CN102052899B - Take trichromatic contrast ratio as method and the device of characteristic frame Matched measurement two-dimension displacement - Google Patents
Take trichromatic contrast ratio as method and the device of characteristic frame Matched measurement two-dimension displacement Download PDFInfo
- Publication number
- CN102052899B CN102052899B CN200910191318.1A CN200910191318A CN102052899B CN 102052899 B CN102052899 B CN 102052899B CN 200910191318 A CN200910191318 A CN 200910191318A CN 102052899 B CN102052899 B CN 102052899B
- Authority
- CN
- China
- Prior art keywords
- frame
- pixel
- primary colours
- comparison window
- reference frame
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Expired - Fee Related
Links
Landscapes
- Image Analysis (AREA)
Abstract
Take trichromatic contrast ratio as method and the device of characteristic frame Matched measurement two-dimension displacement, the computing machine logical by a Daepori and camera thereof form, it resolves into three primary colours frames picture frame, get the feature of edge direction data as this primary colours frame of each primary colours frame, by the self-correlation of the edge direction data of computed image brightness, check the quality details of measured object reflecting surface, choose suitable comparison window pel array, then frame-frame cross correlation the matching primitives of primary colours frame edge direction data is carried out with sampling frame, get its cross correlation coefficient the maximum as optimum matching person, obtain the displacement that each primary colours frame occurs, get their mean value as measurement result, and adjust the position of comparison window accordingly or upgrade reference frame, the scale of adjustment cross correlation operator array, reduce calculated amount, the ambient lighting overcome beyond three primary colours wavelength changes the impact on measuring, ensure that measuring accuracy.
Description
Technical field
The present invention relates to digital picture field of measuring technique, particularly adopt computing machine camera to measure method and the device thereof of the two-dimentional micro-displacement of object.
Background technology
In order to play the function of the photosensor arrays of computing machine camera, " taking contrast as method and the device of characteristic frame Matched measurement two-dimension displacement " (application for a patent for invention number: 200910190926.0) propose a kind of extraction picture contrast feature, frame-frame association matching technique is adopted to measure method and the device of object micro-displacement, adaptating lighting situation can there is change to a certain degree in it, obtain the two-dimension displacement vector velocity that object is small.But, this application for the brightness of just picture frame, not yet have the function making full use of photosensor arrays.
Summary of the invention
In order to play the function of the photosensor arrays of computing machine camera further, the invention provides a kind of take trichromatic contrast ratio as method and the device of characteristic frame Matched measurement two-dimension displacement, it with computing machine camera for photoelectric conversion sensor, can occur in the environment of certain change at illuminating position, measure the small two-dimension displacement vector velocity of object in the plane perpendicular with the optical axis of camera.
The technical solution adopted for the present invention to solve the technical problems is: camera is connected to the logical computing machine of a Daepori by USB interface, and this allocation of computer has USB interface, internal memory, CPU, hard disk, display card and display, keyboard and mouse, operating system, webcam driver program and camera to take and three primary colours edge direction Frame matching displacement measurement program; It take trichromatic contrast ratio as the method that feature implements the small two-dimension displacement of frame Matched measurement object that this program embodies the present invention, comprising:
Step one, with the form of bitmap (M × N, M, N ∈ positive integer), take the image of a frame testee, as with reference to frame;
With the position of first pixel in its upper left corner of pel array of described photographed frame for initial point, horizontal right direction is x-axis direction, and direction is vertically downward y-axis direction;
Choose a region at the middle section of described pel array, size is m
0× n
0, m
0, n
0∈ positive integer, is referred to as comparison window, and described comparison window apart from the horizontal direction of described pel array and each h and v of the edge pixel of a vertical direction pixel, namely has: m respectively
0+ 2h=M, n
0+ 2v=N, h, v ∈ positive integer;
