CN100368767C - Two-dimensional image area positioning method based on grating projection - Google Patents
Two-dimensional image area positioning method based on grating projection Download PDFInfo
- Publication number
- CN100368767C CN100368767C CNB2006100391570A CN200610039157A CN100368767C CN 100368767 C CN100368767 C CN 100368767C CN B2006100391570 A CNB2006100391570 A CN B2006100391570A CN 200610039157 A CN200610039157 A CN 200610039157A CN 100368767 C CN100368767 C CN 100368767C
- Authority
- CN
- China
- Prior art keywords
- grating
- image
- width
- cloth
- coarse
- 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
Images
Abstract
The present invention discloses a fast and reliable two-dimensional image area positioning method based on grating projection, which can identify the source and the characteristics of image noise and inhibit the image noise. A set of rough and fine grating is projected on objects after reference pictures are shot with a camera, rough positioning images are obtained after threshold cutting, image filtering, etc. when each grating is projected and shot, and fine grating gray value images are obtained by filtering fine grating images. Pixel points on all rough positioning images are traversed line by line, the space of detected objects is cut into strip-shaped areas after being processed with compensating algorithm, etc., and rough positioning result images corresponding to the information are obtained. Boundary information on the information is overlapped on all fine grating gray value images to be analyzed for carrying out traverse line by line and secondary curve fitting. Finally, extreme points in all fine grating gray value images are overlapped, and the information of the extreme points is used as the final dividing results for object space.
Description
Technical field
The present invention relates to a kind of two-dimensional image area positioning method when using optical grating projection equipment that object is carried out three-dimensionalreconstruction, relate in particular to a kind of two-dimensional image area positioning method based on optical grating projection.
Background technology
(Reverse Engineering, RE) technology is the later stage eighties 20th century to appear at the new technology in the advanced manufacturing field to reverse-engineering.Great majority mainly concentrate in the reverse reconstruct in kind about the research of reverse-engineering, and promptly the cad model reconstruct of product material object and the manufacture view of final products are called " reverse-engineering in kind ".Its flowage structure is seen Fig. 1.
The optical grating projection method is a kind of in the reverse Engineering Technology in kind, has the complete noncontact of testing process, data space resolution height, disposable moment projection and realizes that directly the three dimensions body form detects and obtain the characteristics of three-dimensional information.Have multiple advantages such as low, with low cost to environmental requirement, that use is easy to operate in actual applications.System's composition diagram is seen Fig. 2.
The basic thought of projection grating method is to throw various structured lights (adopting sinusoidal grating or rectangular raster in practical operation usually) to object, by observing being different from the direction of projection optical axis, utilize the triangle geometric relationship of object, subpoint, observation station to come the three-dimensional look information of Measuring Object.
The final purpose of reverse-engineering is for the reconstruct three dimensional practicality, in theory, does not need by any supplementary means, only needs directly to detect two width of cloth different images that obtained by binocular CCD, and the position parallax of object just can restore mock-up in the comparison diagram.But concerning computing machine, object in Direct Recognition two width of cloth pictures and the potential difference difficulty that compares between the two are too big, therefore need the design grating fringe to come secondary computer that picture is discerned.With regard to the identifying of same object in two width of cloth images,, can be equal to identification to object features zone in the single width picture to the identification of object parallax under different camera lenses.This is because if can identify each characteristic of object on the single width picture, and so only the simple same characteristic area that compares among two figure of need just can obtain the error between the two easily.But for general picture, that the object in the picture does not have itself is regular, obviously be convenient to the feature discerned, thus object is carried out optical grating projection, so that the shape of stripes that is suitable for identification of body surface occurrence law in the image.Since still very difficult to the identification of single width stripe pattern, so adopt usually in actual applications object is carried out one group of optical grating projection, then analyze the method for one group of grating fringe image and go to carry out the object space division, to obtain the object parallax information.
Like this potential difference identification problem of same object in two CCD pictures just is converted into identification problem to striped in single CCD serial picture.Fringe number that can be identified in single CCD image series is many more, the position of striped identification is accurate more, mean that then the number that object space is divided is many more, the object features zone that can identify is many more, and this precision that also just means the object CAD point cloud that finally reconstructs is high more.
