece x to be measured_jAverage:

$u_{x,j}=\frac{1}{N}\underset{i=1}{\overset{N}{\mathrm{\Σ}}}{x}_{j,i}---\left(3\right)$

u_y,jFor standard flame source gray level image piece y_jAverage:

$u_{y,j}=\frac{1}{N}\underset{i=1}{\overset{N}{\mathrm{\Σ}}}{y}_{j,i}---\left(4\right)$

σ_x,jFor flame gray level image piece x to be measured_jStandard deviation:

${\mathrm{\σ}}_{x,j}={\left(\frac{1}{N-1}\underset{i=1}{\overset{N}{\mathrm{\Σ}}}{(x_{j,i}-u_{x,j})}^2\right)}^{1/2}---\left(5\right)$

σ_y,jFor standard flame source gray level image piece y_jStandard deviation:

${\mathrm{\σ}}_{y,j}={\left(\frac{1}{N-1}\underset{i=1}{\overset{N}{\mathrm{\Σ}}}{(y_{j,i}-u_{y,j})}^2\right)}^{1/2}---\left(6\right)$

σ_xy,jFor standard flame source gray level image piece y_jWith flame gray level image piece x to be measured_jBetween covariance:

${\mathrm{\σ}}_{\mathrm{xy},j}=\frac{1}{N-1}\underset{i=1}{\overset{N}{\mathrm{\Σ}}}(x_{j,i}-u_{x,j})(y_{j,i}-u_{y,j})---\left(7\right)$

In formula (3)～(7): N=121 is flame gray level image piece x to be measured_jWith standard flame source gray level image piece y_jPixelNumber;

x_j,iFor flame gray level image piece x to be measured_jThe value of i pixel;

y_j,iFor standard flame source gray level image piece y_jThe value of i pixel.

Embodiment 4

A kind of cement rotary kiln based on flame image structural similarity burns till state identification method. This rotary kiln burns till state recognitionThe concrete steps of method are as shown in Figure 1:

Described in the first step, this step, be filtered into region of interesting extraction; Described standard flame source gray level image y is of a size of225 × 189. All the other are with embodiment 3 first steps.

Second step, from gather cement rotary kiln clinkering zone flame video obtain width colour as shown in Figure 8 flame figure to be measuredPicture P, flame image P to be measured is of a size of 352 × 288 × 3; Flame image P to be measured is carried out to region of interesting extraction and gray scale changeChange, obtain being of a size of as shown in Figure 9 225 × 189 flame gray level image x to be measured.

The 3rd step, this step except 25 average structure likeness coefficient MSSIM (x, y) as shown in table 4 outside, all the other same embodiment3 the 3rd step.

Table 4

The 4th step, according to shown in table 4,25 average structure likeness coefficient MSSIM (x, y) that the 3rd step is obtained compare, choosing average structure likeness coefficient maximum MAXmssim is 0.7720. Due to average structure likeness coefficient maximumThe corresponding standard flame source gray level image of MAXmssim y belongs to abnormal flame gray scale picture library FA, judges flame image P to be measuredThe state that burns till be abnormality.

The 5th step, with the 5th step of embodiment 2.

The 6th step, with the 6th step of embodiment 2.

The computational methods of described average structure likeness coefficient are with the computational methods of the average structure likeness coefficient of embodiment 3.

This detailed description of the invention compared with prior art tool has the following advantages:

This detailed description of the invention adopts structural similarity index first in the technical field of rotary kiln flame image identification, from brightDegree, contrast and three kinds of different angles of structure are differentiated rotary kiln and are burnt till state, and discrimination precision is higher; In addition, the method belongs toMore meet an image quality evaluating method for human visual system's general principle, differentiate the result subjective assessment of more fitting.

The method of discrimination that this detailed description of the invention is used, without flame image is carried out to special training and study, greatly reduces meterCalculate complexity, dwindled handling duration.

This detailed description of the invention can independently be differentiated the state that burns till of single width flame image, meets on-line real time monitoring rotary kiln moderate heatThe condition that flame changes, can find within the extremely short time that rotary kiln flame burns till the variation of state, and make in time corresponding adjustment,Thereby increase the safety and reliability of rotary kiln, make the clinker quality of production more stable, can greatly reduceIndustrial production cost, lays the foundation for realizing the closed-loop control of monitoring in real time to whole rotary system, for really realizing with " machineDevice is seen fire " replacement " is manually seen fire " basic guarantee is provided.

Therefore, this detailed description of the invention has that precision is high, computation complexity is low, processing procedure is short and can realize on-line real time monitoringThe feature that flame status changes.