Step 2, pel array for above-mentioned reference frame, derive the edge direction data of all pixel intensity along X-direction and red, green and blue three kinds of primary colours line by line, respectively with the binary numeral 001 of 3bit, 010 and 100 represent positive limit, marginal and the 3rd class limit, be designated as: { reference (x, y) }, { reference
red(x, y) }, { reference
green(x, y) } and { reference
blue(x, y) }, preserve it;
Step 3, calculate the auto correlation matching factor about pixel intensity of the pel array of comparison window in described reference frame:
In formula, x and y is the coordinate of pixel in comparison window respectively, and sign of operation represents binary logic and computing, its operation result or be logical zero or for logical one, bracket " [] " represents the numerical value that the logical function got wherein is corresponding, or is numerical value 0, or be numerical value 1, individual element calculates, and cumulative result is as this self-correlation auto_correlation (a, b), 3 × 3 association coupling operator: a=-1 are got, 0,1, b=-1,0,1, therefore, raw 9 self-correlations of common property;
Step 4, analyze and be observed the fine degree of the architectural feature of body surface, that is analyze the degree of can distinguishing of the light and shade contrast between most neighborhood pixels, check:
auto_correlation(a，b)≥auto_correlation(0，0)×similarity
In formula, similarity describes comparison window and the similarity degree of the pel array of its contiguous identical scale, such as, get similarity=60%, can pre-set, also can carry out debugging and selecting according to the quality on light conditions and measured object surface;
If the auto correlation matching factor meeting above-mentioned inspection inequality is more than 3, the capable and step row of each step of scope needing progressively to expand comparison window: make m=m
0+ step, n=n
0+ step, recalculate the auto correlation matching factor of new comparison window, and again carry out neighborhood pixels analyze distinguish degree, until no more than 3 of the self-correlation meeting above-mentioned analysis inequality, illustrate that the architectural feature on the surface of subject is enough meticulous, value between most neighborhood pixels can be distinguished, think at this moment have found to may be used for carrying out under current surface situation of object and illuminating position matching ratio compared with best comparison window pel array: m=m
0+ step, n=n
0+ step, 2h=M-m, 2v=N-n, wherein, step is stepped parameter, and initial value is 1, needs the scale expanding comparison window just to increase by 1 at every turn; If to exceed in frame a predetermined scope, also do not find suitable comparison window, then think that this part reflecting surface of this object is unsuitable for the surveying work of this device, and provide prompting warning;
After step 5, above-mentioned shooting, through Δ t after a while, take the second framing bit figure, as sampling frame;
Determine all pixel intensity of pixel along X-direction and the edge direction data of red, green and blue three kinds of primary colours in this sampling frame line by line, respectively with the binary numeral 001 of 3bit, 010 and 100 represent positive limit, marginal and the 3rd class limit, be designated as: (comparison (x, y) }, { comparison
red(x, y) }, { comparison
green(x, y) } and { comparison
blue(x, y) }, preserve it;
Step 6, the edge direction data { reference (x of the pel array in comparison window in described reference frame, y) } ({ comparison (x in described sampling frame scope, y) }) carry out cross correlation coupling according to 9 × 9 association coupling arrays, specific algorithm is:
For a=-4 ,-3 ,-2 ,-1,0 ,+1 ,+2 ,+3, + 4 and b=-4 ,-3 ,-2 ,-1,0 ,+1 ,+2 ,+3, the combination of+4, the comparison window of described reference frame may have 81 kinds of situation of movement, therefore, the picture frame of corresponding three kinds of primary colours, raw 9 × 9 × 3=243 cross correlation matching factor cross_correlation (a, b) of common property;
Step 7, the cross correlation coefficient that frame-frame correlation degree is the highest are maximum:
cross_correlation(a，b)→auto_correlation(0，0)
Therefore the picture frame of three kinds of primary colours corresponding to described sampling frame is obtained relative to the described direction of reference frame movement and the amplitude of movement:
Δ x
red(i, j)=a
1, Δ y
red(i, j)=b
1
Δ x
green(i, j)=a
2, Δ y
green(i, j)=b
2
Δ x
blue(i, j)=a
3, Δ y
blue(i, j)=b
3
Wherein, i represents the sequential counting measuring shooting, and j represents the sequential counting of got reference frame;
The relative displacement vector that in this sample period, object occurs is:
In measuring process, the total relative displacement vector of object is:
Δ x
0(i, j)=Δ x