Relatively more classical space-division method is the Gray division methods.To be Gray put forward in the digital-to-analogue of nineteen fifty-three research digital communication and analog to digital conversion for it, claims gray code method method (Gray F.Pulse code Communications[P.US-Patent 2632058,1953-03-17]) again.Gray code method comes down to by one group of coding of forming of ash exponent number.So-called grey exponent number, have following character: a back numeral is compared with previous numeral has only one-bit digital different.Such as: 000,001,010,011,111,110,010,000 is exactly one group of ash exponent number.When carrying out spatial division, use the Gray method just to be meant one group of grating fringe of design, make the final area of space of dividing satisfy the character of grey exponent number.Use this method that object space is divided concrete principle and see Fig. 3.
Employed in its image space partition process is one group of light and dark black and white strip.By the computer control projection arrangement, black and white strip is projected on the testee successively.Shoot the object fringe gray level image that is projected successively by ccd video camera, and then resulting candy strip is carried out the black and white binary conversion treatment by computing machine, the pixel in white stripes zone in the image is labeled as " 1 ", and the pixel in blackstreak zone is labeled as " 0 ".
In each projection, each pixel of image just can obtain a unique binary number " 0 " or " 1 " like this.Treat that the whole projections of projection pattern obtain the binary number sequential combination with pixel after intact, pixel with identical numerical value has constituted a narrow belt-like zone, like this testee space with regard to corresponding be divided into numerous in the well-determined shaped like narrow of binary number zone.
Its process flow diagram is seen Fig. 4.The conversion accuracy of final 3-D view only depends on the density of grating fringe projection in theory, but then is limited by the quality of figure that video camera is got in actual applications.
In actual projection process, the variation of environment shadow, ground vibration, color of object surface and factor such as reflective all may have influence on the quality of the captured image of final CCD.In such cases, the space orientation effect of Gray method is unsatisfactory, the situation of mistake might occur.Therefore how designing new area positioning method makes it satisfy the key point that high-precision requirement is exactly the division of image two-dimensional space simultaneously under noise conditions.
Want to reduce in the whole partition process noise to the influence of net result, must at first clear and definite process in the character and the source of noise.The error that may comprise in the raster image that finally obtains has:
1) error of video camera generation
The noise that video camera produces comprises photon noise, dark current noise, photoresponse heterogeneity noise, reads noise, clutter noise, A/D transform error etc.
Photon noise is obeyed Poisson distribution, and under low illumination, low contrast condition, after other noises ined all sorts of ways inhibition, photon noise became overriding noise.
Dark current noise refers to sensitization pixel electric current owing to electron production under no optical condition.It is closely related with temperature and time shutter, and it also satisfies Poisson distribution.
Photoresponse heterogeneity noise does not have certain rules, and is different, main relevant with the manufacturing process of device because of device, has very big randomness.
Clutter noise is mainly derived from transmission channel and various device, as clock signal and supply voltage instability, and the electromagnetic interference (EMI) that is subjected in the transmission.This noise satisfies random randomness more, and spectral magnitude is indefinite.
Read noise and be the charge signal that the sensitization pixel produces be read out and the circuit that amplifies in the noise that produces, comprise reset noise that produces by reset circuit and the noise that produces by amplifier.This is a kind of random noise.
The A/D transformed error is independent of picture signal, has Gaussian distribution and additive property.This error is very little usually, compares and can ignore with other errors.
2) air flow
Air flow is mainly produced by two aspects: the firstth, by flowing that walking about of people in the shooting process produces; Second be since the high-power pointolite that system adopts after long-time illumination, what the circumference air was produced flows.Air flow will be brought adverse effect to the captured by camera picture.And, can't know that therefore it determines the result, so be difficult to quantitative test and simulation because this influence determines by the concrete mode of air flow.Generally speaking, air flow can be ignored to the influence of picture.But under the more situation of dust, this phenomenon then can not be out in the cold in the more weak and air of light intensity.
3) pixels dithers of video image acquisition
Owing to the fluctuation of image card interior pixels clock itself has caused the variation in sampling time, and because vision signal changes so variation that has produced the pixel value correspondence position.Pixels dithers shows as a kind of stochastic error.
4) noise that brings of mechanical hook-up
The noise of mechanical hook-up refers to that mainly stepper motor drives slide block and slide caused noise on guide rail, and this noise also can be produced by other vibration.So when taking, the entire environment peace and quiet of will trying one's best.In the case, this noise is quite little to the influence of system, can ignore.