Claims (3)

1. the rotary kiln based on flame image structural similarity burns till a state identification method, and described rotary kiln burns till state recognitionThe concrete steps of method:

The first step, standard flame source image Q are demarcated by kiln operation expert; Standard flame source image Q composition standard flame source picture library L,Standard flame source picture library L is divided into normal flame picture library LN and abnormal flame picture library LA, and normal flame picture library LN is by the normal shape of a widthThe standard flame source image Q composition of state, abnormal flame picture library LA is made up of the standard flame source image Q of b width abnormality; To markAccurate flame image Q carries out filtering processing and greyscale transformation, obtains standard flame source gray level image y; All standard flame source gray level imagesY composition standard flame source gray scale picture library F, standard flame source gray scale picture library F is divided into normal flame gray scale picture library FN and abnormal flame gray scalePicture library FA;

It is characterized in that:

Second step, from the flame video of rotary kiln clinkering zone gathering, obtain a flame image P to be measured; By flame figure to be measuredCarry out filtering processing and greyscale transformation as P, obtain flame gray level image x to be measured;

The computational methods of the 3rd step, employing average structure likeness coefficient, calculate a flame gray level image x to be measured and each widthAverage structure likeness coefficient MSSIM (x, y) between standard flame source gray level image y, obtains a+b average structure similitude systemNumber MSSIM (x, y);

The 4th step, a+b the average structure likeness coefficient MSSIM (x, y) that the 3rd step is obtained compare, and choose average knotStructure likeness coefficient maximum MAXmssim; If the corresponding standard of average structure likeness coefficient maximum MAXmssimFlame gray level image y belongs to normal flame gray scale picture library FN, and the state that burns till of judging flame image P to be measured is normal condition;If the corresponding standard flame source gray level image of average structure likeness coefficient maximum MAXmssim y belongs to abnormal flame gray scalePicture library FA, the state that burns till of judging flame image P to be measured is abnormality;

If it is normal condition that the 5th step is judged the state that burns till of flame image P to be measured, directly carries out next width flame image and burn tillThe identification of state;

If judge, the state that burns till of flame image P to be measured is abnormality, and system alarm, improves rotary kiln flame and burn till after state,Carry out again the identification that next width flame image burns till state;

The 6th step, repetition the second～five step, until finish.

2. burn till state identification method according to the rotary kiln based on flame image structural similarity claimed in claim 1, its featureBe that described a is 1～1000 natural number; Described b is 1～1000 natural number.

3. burn till state identification method according to the rotary kiln based on flame image structural similarity claimed in claim 1, its featureThe computational methods that are described average structure likeness coefficient are: by flame gray level image x to be measured and standard flame source gray level imageY respectively with the window of formed objects by pixel move to the lower right corner from the upper left corner, obtain M flame gray level image piece x to be measured_jWith M standard flame source gray level image piece y_j, calculate each flame gray level image piece x to be measured_jWith corresponding standard flame source gray-scale mapPicture piece y_jBetween structural similarity coefficient S SIM (x_j,y_j), obtain M structural similarity coefficient S SIM (x_j,y_j); Then rightM structural similarity coefficient S SIM (x_j,y_j) carry out cumulative mean, obtain flame gray level image x to be measured and standard flame source gray-scale mapThe average structure likeness coefficient MSSIM (x, y) of picture y:

$MSSIM(x,y)=\frac{1}{M}\underset{j=1}{\overset{M}{\Σ}}SSIM(x_j,y_j)---\left(1\right)$

In formula (1): SSIM (x_j,y_j) be flame gray level image piece x to be measured_jWith corresponding standard flame source gray level image piece y_jKnotStructure likeness coefficient:

$SSIM(x_j,y_j)=\frac{{(2u_{x,j}{u}_{y,j}+C_1)}^2}{{({u_{x,j}}^2+{u_{y,j}}^2+C_2)}^2}\·\frac{2{\mathrm{\σ}}_{xy,j}+C_2}{{{\mathrm{\σ}}_{x,j}}^2+{{\mathrm{\σ}}_{y,j}}^2+C_2}---\left(2\right)$

In formula (2): C₁＝6.5025，C₂＝58.5225；

u_x,jFor flame gray level image piece x to be measured_jAverage:

$u_{x,j}=\frac{1}{N}\underset{i=1}{\overset{N}{\Σ}}{x}_{j,i}---\left(3\right)$

u_y,jFor standard flame source gray level image piece y_jAverage:

$u_{y,j}=\frac{1}{N}\underset{i=1}{\overset{N}{\Σ}}{y}_{j,i}---\left(4\right)$

σ_x,jFor flame gray level image piece x to be measured_jStandard deviation:

${\mathrm{\σ}}_{x,j}={\left(\frac{1}{N-1}\underset{i=1}{\overset{N}{\Σ}}{\left(x_{j,i}-u_{x,j}\right)}^2\right)}^{1/2}---\left(5\right)$

σ_y,jFor standard flame source gray level image piece y_jStandard deviation:

${\mathrm{\σ}}_{y,j}={\left(\frac{1}{N-1}\underset{i=1}{\overset{N}{\Σ}}{\left(y_{j,i}-u_{y,j}\right)}^2\right)}^{1/2}---\left(6\right)$

σ_xy,jFor standard flame source gray level image piece y_jWith flame gray level image piece x to be measured_jBetween covariance:

${\mathrm{\σ}}_{xy,j}=\frac{1}{N-1}\underset{i=1}{\overset{N}{\Σ}}(x_{j,i}-u_{x,j})(y_{j,i}-u_{y,j})---\left(7\right)$

In formula (3)～(7): N is flame gray level image piece x to be measured_jWith standard flame source gray level image piece y_jPixel number;

x_j,iFor flame gray level image piece x to be measured_jThe value of i pixel;

y_j,iFor standard flame source gray level image piece y_jThe value of i pixel.