0(i-1, j)+Δ x (i, j), Δ y
0(i, j)=Δ y
0(i-1, j)+Δ y (i, j), (the Δ x on the right in equation
0(i-1, j), Δ y
0(i-1, j)) be the relative displacement vector that object was accumulated in the past;
The velocity of step 8, calculating ohject displacement:
Δv
x(i，j)＝Δx(i，j)/Δt，Δv
y(i，j)＝Δy(i，j)/Δt；
If step 9 | Δ x
0(i, j)-Δ x
0(k, j-1) |>=2m/5, or | Δ y
0(i, j)-Δ y
0(k, j-1) |>=2n/5, wherein, k=max (i) | (j-1) represents the Order Count value of the last shooting when reference frame sequential counting is j-1, namely, under the condition do not changed at described reference frame, the accumulation of the relative displacement that comparison window wherein occurs has exceeded 2/5 of the amplitude of this comparison window, at this moment, replace described reference frame with up-to-date sampling frame, its comparison window is repositioned at the central part of new reference frame;
If | Δ x
0(i, j)-Δ x
0(k, j-1) | < 2m/5 and | Δ y
0(i, j)-Δ y
0(k, j-1) | < 2n/5, does not upgrade described reference frame, but the comparison window generation relative displacement in described reference frame:
If step 10 have updated reference frame, imitative step 3 calculates the auto correlation matching factor of described new reference frame, and imitative step 4 watches surface texture featur, again determines the size m × n of its best comparison window;
If do not upgrade reference frame, the size of the coupling of cross correlation described in set-up procedure six operator array:
If
And
Changing to take is 7 × 7:a=-3 ,-2 ,-1,0 ,+1 ,+2 ,+3, b=-3 ,-2 ,-1,0 ,+1 ,+2 ,+3,
If
And
Changing to take is 5 × 5:a=-2 ,-1,0 ,+1 ,+2, b=-2 ,-1,0 ,+1 ,+2, otherwise be still taken as 9 × 9 association coupling operator array;
After step 11, above-mentioned shooting, through Δ t after a while, shooting the 3rd framing bit figure, as newly sampling frame;
Determine all pixel intensity of pixel along X-direction and the edge direction data of red, green and blue three kinds of primary colours in this sampling frame line by line, respectively with the binary numeral 001 of 3bit, 010 and 100 represent positive limit, marginal and the 3rd class limit, be designated as: { comparison (x, y) }, { comparison
red(x, y) }, { comparison
green(x, y) } and { comparison
blue(x, y) }, preserve it;
Step 12, according to determine in step 10 cross correlation coupling operator array, the pel array in comparison window in described reference frame is carried out three primary colours cross correlation matching primitives in described sampling frame scope, the same step 6 of specific algorithm;
Step 13, jump to step 7, continue to measure.
In above-mentioned steps two, five and 11 about the definition of " along all pixel intensity of X-direction and the edge direction data of red, green and blue three kinds of primary colours " be:
With function I (X, Y) brightness value of all pixels is represented, or represent the brightness value of certain wavelength (red, green or blue) of all pixels, wherein, (X, Y) coordinate of this pixel is represented, if the brightness value of a pixel is than a brightness value also little error margin value error of the pixel of second after it, if namely
I(X，Y)＜I(X+2，Y)-error
Then define between these two pixels and there is the brightness of this pixel or the positive limit corresponding to these primary colours of this pixel;
If the brightness value of a pixel is than a brightness value also large error margin value error of the pixel of second after it, if i.e. I (X, Y) > I (X+2, Y)+error
Then define that to exist corresponding to the brightness of this pixel or these primary colours of this pixel between these two pixels marginal;
If the brightness value of the pixel brightness value corresponding to after it second pixel is close, its difference is no more than an error margin value error, if i.e. I (X+2, Y)-error < I (X, Y) < I (X+2, Y)+error
Then think there is not the brightness of this pixel or " limit " corresponding to these primary colours of this pixel between these two pixels, or be referred to as the 3rd class limit;
Error margin value in above formula can according to concrete light conditions, is predisposed to a little numerical value, such as: error=10; The limit of acquisition like this is positioned at the position of first pixel after this pixel, is also namely positioned in that pixel (X, Y) in the centre position participating in two pixels compared;
Along X-direction, positive limit corresponding to certain primary colours of the brightness of each pixel or each pixel, marginal and the 3rd class limit are respectively with the binary numeral 001 of 3bit, 010 and 100 represent, their set accordingly form the brightness of pixel or the edge direction data corresponding to these primary colours of pixel in this direction, be designated as set function { fram λ (x, y) }, subscript wherein represents wavelength (i.e. color).