5) noise that brings of lenticular lenses
Noise on the lenticular lenses mainly refers to the noise that dust brought on the lenticular lenses, and the performance of this noise on image is exactly the appearance of gray corrosion.In the image series of gained, this gray corrosion and because the gray face of object own or since the gray shade that object height change to form distinguish to some extent.In whole a series of figure, it is motionless all the time that the gray face of object and object height change the gray shade that forms, that is to say if they occur at first width of cloth figure, that in the end a width of cloth figure also should occur, and the position in image is motionless all the time, and this spot appears in the captured all images.Because the performance of the gray corrosion on the lenticular lenses in the caused image of dust is then different, it shows as in raster image has in certain width of cloth image, does not have in certain width of cloth image.
Summary of the invention
The invention provides a kind of clearly source of picture noise and characteristic in the fringe location process, the noise that is comprised in the image is contained the rapid and reliable object space that obtains high-precision requirement is divided the two-dimensional image area positioning method based on optical grating projection.
The present invention adopts following technical scheme:
A kind of two-dimensional image area positioning method that is used for the stripe pattern coding based on optical grating projection:
1) at first takes the picture of a width of cloth testee as reference base picture with the CCD camera, thick by projector equipment then one group, fine grating successively projects on the object, every projection one amplitude grating, the CCD camera carries out a width of cloth to it and takes, above-mentioned coarse grating is the grating of wall scroll width of fringe greater than 10 pixels, in the 1st width of cloth coarse grating picture black and white 2 stripe are arranged, the 2nd width of cloth coarse grating picture carries out five equilibrium to the black and white strip in the 1st width of cloth, obtain the striped that 4 black and white replace, the rest may be inferred, n width of cloth coarse grating picture carries out five equilibrium to the striped in the n-1 width of cloth coarse grating picture, obtains 2
nThe bar black and white replaces striped, above-mentioned fine grating is that width of fringe is 1/2nd a grating of width of fringe in last width of cloth coarse grating picture, in the wherein back width of cloth fine grating picture fringe position successively with a last width of cloth fine grating picture in fringe position skew l/k distance on equidirectional, l is the width of fringe in the fine grating, k is total width of cloth number of fine grating picture, and the l/k value is 0.4 to 0.6 pixel wide;
2) earlier reference base picture is carried out Threshold Segmentation, obtain the binary image of reference base picture; The coarse grating image is carried out filtering, again Threshold Segmentation is carried out in the corresponding zone of the white portion with the reference base picture binary image of coarse grating image after the filtering, and give the third color beyond the black and white for the corresponding zone of the black region with the reference base picture binary image of coarse grating image after the filtering, obtain the coarse positioning image; The fine grating image is carried out filtering, obtain subject image is carried out the accurately used fine grating gray level image in location of space;
3) pixel on first width of cloth coarse positioning image is traveled through line by line, sudden change information according to the black and white color, determine and obtain striped border on first width of cloth coarse positioning image, and the discontinuous part in striped border is carried out the serialization compensation deals with backoff algorithm, obtain having first width of cloth coarse positioning compensating images on the 1st continuous black and white border; Before this each continuous black and white border is added on second width of cloth coarse positioning image, pixel on the back coarse positioning image that superposes is traveled through line by line, and according to the sudden change information of black and white color, on this stack back coarse positioning image, determine the striped border on second width of cloth coarse positioning image, and the discontinuous part in striped border is carried out the serialization compensation deals with backoff algorithm, obtain having the 2nd, the 3rd continuous black and white border; And the like, on before this each continuous black and white border is added to the coarse positioning image current, pixel on the back coarse positioning image that superposes is traveled through line by line, and according to the sudden change information of black and white color, on this stack back coarse positioning image, determine the striped border on the current coarse positioning image, and the discontinuous part in striped border is carried out the serialization compensation deals with backoff algorithm, until obtaining having the continuous the 2nd
nThe-2 and the 2nd
n-1 black and white border becomes the shaped like narrow zone with the testee space segmentation thus, and has obtained the corresponding coarse positioning result images of information therewith;
Above-mentioned backoff algorithm is: the coarse grating picture of being obtained by video camera passes through the denoising of image, the Threshold Segmentation of integral image, pass through expansive working (the Rafael C.Gonzalez of profile extraction and edge thinning, image again, Richard E.Woods.Digital Image Processing.Beijing:Publishing House of Electronics Industry) after, obtains needed compensation contrast figure; Travel through the pixel in the present image line by line, find the noncoherent boundary point, the compensation reference point that finds the position on the compensation contrast figure to be complementary again with this noncoherent boundary point, the positional information according to the compensation reference point compensates noncoherent boundary;
4) boundary information on the coarse positioning result images is added on all fine grating gray level images to be analyzed, in each width of cloth fine grating gray level image, zone between per two borders in the coarse positioning result images is traveled through line by line, use least square method that conic fitting is carried out in this interval line by line, obtain the extreme point on every row in this interval, superpose at last extreme point in all fine grating gray level images, and use this extreme point information to divide the result as final object space.
