CN100383502C - Historical trend analysis method for infrared sequence image - Google Patents

Abstract

The present invention relates to a historical trend analysis method for infrared sequence images, which belongs to the field of equipment fault diagnosis and solves the problem of registering the infrared thermal images so as to analyze a historical trend on the basis. The present invention has the following steps: 1, the infrared thermal images which are shot by equipment to be analyzed in different periods are sorted to an analysis file folder; 2, a user manually inputs position points and utilizes an affine transformation matrix, a projection transformation matrix or a quadric transformation matrix and bilinear interpolation to geometrically correct the images and register the images according to the spatial positions of reference images; 3, the characteristics of the surface scanning of a thermal imager is utilized to search on the images; thus, the position and the temperature of each interest point can be obtained, and a temperature change curve of the equipment in each of the interest points in different periods is drawn and displayed. The present invention realizes the historical trend analysis of the infrared thermal images of the equipment and adds a dynamic analysis capability for the infrared thermal imager; moreover, the present invention can observe abnormal temperature changes and find early fault symptoms through the comparison with the changes of history temperatures for precaution.

Description

Infrared sequence image historical trend analytical approach

Technical field

The invention belongs to the equipment fault diagnosis field, propose a kind of historical trend analytical approach of utilizing infrared sequence image to carry out equipment fault diagnosis.

Background technology

Thermal infrared imager and image analysis software thereof have been widely used in the infrared equipment fault diagnosis field at present, in the practicality mainly are the temperature of equipment to be measured the hot stall of discovering device by the temp sensing function of thermal infrared imager.With respect to the contact type measurement means, thermal imaging system is contactless online detection, promptly can carry out online non-contacting temperature survey under the non-stop-machine situation of equipment; With respect to a Wen Yi, thermal imaging system is face scanning, and what it was gathered is a temperature value of being had a few on the projection plane; And its visual result that provides, strengthening by pseudo-colours, high temperature and low-temperature region can show in mode intuitively.

Thermal infrared imager and software thereof only limit to the free hand drawing analysis in the use at present, and the major function of software is that the single image of gathering is carried out basic Flame Image Process (as processing such as filtering sharpening and pseudo-colours enhancings), thermometric and assistant analysis and generation report etc.This method is a kind of method of static state in essence, being applicable to the quick localization of fault after fault has taken place, is exactly specifically when equipment hot stall takes place, uses thermal imaging system that equipment is scanned, utilize the above-mentioned advantage of thermal imaging system, find the position that hot stall takes place rapidly.

If the user need inquire about the account of the history of certain equipment, must search historical thermal-induced imagery by hand at present, analyze according to the function of free hand drawing analysis again.Because the relative position of thermal imaging system and equipment, angle etc. are all multifactor impossible just the same in the time of each the shooting, the infrared chart of the equipment of the same race that obtains like this exists difference on the locus.Therefore, in present infrared software, if the user need do long-term observation and record to equipment, can only adopt manual interest point annotation and use other means records, inefficiency so on the one hand, lost the advantage of thermal imaging system distinctive " face scanning " on the other hand, thermal imaging system has been degenerated to the function of some Wen Yi and has been used.User's needs of history curve are analyzed and drawn to the historical process that device temperature is changed, and present application all can't be satisfied.

Summary of the invention

The invention provides a kind of infrared sequence image historical trend analytical approach, solve the registration problems of thermal-induced imagery, on the infrared image of registration, eliminate the difference of locus, carry out the historical trend analysis on this basis again, raise the efficiency, give full play to the advantage of thermal imaging system " face scanning ", give thermal infrared imager with the performance analysis ability.

A kind of infrared sequence image historical trend analytical approach of the present invention comprises the steps: in regular turn

(1) thermal-induced imagery that the different times of equipment to be analyzed is taken is put in order in the Study document folder;

(2) user imports anchor point by hand, utilize affine transformation matrix, projective transformation matrix or quadratic form transformation matrix, use bilinear interpolation method then, carry out the geometry correction operation, to the above-mentioned geometric correction of imagery, with the locus registration of these images according to reference picture;

(3) utilize the characteristics of thermal imaging system " face scanning ", search and pick up point of interest on image obtains each point of interest position and temperature, draws and shows that this equipment is at this point of interest different times variation of temperature curve map.

Described infrared sequence image historical trend analytical approach is characterized in that: the temperature variation curve data importing of described point of interest is further analyzed in the Excel software and reported; The user can also be comprising that all of proofreading and correct back image, point of interest position and temperature analyze data and be saved in the video file, makes things convenient for the user to analyze the interchange of data and share.