In actual measurement process, can also implement further to measure calibration, take this to obtain direct measurement result.
The invention has the beneficial effects as follows, it is based on tristimulus image frame, extract the contrast metric of edge direction data as each primary colour picture picture frame of each primary colours respectively, then respectively contrast frame-frame coupling is implemented to each primary colour picture picture frame, thus obtain separately corresponding relative displacement, the mean value getting them as measurement result, than image intensity contrast's degree feature, effectively overcome other wavelength in ambient lighting change beyond three primary colours to the impact of image pixel, ensure that measuring accuracy.
Accompanying drawing explanation
Fig. 1 is measurement mechanism block scheme of the present invention.
Fig. 2 is the schematic diagram of the pel array produced after photoelectric sensor chip carries out opto-electronic conversion.
In Fig. 1,1. computing machine camera, 2. optical lens, 3. photoelectric sensor chip, 4.USB interface, 5. computer system, 6.USB interface, 7.CPU, 8. internal memory, 9. display card and display, 10. hard disk, 11. keyboards and mouse, 12. operating systems, 13. webcam driver programs, 14. camera shooting and three primary colours edge direction Frame matching displacement measurement programs, 15. light fixture.
In Fig. 2,20. 1 frame pel arrays, comparison window in 21. reference frames, association matching area illustrated embodiment in 22. sampling frames, the contingent extreme position of comparison window in 23. reference frames.
Embodiment
Fig. 1 is the block scheme of measurement mechanism of the present invention.First, above run the webcam driver program (13) of rationing with camera (1) in computer system (5), be connected camera (1) to computing machine (5) by USB interface (4) with (6).Then, camera focal imaging object being measured is allowed.Select measurement environment and select light fixture (15), preferably eliminating obvious jamming light source, avoid the light and shade contrast of picture, in measuring process, obvious change does not occur.Then, run camera shooting and three primary colours edge direction Frame matching displacement measurement program (14), implement to measure in real time.It is the method for characteristic frame Matched measurement two-dimension displacement that this program (14) embodies with trichromatic contrast ratio, specifically sees described by " summary of the invention ", selects its general condensed summary below as follows.
Fig. 2 represents the corresponding pel array (20) produced after photoelectric sensor chip (3) carries out opto-electronic conversion.The size of this pel array (20) is M × N, M, N ∈ integer, correspond to bitmap format, wherein each lattice and pixel, and its brightness number is 0-255.In diagram coordinate system, each lattice and pixel orientate a pair coordinate points (x, y) as, 0≤x≤M, 0≤y≤N, x, y ∈ integer.Middle section has in fig. 2 opened up a comparison window (21), size is taken as m × n, m, n ∈ integer, comparison window (21) and the horizontal direction of picture element matrix (20) and an edge pixel distance h and v pixel respectively of vertical direction, namely have: m+2h=M, n+2v=N, h, v ∈ positive integer.Therefore, the coordinate of that first pixel grid of the upper left corner of comparison window is: x=h+1, y=v+1.
Choose m, n is relevant with the fine degree of the architectural feature in reflections off objects face, is related to (coupling) measuring accuracy, decides amount of calculation, affect the speed of response of measurement mechanism.According to resolution and the frame per second selecting index initial value of concrete camera, such as, choose initial value: m
0=80, n
0=80, then, by the auto correlation matching factor of pixel intensity contrast in comparison window in computing reference frame, find to may be used for carrying out under current surface situation of object and illuminating position matching ratio compared with best comparison window pel array, this is namely described in step 3, four.On the one hand, whether this is by analyzing the self-correlation of each self-correlation and comparison window itself close to judging, on the other hand, the parameter s imilarity that can be pre-set by is characterized.This parameter describe comparison window and the similarity degree of the pel array of its contiguous identical scale, can carry out debugging and selecting according to the quality on light conditions and measured object surface.Comprehensive is following analysis inequality:
auto_correlation(a，b)≥auto_correlation(0，0)×similarity
But " meeting the number of the self-correlation of above-mentioned analysis inequality " also not necessarily have to equal 3, can adjust according to actual conditions.
After object is subjected to displacement, in reference frame comparison window be likely moved to sampling frame in somewhere, such as sample association matching area illustrated embodiment (22) place in frame, but, do not allow it to move and exceed the contingent extreme position of comparison window in reference frame (23).