Compared with prior art, the present invention has following advantage:
The striped object space that the present invention is mainly used under the noise situations is divided.Utilize invention, can effectively contain in the image noise to spatial division result's influence, the mistake that may occur in the time of can not occurring using the Gray method has improved the precision of striped spatial division.Two kinds of situations that the Gray method may be made mistakes are seen Figure 11, Figure 12.
In Figure 11, we can see: the ellipticity black region that the rightmost side of every amplitude grating has piece to be caused by noise be 000 according to Gray method elliptic region net result, and ideally this zone is 111.
In Figure 12, we can see because the carrying out image threshold segmentation effect is bad the uneven phenomenon of black and white strip may occur, thisly inhomogeneously will cause mistake.If in last figure, black with 0, white represents that with 1 finally coding result has two 110 zone.
For first kind of situation, our method has adopted the thinking of zone errors not being divided in the striped coarse positioning stage; In the striped coarse positioning stage result the same with the Gray method can appear too for second kind of situation, just the border is not accurate enough, but because final we use is the gray level image of stria, division result when giving up coarse positioning will not be so cord location time error and noise will can be brought among the result of final spatial division.
In the middle of the gray-scale map of stria image is handled, adopted the spin filter method of stripe pattern, effectively removed the random noise of striped tangential direction in the stripe pattern.The basic thought of spin filter is as follows: in the normal direction of striped, intensity profile changes maximum, and on the tangential direction of striped, it is minimum that intensity profile changes, and occupy between the two in the other direction.In the striped normal direction, grey scale change is bigger, and corresponding frequency spectrum is a broadband, and the frequency spectrum of stripe signal and noise is superimposed, can't be well-separated.In the tangential direction of striped, grey scale change is very little, is approximately constant, and corresponding frequency spectrum is a near arrowband the zero-frequency, and random noise still is distributed in high band.This situation is used conventional low-pass filter,, just the striped of zero-frequency and the noise spectrum of high frequency can be separated neatly as medium filtering or mean filter.This way can reach the effect of removing garbage and keeping useful information.The steps include: at first to determine the tangential direction of striped, only bar graph is carried out low-pass filtering then in the tangential direction of striped.That is to say, at first find the isoline of fringe gray level, on this gray scale isoline, make low-pass filtering then.
When the stria image was carried out conic fitting, we had considered the influence of noise to striped normal direction in the image.Therefore the analysis showed that in big its random noise of place of image striped normal black and white gray scale variable gradient also greatlyyer, the quadratic fit curved line arithmetic is improved.At first use in the image on the striped normal direction grey scale change to carry out conic fitting, then interpolation arithmetic is carried out in institute's match, use interpolation point and all original point to carry out match again at last than point.Algorithm has strengthened the weighted value that gradient of image and gray scale in the conic fitting algorithm changes mild point, makes the fitting result that finally obtains more accurate.
Description of drawings
Fig. 1 is the reverse-engineering process flow diagram.
Fig. 2 is a grating style three-dimension scanning system composition diagram.
Fig. 3 is Gray coding principle figure.
Fig. 4 is a Gray coding process flow diagram.
Fig. 5 is the fringe location process flow diagram.
Fig. 6 is a cord grating diagrammatic series of views.
Fig. 7 is a stria grating diagrammatic series of views.
Fig. 8 is a border algorithm synoptic diagram.
Fig. 9 is the border algorithm flow chart.
Figure 10 is the definition of the discontinuous situation of border arithmetic result.
Figure 11 is a Gray coding error situation one.