This method can realize the historical trend analysis of equipment infrared chart, help user's apparatus for establishing temperature variation archives, put out the device temperature Changing Pattern in order, know " habit " of equipment well, for this static analysis tools of thermal infrared imager increases the performance analysis ability.On the other hand, the generation of fault normally has a process, and this method can have the early stage sign that unusual temperature variation is found fault by the comparative observation with historical temperature variations before fault takes place as yet, accomplish " taking precautions against in possible trouble ".

Description of drawings

Fig. 1 is a FB(flow block) of the present invention;

Fig. 2 is the realization flow block diagram of folder management;

Fig. 3 is that infrared file management facilities realize block diagram;

Fig. 4 is an infrared sequence image of taking, do not do geometry correction behind the daylight lamp powered-down continuously;

Fig. 5 is the temperature variation of image shown in Figure 4;

Fig. 6 is the sequence image behind the image rectification shown in Figure 4;

Fig. 7 is the temperature variation of image shown in Figure 6.

Embodiment

Below in conjunction with Fig. 1 flow process of the present invention is described:

The applicant has realized infrared file management facilities in the infrared analysis software of computing machine, the thermal-induced imagery that the user can be in analysis software takes different times carries out Classification Management, with the different times image of equipment to be analyzed by in this finishing functions to one Study document folder; Secondly, by the anchor point of the manual input of user, utilize affine, projection or quadratic form transformation matrix that these images are carried out geometry corrections operation, with the locus registration of these images according to reference picture.Then, the user can utilize the advantage of thermal imaging system " face scanning " to search for and pick up point of interest on image, and software shows the temperature changing curve diagram of this equipment at this spatial point different times automatically.End user can further analyze the temperature variation curve data importing of point of interest in other softwares such as Excel or Avi and report, the user can also analyze data (comprise and proofread and correct back image, point of interest position and temperature) to all and be saved in the video file, makes things convenient for the user to analyze the interchange of data and shares.Excel is the graphic analyses instrument of the Office of MS external member; Avi is a Windows normal video file layout.

1. file management facilities:

File management facilities comprise two parts: folder management function and infrared file management facilities.This part function realizes by accessing operation system file system function.At first call the file directory tree that obtains in the computer system by file system function, the navigation feature of carrying out catalogue as the user after certain file is double-clicked mouse enters this file, and then obtain all infrared files in the presents folder by system function call, read in infrared file successively and thermal map is presented on GUI (graphical user interface) interface.When user's click right on the directory tree assembly ejects the folder management menu, comprise: newly-built, rename, deleted file folder and menu such as all choose, on gui interface, after the user chooses certain width of cloth infrared chart, the right button of clicking the mouse on the interface ejects the file management menu, comprise: copy, paste, deletion, functions such as shearing and rename, in gui interface, also shown the current attribute of choosing thermal map, these attributes comprise the thermometric shelves that obtain from infrared file, emissivity, information such as environment temperature, also comprise the device name that the user imports, the image acquisition time, information such as apparatus of load, these information are kept in the current file folder with the form of text.

Fig. 2 is the realization flow block diagram of folder management, the function of wherein indicating (os) is the realization of call operation system function call, do not comprise abnormality processing (such as treatment of special situation such as file duplications of name) in the block diagram, to take place all are unusual, all stop current operation, and eject the prompted dialog frame:

The right button of clicking the mouse in the image workspace of GUI ejects the file management menu, and Fig. 3 is that infrared file management facilities realize block diagram." common data structures " comprises following information among the figure: the command word of filename tabulation, place file, operation information such as (being to shear or duplicate)." attribute " refers to the attribute of infrared chart among the figure, these attributes comprise information such as the thermometric shelves that obtain, emissivity, environment temperature from the infrared chart file, also comprise information such as device name that the user imports, image acquisition time, apparatus of load, these information are kept in the current file folder with the form of text, when the execute file bookkeeping, need safeguard these attribute informations, guarantee between attribute and the file synchronously.

The realization of these functions is all called file system function and is called realization, is example with shearing, paste operation below, and implementation procedure is described.After the user chooses one or more infrared charts, after clicking shearing, the button.onrelease of shearing the menu correspondence with filename be saved in above-mentioned " common data structures " in, after navigating to another file during click paste, software is inquired about this data structure and is obtained the filename tabulation, and calling system funcall execution replicate run is duplicated these files to current file folder (also need duplicate former figure attribute information), if action type is to shear, then the calling system funcall is deleted these files (also need delete former figure attribute information) in the original folder.