Described in step 2, five and 11, edge direction data reflect light and shade contrast's feature on testee surface, application for a patent for invention " peak valley motion detection method of measurement Displacement and the device " (application number: 200910190924.1) give detailed description to this submitted to recently.
Basic thought in step 6, ten and 12 is, first adopts the search of fairly large cross correlation coupling operator matrix and the region of comparing frame in reference frame and matching most, takes this displacement determining that frame of pixels occurs.Then, according to obtained displacement size, the scale of adjustment cross correlation operator matrix, in the hope of reducing calculated amount.The scale of cross correlation operator matrix is greater than contingent displacement range.
Teach the method for the position of comparison window in mobile reference frame in step 9 or upgrade the condition of reference frame, its objective is to ensure that in described reference frame, comparison window has considerable overlapping region with respective associated region in sampling frame, the measuring accuracy being less than a pixel unit can be reflected, and can not after repeatedly taking measurement cumulative measurement error." use computing machine camera measures method and the device of small two-dimension displacement " (application for a patent for invention number: 2009101042778) this has been done to detailed analysis.Wherein said " under the condition that described reference frame does not change; the accumulation of the relative displacement that comparison window wherein occurs has exceeded 2/5 of the amplitude of this comparison window; at this moment; replace described reference frame with up-to-date sampling frame ", so-called super 2/5 this numerical value, specifically can adjust according to concrete conditions such as the scale of comparison window, tested movement velocitys.
Claims (1)
1. one kind take trichromatic contrast ratio as the method for characteristic frame Matched measurement two-dimension displacement, it provide the method for the computing machine cooperation camera measurement two-dimension displacement that a kind of use one Daepori leads to, this camera is connected to computing machine by the USB interface of camera, it is characterized in that: the method is taken by camera and adopted frame matching process to measure two-dimension displacement according to the trichromatic contrast ratio of image, comprises the following steps:
Step one, take the image of a frame testee with the form of bitmap M × N, M, N ∈ positive integer, is called reference frame; For reference frame and follow-up photographed frame choose planimetric rectangular coordinates axle system;
Choose a region at the middle section of the pel array of reference frame, size is m
0× n
0, m
0, n
0∈ positive integer, is referred to as comparison window, and described comparison window apart from the horizontal direction of described pel array and each h and v of the edge pixel of a vertical direction pixel, namely has: m respectively
0+ 2h=M, n
0+ 2v=N, h, v ∈ positive integer;
Step 2, pel array for above-mentioned reference frame, derive the edge direction data of all pixel intensity along X-direction and red, green and blue three kinds of primary colours line by line, be specifically expressed as: { reference (x, y) }, { reference
red(x, y) }, { reference
green(x, y) } and { reference
blue(x, y) }, wherein, represent positive limit, marginal and the 3rd class limit with the binary numeral 001,010 and 100 of 3bit respectively; Preserve this four groups of edge direction data;
Step 3, calculate the auto correlation matching factor about pixel intensity of the pel array of comparison window in described reference frame:
In formula, x and y is the coordinate of pixel in comparison window respectively, sign of operation represents binary logic and computing, its operation result or be logical zero or for logical one, bracket " [] " represents the numerical value that the logical function got wherein is corresponding, or is numerical value 0, or be numerical value 1, individual element calculates, and cumulative result is as this auto correlation matching factor auto_correlation (a, b), get 3 × 3 association coupling operator: a=-1,0,1, b=-1,0,1, therefore, raw 9 the auto correlation matching factors of common property;
Step 4, analyze and be observed the fine degree of the architectural feature of body surface, that is analyze the degree of can distinguishing of the light and shade contrast between most neighborhood pixels array, check:
auto_correlation(a，b)≥auto_correlation(0，0)×similarity
In formula, similarity describes comparison window is close to the pel array of identical scale similarity degree with it, and similarity is set up in advance, and the quality according to light conditions and measured object surface carries out debugging and selecting;