Figure 12 is a Gray coding error situation two.
Figure 13 is the obtaining of compensation image of coarse grating series.
Figure 14 slightly divides space compensation front and back effect contrast figure.
Figure 15 is a design sketch before and after the filtering of fine grating striped.
Figure 16 is that final raster image is divided the result.
Embodiment
A kind of two-dimensional image area positioning method that is used for the stripe pattern coding based on optical grating projection:
1) at first takes the picture of a width of cloth testee as reference base picture with the CCD camera, thick by projector equipment then one group, fine grating successively projects on the object, every projection one amplitude grating, the CCD camera carries out a width of cloth to it and takes, above-mentioned coarse grating is the grating of wall scroll width of fringe greater than 10 pixels, in the 1st width of cloth coarse grating picture black and white 2 stripe are arranged, the 2nd width of cloth coarse grating picture carries out five equilibrium to the black and white strip in the 1st width of cloth, obtain the striped that 4 black and white replace, the rest may be inferred, n width of cloth coarse grating picture carries out five equilibrium to the striped in the n-1 width of cloth coarse grating picture, obtains 2
nThe bar black and white replaces striped, and the concrete value of n is by the decision of CCD camera parameter, for high-resolution CCD camera, and desirable bigger of its corresponding n value, but the cord picture that should keep fine grating is after Threshold Segmentation, and stripe information is high-visible.In general, the value of n maximum will guarantee at least after the cord grating is got figure, wall scroll width of fringe among its thinnest amplitude grating figure is greater than 10 pixels, above-mentioned fine grating is that width of fringe is 1/2nd a grating of width of fringe in last width of cloth coarse grating picture, in the wherein back width of cloth fine grating picture fringe position successively with a last width of cloth fine grating picture in fringe position skew l/k distance on equidirectional, l is the width of fringe in the fine grating, k is total width of cloth number of fine grating picture, and the l/k value is 0.4 to 0.6 pixel wide;
2) earlier reference base picture is carried out Threshold Segmentation, obtain the binary image of reference base picture; The coarse grating image is carried out filtering, again Threshold Segmentation is carried out in the corresponding zone of the white portion with the reference base picture binary image of coarse grating image after the filtering, and give the third color beyond the black and white for the corresponding zone of the black region with the reference base picture binary image of coarse grating image after the filtering, obtain the coarse positioning image; The fine grating image is carried out filtering, obtain subject image is carried out the accurately used fine grating gray level image in location of space;
3) pixel on first width of cloth coarse positioning image is traveled through line by line, sudden change information according to the black and white color, determine and obtain striped border on first width of cloth coarse positioning image, and the discontinuous part (see figure 10) in striped border is carried out the serialization compensation deals with backoff algorithm, obtain having first width of cloth coarse positioning compensating images on the 1st continuous black and white border; Before this each continuous black and white border is added on second width of cloth coarse positioning image, pixel on the back coarse positioning image that superposes is traveled through line by line, and according to the sudden change information of black and white color, on this stack back coarse positioning image, determine the striped border on second width of cloth coarse positioning image, and the discontinuous part in striped border is carried out the serialization compensation deals with backoff algorithm, obtain having the 2nd, the 3rd continuous black and white border; And the like, on before this each continuous black and white border is added to the coarse positioning image current, pixel on the back coarse positioning image that superposes is traveled through line by line, and according to the sudden change information of black and white color, on this stack back coarse positioning image, determine the striped border on the current coarse positioning image, and the discontinuous part in striped border is carried out the serialization compensation deals with backoff algorithm, until obtaining having the continuous the 2nd
nThe-2 and the 2nd
n-1 black and white border becomes the shaped like narrow zone with the testee space segmentation thus, and has obtained the corresponding coarse positioning result images of information therewith;
Above-mentioned backoff algorithm is: the coarse grating picture of being obtained by video camera is through the denoising of image, the Threshold Segmentation of integral image, pass through the expansive working of profile extraction and edge thinning, image again after, obtain needed compensation contrast figure; Travel through the pixel in the present image line by line, find the noncoherent boundary point, the compensation reference point that finds the position on the compensation contrast figure to be complementary again with this noncoherent boundary point, the positional information according to the compensation reference point compensates noncoherent boundary;
The discontinuous situation that we define the striped border among Figure 10 is the non-striped border crack conditions in the image that causes that changed by object height.