2. the geometry correction CALCULATION OF PARAMETERS of affined transformation:

If find out three point (r by reference map ₁, s ₁), (r ₂, s ₂), (r ₃, s ₃), with last three the point coordinate (u of distortion figure ₁, v ₁), (u ₂, v ₂), (u ₃, v ₃) corresponding one by one, then

x′＝h ₁(x，y)＝a ₀+a ₁x+a ₂y

y′＝h ₂(x，y)＝b ₀+b ₁x+b ₂y

The corresponding point substitution, write as matrix form

$\begin{array}{c}{\mathrm{<figure-callout\; id="1"\; label="u"\; filenames="C20041006138300152.png"\; state="\{\{state\}\}">u</figure-callout>}}_1=a_0+a_1{\mathrm{<figure-callout\; id="1"\; label="r"\; filenames="C20041006138300152.png"\; state="\{\{state\}\}">r</figure-callout>}}_1+a_2{s}_1\\ u_2=a_0+a_1{\mathrm{<figure-callout\; id="2"\; label="r"\; filenames="C20041006138300142.png"\; state="\{\{state\}\}">r</figure-callout>}}_2+a_2{s}_2\\ u_3=a_0+a_1{\mathrm{<figure-callout\; id="3"\; label="r"\; filenames="C20041006138300142.png,C20041006138300152.png"\; state="\{\{state\}\}">r</figure-callout>}}_3+a_2{s}_3\end{array}\left[\begin{array}{c}{u}_1\\ u_2\\ u_3\end{array}\right]=\left[\begin{array}{ccc}1& r_1& {\mathrm{<figure-callout\; id="1"\; label="s"\; filenames="C20041006138300152.png"\; state="\{\{state\}\}">s</figure-callout>}}_1\\ 1& r_2& {\mathrm{<figure-callout\; id="2"\; label="s"\; filenames="C20041006138300142.png"\; state="\{\{state\}\}">s</figure-callout>}}_2\\ 1& r_3& s_3\end{array}\right]\left[\begin{array}{c}{a}_0\\ a_1\\ a_2\end{array}\right]$

$\begin{array}{c}{\mathrm{<figure-callout\; id="1"\; label="v"\; filenames="C20041006138300152.png"\; state="\{\{state\}\}">v</figure-callout>}}_1={\mathrm{<figure-callout\; id="0"\; label="b"\; filenames="C20041006138300142.png,C20041006138300152.png"\; state="\{\{state\}\}">b</figure-callout>}}_0+b_1{\mathrm{<figure-callout\; id="1"\; label="r"\; filenames="C20041006138300152.png"\; state="\{\{state\}\}">r</figure-callout>}}_1+b_2{s}_1\\ {\mathrm{<figure-callout\; id="2"\; label="v"\; filenames="C20041006138300142.png"\; state="\{\{state\}\}">v</figure-callout>}}_2={\mathrm{<figure-callout\; id="0"\; label="b"\; filenames="C20041006138300142.png,C20041006138300152.png"\; state="\{\{state\}\}">b</figure-callout>}}_0+b_1{\mathrm{<figure-callout\; id="2"\; label="r"\; filenames="C20041006138300142.png"\; state="\{\{state\}\}">r</figure-callout>}}_2+b_2{s}_2\\ {\mathrm{<figure-callout\; id="3"\; label="v"\; filenames="C20041006138300142.png,C20041006138300152.png"\; state="\{\{state\}\}">v</figure-callout>}}_3={\mathrm{<figure-callout\; id="0"\; label="b"\; filenames="C20041006138300142.png,C20041006138300152.png"\; state="\{\{state\}\}">b</figure-callout>}}_0+b_1{\mathrm{<figure-callout\; id="3"\; label="r"\; filenames="C20041006138300142.png,C20041006138300152.png"\; state="\{\{state\}\}">r</figure-callout>}}_3+b_2{s}_3\end{array}\left[\begin{array}{c}{v}_1\\ v_2\\ {\mathrm{<figure-callout\; id="3"\; label="v"\; filenames="C20041006138300142.png,C20041006138300152.png"\; state="\{\{state\}\}">v</figure-callout>}}_3\end{array}\right]=\left[\begin{array}{ccc}1& r_1& {\mathrm{<figure-callout\; id="1"\; label="s"\; filenames="C20041006138300152.png"\; state="\{\{state\}\}">s</figure-callout>}}_1\\ 1& r_2& {\mathrm{<figure-callout\; id="2"\; label="s"\; filenames="C20041006138300142.png"\; state="\{\{state\}\}">s</figure-callout>}}_2\\ 1& r_3& s_3\end{array}\right]\left[\begin{array}{c}{b}_0\\ b_1\\ {\mathrm{<figure-callout\; id="2"\; label="b"\; filenames="C20041006138300142.png"\; state="\{\{state\}\}">b</figure-callout>}}_2\end{array}\right]$

But simultaneous solves a ₀, a ₁, a ₂, b ₀, b ₁, b ₂6 coefficients are established h ₁(x, y) and h ₂(x y), uses following bilinear interpolation method then, carries out the geometry correction operation.