If the auto correlation matching factor meeting above-mentioned inequality is more than 3, the capable and step row of each step of scope needing progressively to expand comparison window: make m=m
0+ step, n=n
0+ step, recalculate the auto correlation matching factor of new comparison window, and again carry out neighborhood pixels array analyze distinguish degree, until no more than 3 of the auto correlation matching factor meeting above-mentioned inequality, illustrate that the architectural feature on the surface of subject is enough meticulous, value between most neighborhood pixels array can be distinguished, think at this moment have found to may be used for carrying out under current surface situation of object and illuminating position matching ratio compared with best comparison window pel array: m=m
0+ step, n=n
0+ step, 2h=M-m, 2v=N-n, wherein, step is stepped parameter, and initial value is 1, needs the scale expanding comparison window just to increase by 1 at every turn; If to exceed in frame a predetermined scope, also do not find suitable comparison window, then think that this part reflecting surface of this object is unsuitable for this surveying work, and provide prompting warning;
After step 5, above-mentioned shooting, through Δ t after a while, take the second framing bit figure, as sampling frame;
Determine all pixel intensity of pixel along X-direction and the edge direction data of red, green and blue three kinds of primary colours in this sampling frame line by line, be specifically expressed as: { comparison (x, y) }, { comparison
red(x, y) }, { comparison
green(x, y) } and { comparison
blue(x, y) }, wherein, represent positive limit, marginal and the 3rd class limit with the binary numeral 001,010 and 100 of 3bit respectively, preserve this four groups of edge direction data;
Step 6, the edge direction data of the pel array in comparison window in described reference frame in described sampling frame scope according to 9 × 9 association coupling operators carry out cross correlation coupling, specific algorithm is:
For 9 × 9 association coupling operators, have: a=-4 ,-3 ,-2,-1,0 ,+1 ,+2, + 3 ,+4, b=-4 ,-3,-2 ,-1,0 ,+1, + 2 ,+3 ,+4, the comparison window of described reference frame has 81 kinds of situation of movement, therefore, the picture frame of corresponding three kinds of primary colours, raw 9 × 9 × 3=243 cross correlation matching factor cross_correlation (a, b) of common property;
Step 7, the cross correlation coefficient that frame-frame correlation degree is the highest are maximum:
cross_correlation(a，b)→auto_correlation(0，0)
Therefore the picture frame of three kinds of primary colours corresponding to described sampling frame is obtained relative to the direction of described reference frame movement and amplitude thereof:
Δ x
red(i, j)=a
1, Δ y
red(i, j)=b
1
Δ x
green(i, j)=a
2, Δ y
green(i, j)=b
2
Δ x
blue(i, j)=a
3, Δ y
blue(i, j)=b
3
Wherein, i represents the sequential counting taken in measuring process, and j represents the sequential counting of got reference frame, and a, b correspond to corresponding association coupling operator, and the subscript of a, b is for distinguishing each self-corresponding primary colours;
The relative displacement vector that in this sample period, object occurs is:
In measuring process, the total relative displacement vector of object is:
Δx
0(i，j)＝Δx
0(i-1，j)+Δx(i，j)，Δy
0(i，j)＝Δy
0(i-1，j)+Δy(i，j)，
(the Δ x on the right in equation
0(i-1, j), Δ y
0(i-1, j)) be the relative displacement vector that object was accumulated in the past;
The velocity of step 8, calculating ohject displacement:
Δv
x(i，j)＝Δx(i，j)/Δt，Δv
y(i，j)＝Δy(i，j)/Δt；
If step 9 | Δ x
0(i, j)-Δ x
0(k, j-1) |>=2m/5, or | Δ y
0(i, j)-Δ y
0(k, j-1) |>=2n/5, wherein, k=max (i) | (j-1) represents the Order Count value of the last shooting when reference frame sequential counting is j-1, namely, under the condition do not changed at described reference frame, the accumulation of the relative displacement that comparison window wherein occurs has exceeded 2/5 of the amplitude of this comparison window, at this moment, replace described reference frame with up-to-date sampling frame, its comparison window is repositioned at the central part of new reference frame;
If | Δ x
0(i, j)-Δ x
0(k, j-1) | < 2m/5 and | Δ y
0(i, j)-Δ y
0(k, j-1) | < 2n/5, does not upgrade described reference frame, but the comparison window generation relative displacement in described reference frame:
If step 10 have updated reference frame, calculate the auto correlation matching factor of described new reference frame according to the pel array of comparison window in computing reference frame described in step 3 about the method for the auto correlation matching factor of pixel intensity; Watching surface texture featur according to analyzing the method being observed the fine degree of the architectural feature of body surface described in step 4, again determining the size m × n of its best comparison window;