4) boundary information on the coarse positioning result images is added on all fine grating gray level images to be analyzed, in each width of cloth fine grating gray level image, zone between per two borders in the coarse positioning result images is traveled through line by line, use least square method that conic fitting is carried out in this interval line by line, obtain the extreme point on every row in this interval, superpose at last extreme point in all fine grating gray level images, and use this extreme point information to divide the result as final object space.
Show below in conjunction with accompanying drawing the specific embodiment of the present invention is further described.This example is to the coding of two dimensional image in the three-dimensionalreconstruction process.Mainly may further comprise the steps:
A) grating design
We have designed one group of grating fringe when reality is used, 11 width of cloth altogether, cord series 7 width of cloth wherein, stria series 4 width of cloth.They satisfy the requirement to grating mentioned in the technical scheme.Grating fringe that can certainly other width of cloth numbers of design and use.
B) use the ccd video camera photographic images
At first take the reference base picture of an amplitude object, drag lenticular lenses by stepper motor then, for each amplitude grating, video camera is all got figure once to object.In order to guarantee to get the quality of figure, reduce during shooting as far as possible and walk about, not in the light acute variation, under the extremely strong situation of surround lighting object is taken.
C) ccd image being carried out object space divides
The coarse grating serial picture is carried out Filtering Processing, Threshold Segmentation.Be divided into two-way then and handle, the one tunnel is used for the coarse positioning of encoding, and one the tunnel is used for obtaining encoding compensation image (seeing Figure 13).Next calculate the division result (seeing Figure 14) of cord series with the two resulting result.
Fine positioning is to come analytical calculation according to the result of coarse positioning and subsequent stria grating image series.Key step has the filtering (seeing Figure 15) of fine grating image, the conic fitting of striped.Net result is seen Figure 16.
Claims (1)
1. two-dimensional image area positioning method based on optical grating projection that is used for stripe pattern is characterized in that:
1) at first takes the picture of a width of cloth testee as reference base picture with the CCD camera, then by one group of coarse grating of projector equipment elder generation's projection, one group of fine grating of back projection is on object, every projection one amplitude grating, the CCD camera carries out a width of cloth to the object that is projected this grating and takes, above-mentioned coarse grating is the grating of wall scroll width of fringe greater than 10 pixels, in the 1st width of cloth coarse grating picture black and white 2 stripe are arranged, the 2nd width of cloth coarse grating picture carries out five equilibrium to the black and white strip in the 1st width of cloth, obtain the striped that 4 black and white replace, the rest may be inferred, and n width of cloth coarse grating picture carries out five equilibrium to the striped in the n-1 width of cloth coarse grating picture, obtains 2
nThe bar black and white replaces striped, above-mentioned fine grating is that width of fringe is 1/2nd a grating of width of fringe in last width of cloth coarse grating picture, in the wherein back width of cloth fine grating picture fringe position successively with a last width of cloth fine grating picture in fringe position skew l/k distance on equidirectional, l is the width of fringe in the fine grating, k is total width of cloth number of fine grating picture, and the l/k value is 0.4 to 0.6 pixel wide;
2) earlier reference base picture is carried out Threshold Segmentation, obtain the binary image of reference base picture; The coarse grating image is carried out filtering, again Threshold Segmentation is carried out in the corresponding zone of the white portion with the reference base picture binary image of coarse grating image after the filtering, and give the third color beyond the black and white for the corresponding zone of the black region with the reference base picture binary image of coarse grating image after the filtering, obtain the coarse positioning image; The fine grating image is carried out filtering, obtain subject image is carried out the accurately used fine grating gray level image in location of space;