3. the geometry correction CALCULATION OF PARAMETERS of projective transformation:

If find out four points by reference map, corresponding one by one with last four point coordinate of distortion figure, then

x′＝a ₀+a ₁x+a ₂y+a ₃xy

y′＝b ₀+b ₁x+b ₂y+b ₃xy

8 unknown quantitys are arranged, need 4 corresponding known coordinate points

(x ₁，y ₁)(x ₂，y ₂)(x ₃，y ₃)(x ₄，y ₄)

(x ' ₁, y ' ₁) (x ' ₂, y ' ₂) (x ' ₃, y ' ₃) (x ' ₄, y ' ₄) matrix form:

$\left[\begin{array}{c}{x_1}^{\&prime;}\\ {x_2}^{\&prime;}\\ {x_3}^{\&prime;}\\ {x_4}^{\&prime;}\end{array}\right]=\left[\begin{array}{cccc}1& x_1& y_1& x_1{y}_1\\ 1& x_2& y_2& x_2{y}_2\\ 1& x_3& y_3& x_3{y}_3\\ 1& x_4& y_4& x_4{y}_4\end{array}\right]\left[\begin{array}{c}{a}_0\\ a_1\\ a_2\\ a_3\end{array}\right]$

But simultaneous solves a ₀, a ₁, a ₂, a ₃, b ₀, b ₁, _B2, b ₃8 coefficients are established h ₁(x, y) and h ₂(x y), uses following bilinear interpolation method then, carries out the geometry correction operation.

4. the geometry correction CALCULATION OF PARAMETERS of quadratic form:

If find out six points by reference map, corresponding one by one with last six point coordinate of distortion figure, then

x′＝a ₀+a ₁x+a ₂y+a ₃x ²+a ₄xy+a ₅y ²

y′＝b ₀+b ₁x+b ₂y+b ₃x ²+b ₄xy+b ₅y ²

12 unknown quantitys are arranged, need 6 corresponding known coordinate points

(x ₁，y ₁)(x ₂，y ₂)(x ₃，y ₃)(x ₄，y ₄)(x ₅，y ₅)(x ₆，y ₆)

(x ' ₁, y ' ₁) (x ' ₂, y ' ₂) (x ' ₃, y ' ₃) (x ' ₄, y ' ₄) (x ' ₅, y ' ₅) (x ' ₆, y ' ₆) matrix form:

$\left[\begin{array}{c}{x}_1^{\&prime;}\\ x_2^{\&prime;}\\ x_3^{\&prime;}\\ x_4^{\&prime;}\\ x_5^{\&prime;}\\ x_6^{\&prime;}\end{array}\right]=\left[\begin{array}{cccccc}1& x_1& y_1& {x_1}^2& x_1{y}_1& {y_1}^2\\ 1& x_2& y_2& {x_2}^2& x_2{y}_2& {y_2}^2\\ 1& x_3& y_3& {x_3}^2& x_3{y}_3& {y_3}^2\\ 1& x_4& y_4& {x_4}^2& x_4{y}_4& {y_4}^2\\ 1& x_5& y_5& {x_5}^2& x_5{y}_5& {y_5}^2\\ 1& x_6& y_6& {x_6}^2& x_6{y}_6& {y_6}^2\end{array}\right]\left[\begin{array}{c}{a}_0\\ a_1\\ a_2\\ a_3\\ a_4\\ a_5\end{array}\right]$

But simultaneous solves a ₀, a ₁, a ₂, a ₃, a ₄, a ₅, b ₀, b ₁, b ₂, b ₃, b ₄, b ₅12 coefficients are established h ₁(x, y) and h ₂(x y), uses following bilinear interpolation method then, carries out the geometry correction operation.

5. the bilinear interpolation of geometry correction:

If Basic Reference Coordinate System is an isometric net, from this grid epigraph point (x ₀, y ₀) set out, calculate on known distortion coordinate diagram picture a bit (α, β)

(x′ ₁，y′ ₁) (x′ ₁+1，y′ ₁)

(α，β)

(x′ ₁，y′ ₁+1) (x′ ₁+1，y′ ₁+1)

If (α β) drops in the middle of above-mentioned four points.