If do not upgrade reference frame, described in set-up procedure six, carry out the association coupling operator in the specific algorithm of cross correlation coupling:
If
And
Changing to take is 7 × 7 association coupling operator: a=-3 ,-2 ,-1,0 ,+1 ,+2 ,+3, b=-3 ,-2 ,-1,0 ,+1 ,+2 ,+3, if
and
changing to take is 5 × 5 association coupling operator: a=-2 ,-1,0 ,+1 ,+2, b=-2 ,-1,0 ,+1 ,+2,
Otherwise be still taken as 9 × 9 association coupling operator array;
After step 11, above-mentioned shooting, through Δ t after a while, shooting the 3rd framing bit figure, as newly sampling frame;
Determine all pixel intensity of pixel along X-direction and the edge direction data of red, green and blue three kinds of primary colours in this sampling frame line by line, be specifically expressed as: { comparison (x, y) }, { comparison
red(x, y) }, { comparison
green(x, y) } and { comparison
blue(x, y) }, wherein, represent positive limit, marginal and the 3rd class limit with the binary numeral 001,010 and 100 of 3bit respectively, preserve this four groups of edge direction data;
Step 12, according to determine in step 10 carry out cross correlation coupling association coupling operator, the pel array in comparison window in described reference frame is carried out three primary colours cross correlation matching primitives in described sampling frame scope, the same step 6 of specific algorithm;
Step 13, jump to step 7, continue to measure;
In above-mentioned steps two, five and 11 about the definition of " along all pixel intensity of X-direction and the edge direction data of red, green and blue three kinds of primary colours " be:
For a kind of primary colours, namely one of red, green or blue, represent with function I (X, Y) brightness value that all pixels are corresponding, wherein, (X, Y) represents the coordinate of this pixel; If the brightness value of a pixel is than a brightness value also little error margin value error of the pixel of second after it, if namely
I(X，Y)＜I(X+2，Y)-error
Then define the positive limit that the described primary colours of existence this pixel between these two pixels are corresponding;
If the brightness value of a pixel is than a brightness value also large error margin value error of the pixel of second after it, if i.e. I (X, Y) > I (X+2, Y)+error
Then define and between these two pixels, to there is corresponding marginal of the described primary colours of this pixel;
If the brightness value of the pixel brightness value corresponding to after it second pixel is close, its difference is no more than an error margin value error, if i.e. I (X+2, Y)-error < I (X, Y) < I (X+2, Y)+error
Then think on " limit " that the described primary colours that there is not this pixel between these two pixels are corresponding, or be referred to as the 3rd class limit;
Be a little numerical value according to the preset above-mentioned error margin value of concrete light conditions; The limit of acquisition like this is positioned at the position of first pixel after this pixel, is also namely positioned in that pixel (X+1, Y) in the centre position participating in two pixels compared;
Along X-direction, for certain primary colours, the positive limit corresponding to each pixel, marginal and the 3rd class limit represent with the binary numeral 001,010 and 100 of 3bit respectively, their set forms the edge direction data corresponding to corresponding primary colours of the pixel in this direction, is designated as set function { fram
λ(x, y) }, subscript wherein represents corresponding primary colours; For the brightness of all pixels, represent with function I (X, Y) brightness value that all pixels are corresponding, wherein, (X, Y) represents the coordinate of this pixel, similar above-mentioned definition, along X-direction, obtains the edge direction data corresponding to brightness of the pixel in this direction.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN200910191318.1A CN102052899B (en) | 2009-11-02 | 2009-11-02 | Take trichromatic contrast ratio as method and the device of characteristic frame Matched measurement two-dimension displacement |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN200910191318.1A CN102052899B (en) | 2009-11-02 | 2009-11-02 | Take trichromatic contrast ratio as method and the device of characteristic frame Matched measurement two-dimension displacement |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN102052899A CN102052899A (en) | 2011-05-11 |
| CN102052899B true CN102052899B (en) | 2015-09-30 |
Family
ID=43957467