3) pixel on first width of cloth coarse positioning image is traveled through line by line, sudden change information according to the black and white color, determine and obtain striped border on first width of cloth coarse positioning image, and the discontinuous part in striped border is carried out the serialization compensation deals with backoff algorithm, obtain having first width of cloth coarse positioning compensating images on the 1st continuous black and white border; The black and white border that this is continuous is added on second width of cloth coarse positioning image, pixel on the back coarse positioning image that superposes is traveled through line by line, and according to the sudden change information of black and white color, on this stack back coarse positioning image, determine the striped border on second width of cloth coarse positioning image, and the discontinuous part in striped border is carried out the serialization compensation deals with backoff algorithm, obtain the 2nd, the 3rd continuous black and white border; The rest may be inferred, on before this each continuous black and white border is added to the coarse positioning image current, pixel on the back coarse positioning image that superposes is traveled through line by line, and according to the sudden change information of black and white color, on this stack back coarse positioning image, determine the striped border on the current coarse positioning image, and the discontinuous part in striped border is carried out the serialization compensation deals with backoff algorithm, until obtaining the continuous the 2nd
nThe-2 and the 2nd
n-1 black and white border becomes the shaped like narrow zone with the testee space segmentation thus, and has obtained and the corresponding coarse positioning result images of this shaped like narrow area information;
Above-mentioned backoff algorithm is: the coarse grating picture of being obtained by video camera is through the denoising of image, the Threshold Segmentation of integral image, pass through the expansive working of profile extraction and edge thinning, image again after, obtain needed compensation contrast figure; Pixel among the traversal compensation contrast figure finds the noncoherent boundary point line by line, the compensation reference point that finds the position on the compensation contrast figure to be complementary again with this noncoherent boundary point, and the positional information according to the compensation reference point compensates noncoherent boundary;
4) boundary information on the coarse positioning result images is added on all fine grating gray level images to be analyzed, in each width of cloth fine grating gray level image, zone between per two borders in the coarse positioning result images is traveled through line by line, use least square method that conic fitting is carried out in this interval line by line, obtain the extreme point on every row in this interval, superpose at last extreme point in all fine grating gray level images, and use this extreme point information to divide the result as final object space.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CNB2006100391570A CN100368767C (en) | 2006-03-29 | 2006-03-29 | Two-dimensional image area positioning method based on grating projection |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CNB2006100391570A CN100368767C (en) | 2006-03-29 | 2006-03-29 | Two-dimensional image area positioning method based on grating projection |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN1847782A CN1847782A (en) | 2006-10-18 |
| CN100368767C true CN100368767C (en) | 2008-02-13 |
Family
ID=37077434
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CNB2006100391570A Expired - Fee Related CN100368767C (en) | 2006-03-29 | 2006-03-29 | Two-dimensional image area positioning method based on grating projection |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN100368767C (en) |
Families Citing this family (11)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US8079656B2 (en) * | 2006-12-22 | 2011-12-20 | Palo Alto Research Center Incorporated | Method for decimation of images |
| CN102052899B (en) * | 2009-11-02 | 2015-09-30 | 重庆工商大学 | Take trichromatic contrast ratio as method and the device of characteristic frame Matched measurement two-dimension displacement |
| TW201218047A (en) * | 2010-10-26 | 2012-05-01 | Xiroku Accupoint Technology Inc | Object sensing device |
| CN102779337A (en) * | 2011-04-13 | 2012-11-14 | 南京大学 | Method for light band separation and peak positioning of structured light |
| CN102393964B (en) * | 2011-08-02 | 2013-09-25 | 中国科学院长春光学精密机械与物理研究所 | Strip gap detection method |
| CN102508578B (en) * | 2011-10-09 | 2015-07-22 | 清华大学深圳研究生院 | Projection positioning device and method as well as interaction system and method |
| CN103034373B (en) * | 2012-11-23 | 2015-09-09 | 广东威创视讯科技股份有限公司 | The automatic selecting method of battle array camera positioning image effective coverage, face and system |
| CN103344651B (en) * | 2013-05-08 | 2015-04-29 | 中北大学 | Method of detecting glass defects based on phase image processing |
| CN104949621B (en) * | 2015-06-04 | 2017-08-29 | 广东工业大学 | A kind of boundary alignment method of grating scale striped |