If

x ₁′≤α≤x′ ₁+1 α′≤α-x′ ₁

y′ ₁≤β≤y′ ₁+1 β′≤β-y′ ₁

Then the infrared temperature value is on the return contact

f(x ₀，y ₀)＝(1-α′)(1-β′)g(x′ ₁，y′ ₁)+α′(1-β′)g(x′ ₁+1，y′ ₁)

+(1-α′)β′g(x′ ₁，y′ ₁+1)+α′β′g(x′ ₁+1，y′+1)

6. the acquisition of temperature inversion curve:

First width of cloth in the image workspace of GUI is a reference picture, other thermal maps are sample image, on the sample image that reference picture and needs are proofreaied and correct, import anchor point and (import 3 successively respectively according to distinct methods such as affine, projection, quadratic forms, 4,6 anchor points) after the execution geometry correction, the position, space possesses comparability on the image, and the thermal map spatial information after the correction is from reference picture, temperature information has promptly been finished the registration of image from uncorrected original sample figure.

When on the image of user behind registration during rolling mouse, software carries out thermometric to the same spatial location of every width of cloth thermal map of Study document folder successively, obtaining a transverse axis is the temperature inversion curve map of temperature for the time longitudinal axis, and this curve map has reflected the temperature variation of this equipment of different times in this locus.

So far can see that the thermal map that this equipment different times is taken is through registration, the locus unanimity, then the user can with custom, thermal imaging system distinctive " face scanning " mode obtains the temperature inversion curve map of this equipment point.

The groundwork amount of considering this analysis method is the process of manual input anchor point, this method is only just carried out geometry correction to the image of having imported anchor point, like this to corrected image, the user needn't repeat to import anchor point, only need get final product, avoid repeating work reference picture and new image input anchor point that gather, that do not have registration.

7. import to Excel and generate Avi:

After the user imports point of interest, software is saved in information such as the temperature data of each point of interest and filename, thermometric shelves, emissivity, environment temperature, load in the text of Excel compatible format, can use " importing external data " function with this data importing in Excel.The user can further analyze in Excel like this, for example: obtain difference variation curve of two points of interest or the like.

The user can also and analyze data view data and output in the Avi file.The implementation method of this function is, to the every width of cloth thermal map in the Study document folder, carry out pseudo-colours according to temperature scale earlier and strengthen the irrelevant bitmap of forming device, interest point annotation position on this a series of bitmap, and at new bitmaps of information acquisition such as bitmap white space demonstration point of interest temperature, environment temperature, emissivity, bitmaps this is a series of at the using system funcall generate an Avi file.

The purpose of doing like this is not only more intuitive display analysis result, the interchange of more convenient infrared fault diagnosis knowledge and sharing, for example: higher leveled infrared fault diagnosis research institution can become animation file to the long-term observation result's of some exemplary apparatus preservation, the infrared failed operation mechanism of low one-level can obtain these data by Internet, and and oneself analysis result compare, better carry out equipment fault diagnosis work.

Fig. 4 is that powered-down served as an infrared sequence image of taking, do not do geometry correction at interval continuously with 5 seconds after daylight lamp steady operation a period of time.On same spatial location, the 1st, point of interest is in daylight lamp electrode place among 2 width of cloth figure, but in the figure of back, point of interest is obviously outside electrode, take the temperature at electrode place and the temperature of background to compare like this, can see that from the temperature variation of Fig. 5 this temperature variation has no rule and can say without any value.

Fig. 6 is the sequence image behind the image rectification shown in Figure 4.As can be seen from Figure 6, all points of interest all are in the hot spot region, at the Changing Pattern that the negative exponent of daylight lamp electrode place temperature descends behind the powered-down as can be seen on the temperature changing curve diagram of Fig. 7.

Claims (2)

1. an infrared sequence image historical trend analytical approach comprises the steps: in regular turn

(1) thermal-induced imagery that the different times of equipment to be analyzed is taken is put in order in the Study document folder;

(2) user imports anchor point by hand, utilize affine transformation matrix, projective transformation matrix or quadratic form transformation matrix, use bilinear interpolation method then, carry out the geometry correction operation, to the above-mentioned geometric correction of imagery, with the locus registration of these images according to reference picture;

(3) utilize the characteristics of thermal imaging system " face scanning ", on image, search for, obtain each point of interest position and temperature, draw and show that this equipment is at this point of interest different times variation of temperature curve map.

2. infrared sequence image historical trend analytical approach as claimed in claim 1 is characterized in that: the temperature variation curve data importing of described point of interest is further analyzed in the Excel software and reported; To comprise that all of proofreading and correct back image, point of interest position and temperature analyze data and be saved in the video file, and make things convenient for the user to analyze the interchange of data and share.

Images