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN200910191318.1A Expired - Fee Related CN102052899B (en) | 2009-11-02 | 2009-11-02 | Take trichromatic contrast ratio as method and the device of characteristic frame Matched measurement two-dimension displacement |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN102052899B (en) |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1198217A (en) * | 1996-05-31 | 1998-11-04 | 埃勒夫勘探产品公司 | Method for automatically determining stratification beds in a site |
| CN1201910A (en) * | 1997-06-11 | 1998-12-16 | 三星电子株式会社 | Target tracking method and device thereof |
| CN1673666A (en) * | 2005-04-15 | 2005-09-28 | 天津大学 | Micro-structural 3D information obtaining method based on phase shifting interference image sequence analysis |
| CN1847782A (en) * | 2006-03-29 | 2006-10-18 | 东南大学 | 2D Image Region Location Method Based on Raster Projection |
-
2009
- 2009-11-02 CN CN200910191318.1A patent/CN102052899B/en not_active Expired - Fee Related
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1198217A (en) * | 1996-05-31 | 1998-11-04 | 埃勒夫勘探产品公司 | Method for automatically determining stratification beds in a site |
| CN1201910A (en) * | 1997-06-11 | 1998-12-16 | 三星电子株式会社 | Target tracking method and device thereof |
| CN1673666A (en) * | 2005-04-15 | 2005-09-28 | 天津大学 | Micro-structural 3D information obtaining method based on phase shifting interference image sequence analysis |
| CN1847782A (en) * | 2006-03-29 | 2006-10-18 | 东南大学 | 2D Image Region Location Method Based on Raster Projection |
Non-Patent Citations (1)
| Title |
|---|
| 谢勤岚 等.结合两步估算和运动补偿的子像素运动估计.《计算机工程与应用》.2009,第45卷(第7期),190-191,198. * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN102052899A (en) | 2011-05-11 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN103575227B (en) | A kind of vision extensometer implementation method based on digital speckle | |
| CN103471910B (en) | A kind of elongation at break of metal material intelligent test method followed the tracks of based on random point | |
| CN111191629B (en) | Image visibility detection method based on multiple targets | |
| CN108562250B (en) | Keyboard keycap flatness rapid measurement method and device based on structured light imaging | |
| CN104616325B (en) | A kind of large surfaces Light stripes center extraction method of quick high accuracy | |
| Lu et al. | Vision-based structural displacement measurement under ambient-light changes via deep learning and digital image processing | |
| CN104240264A (en) | Height detection method and device for moving object | |
| CN111998910A (en) | A multi-level water gauge water level visual measurement method and measurement system | |
| CN101819024A (en) | Machine vision-based two-dimensional displacement detection method | |
| CN105572143B (en) | The detection method of rolled material surface periodic defect in calender line | |
| CN103281513B (en) | Pedestrian recognition method in the supervisory control system of a kind of zero lap territory | |
| CN101929843B (en) | Device and method for automatically measuring spike length of rice | |
| CN106012778B (en) | Digital image acquisition analysis method for express highway pavement strain measurement | |
| CN108286948A (en) | A kind of deflection of bridge span detection method based on image procossing | |
| CN113095323A (en) | SIFT-improvement-based digital image correlation method real-time detection method | |
| CN102052899B (en) | Take trichromatic contrast ratio as method and the device of characteristic frame Matched measurement two-dimension displacement | |
| CN102095375A (en) | Method and device for measuring displacement along optical axis direction by using contrasts of three primary colors as characteristics | |
| CN112485262B (en) | Method and device for detecting apparent crack width and expansion evolution of concrete | |
| CN105513071A (en) | Topographic map schema quality evaluation method | |
| CN104268536B (en) | A kind of image method for detecting human face | |
| CN111583191A (en) | Refraction feature detection method based on light field EPI Fourier transform | |
| CN102095374B (en) | Method and device for measuring displacement along optical axis direction by using peaks and valleys of three primary colors as characteristics | |
| CN102102981B (en) | Method and device for measuring displacement using two-dimensional single primary color contrast as feature frame matching | |
| CN102022982B (en) | Method and device for measuring displacement using two-dimensional contrast as feature frame matching | |
| CN115035071A (en) | Visual detection method for black spot defect of PAD light guide plate |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C14 | Grant of patent or utility model | ||
| GR01 | Patent grant | ||
| CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20150930 Termination date: 20191102 |
|
| CF01 | Termination of patent right due to non-payment of annual fee |