| CN105674893B (en) * | 2016-03-18 | 2018-10-19 | 广东工业大学 | Absolute grating scale based on cmos image sensor and its measurement method |
| CN109917601B (en) * | 2019-02-13 | 2021-05-28 | 盎锐（上海）信息科技有限公司 | Camera setting method and device based on rolling shutter |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US2632058A (en) * | 1946-03-22 | 1953-03-17 | Bell Telephone Labor Inc | Pulse code communication |
| CN1483999A (en) * | 2003-08-15 | 2004-03-24 | 清华大学 | Method and system for measruing object two-dimensiond surface outline |
-
2006
- 2006-03-29 CN CNB2006100391570A patent/CN100368767C/en not_active Expired - Fee Related
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US2632058A (en) * | 1946-03-22 | 1953-03-17 | Bell Telephone Labor Inc | Pulse code communication |
| CN1483999A (en) * | 2003-08-15 | 2004-03-24 | 清华大学 | Method and system for measruing object two-dimensiond surface outline |
Non-Patent Citations (3)
| Title |
|---|
| Accurate 3D measurement using a structured light system. R.J. Valkenburg,A.M. McIvor.Image and Vision Computing,No.16. 1998 * |
| 三维测量中光栅投影条纹边界编码方法的改进. 金亚，达飞鹏，盖绍彦.计算机与现代化，第1期. 2006 * |
| 结构光法形体检测系统中的条纹图像处理. 盖绍彦.东南大学硕士学位论文. 2004 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN1847782A (en) | 2006-10-18 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN100368767C (en) | Two-dimensional image area positioning method based on grating projection | |
| DE112012001243B4 (en) | Method for measuring depth values in a scene | |
| Ouyang et al. | Pavement cracking measurements using 3D laser-scan images | |
| KR101733228B1 (en) | Apparatus for three dimensional scanning with structured light | |
| CN105844633B (en) | Single frames structure optical depth acquisition methods based on De sequence and phase code | |
| CN103292741A (en) | Structured light vision measurement method for 3D surface profiles of objects on the basis of K-means color clustering | |
| Pages et al. | A new optimised De Bruijn coding strategy for structured light patterns | |
| CN111986162A (en) | Hyperspectral abnormal point rapid detection method based on rough positioning and collaborative representation | |
| CN113506348A (en) | Gray code-assisted three-dimensional coordinate calculation method | |
| Wujanz | Towards transparent quality measures in surface based registration processes: Effects of deformation onto commercial and scientific implementations | |
| Sabater et al. | Review of low-baseline stereo algorithms and benchmarks | |
| CN103728022A (en) | Correction method for poor image elements | |
| Koch et al. | Heightmap generation for printed circuit boards (PCB) using laser triangulation for pre-processing optimization in industrial recycling applications | |
| CN100449571C (en) | Threshold value dividing method based on single-pixel in three-dimensional scanning system | |
| US7257248B2 (en) | Non-contact measurement system and method | |
| CN111307069B (en) | Compact parallel line structured light three-dimensional scanning method and system | |
| CN114943761A (en) | Method and device for extracting center of light stripe of central line structure of FPGA (field programmable Gate array) | |
| CN114419317A (en) | Light strip center extraction method for light with complex environment line structure | |
| Zhang et al. | Discontinuity-preserving decoding of one-shot shape acquisition using regularized color | |
| CN111508022A (en) | Line laser stripe positioning method based on random sampling consistency | |
| CN111121637A (en) | Grating displacement detection method based on pixel coding | |
| Zhao et al. | Analysis of Data Point Cloud Preprocessing and Feature Angle Detection Algorithm | |
| CN105654068B (en) | A kind of target detection background estimating method based on fractal theory | |
| CN113689400B (en) | Method and device for detecting profile edge of depth image section | |
| Eastwood et al. | Autonomous image background removal for accurate and efficient close-range photogrammetry |
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 | ||
| ASS | Succession or assignment of patent right |
Owner name: SOWTHEAST UNIV. Effective date: 20131022 Owner name: SHENLING ELECTRIC MANUFACTURING CO., LTD., HAIAN Free format text: FORMER OWNER: SOWTHEAST UNIV. Effective date: 20131022 |
|
| C41 | Transfer of patent application or patent right or utility model | ||
| COR | Change of bibliographic data |
Free format text: CORRECT: ADDRESS; FROM: 210096 NANJING, JIANGSU PROVINCE TO: 226600 NANTONG, JIANGSU PROVINCE |
|
| TR01 | Transfer of patent right |
Effective date of registration: 20131022 Address after: 226600 Haian, Jiangsu province Haian Zhenhai Road, No. 88, South Road, No. Patentee after: Haian Shenling Electrical Appliance Manufacturing Co., Ltd. Patentee after: Southeast University Address before: 210096 Jiangsu city Nanjing Province four pailou No. 2 Patentee before: Southeast University |
|
| CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20080213 Termination date: 20180329 |
|
| CF01 | Termination of patent right due to non-payment of annual fee |