CN100491986C - Interference stripe automatic focusing and automatic detection method - Google Patents
Interference stripe automatic focusing and automatic detection method Download PDFInfo
- Publication number
- CN100491986C CN100491986C CNB2005101374559A CN200510137455A CN100491986C CN 100491986 C CN100491986 C CN 100491986C CN B2005101374559 A CNB2005101374559 A CN B2005101374559A CN 200510137455 A CN200510137455 A CN 200510137455A CN 100491986 C CN100491986 C CN 100491986C
- Authority
- CN
- China
- Prior art keywords
- mentioned
- interference
- contrast
- sharpness
- pixel
- 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.)
- Active
Links
Images
Landscapes
- Automatic Focus Adjustment (AREA)
Abstract
An interference fringe automatic focusing and automatic detecting method are disclosed that fit for optical interference system. It includes: gathering several interference patterns. Every interference pattern consists of many pixels, and every pixel has corresponding pixel intensity. And the location, articulation and contrast of the interference patterns are recorded. The articulation and contrast are obtained according to the calculation of articulation and contrast. The articulation of the interference pattern is analyzed, and the articulation distribution curve relative to the interference pattern can be obtained to get an optimal articulation location as the focusing location of the optical interference system. The contrast of the interference pattern is analyzed, and the contrast distribution curve relative to the interference pattern can be obtained. The interference scanning extent is determined, the contrast distribution curve is analyzed separately, or the contrast distribution curve and articulation distribution curve are hybrid analysis to determine the interference scanning extent.
Description
Technical field
The present invention relates to the detection technique of a kind of optical interference system, and be particularly related to a kind of interference fringe focusing and automatic testing method automatically.
Background technology
In recent years, the checkout equipment industry development was mainly towards the combining of photoelectricity science and technology and automatic technology, and along with the development of semiconductor technology, system's microminiaturization technology and nanosecond science and technology, traditional measurement and detection method can't satisfy the demand of industrial community.Therefore, emerging detection technique all utilizes nondestructive optical detecting method to be the basis, and it not only can reach minute sized fast detecting, cooperates signal processing technology can also improve the automaticity of trace routine.In the robotization optical detective technology, one of important step with detection method is measured in focusing for optical profile type automatically.No matter in the testing process on technology or on the finished product, using fast, the image-type detection method has become main flow.
In order correctly to inspect the feature of product, need obtain the preferable image of sharpness.And slower with the speed of manual method differentiation, and be subjected to human factor to cause accuracy to differ easily, reduce the fiduciary level that detects quality.Therefore in order to reach the purpose that the robotization optical profile type detects, the detection method that needs good automatic focus algorithm to come the auxiliary optical image-type.In the field of micro-size measurement, the most valued method is an interference of light technology, and it is very practical in the optical measurement field and important survey instrument, and measuring limit even can reach time nanometer (0.5nm) grade.
The principle of interference of light technology is to utilize the interference fringe that optical path difference produced between determinand surface and the reference surface to assess the optical characteristics on determinand surface, when collocation array light intensity sensor and position transducer, can reach the optical characteristic measurement of 2D and 3D, and then be applied to detect the feature such as surface quality, size, defective of determinand.
Commercially available optical microscope how based on manual focus, for example can add focusing (Target Mark) in light path at present, and chimeric with the focusing image is the focusing reference.This method can increase the efficient of artificial adjustment focusing, but the method still is subjected to the influence of peopleware, makes that the quality of repeatable accuracy is inconsistent.And in addition in the microscope framework, also can add the eyeglass of compensation focusing error.In addition, at the measured interference fringe of single-point photodetector, interference range (fringewindow) can be set also, and search the position that maximum interference fringe intensity maximum takes place in interference range, be best focusing with this position.
Yet, in existing technology, still having various problems, the object lens field depth of for example optical interference system is little, and the precision of manual operation is coarse, is not easy to capture efficiently the position of interference fringe or focus point.In addition, the method for the automatic focusing in the general microscope framework is subjected to the influence of interference fringe easily and loses accuracy.Therefore, how reaching interference fringe focusing and automatic testing method automatically, is to need the actively direction of research and development.
Summary of the invention
The invention provides a kind of interference fringe automatically focusing and automatic testing method, can be automatically and draw preferable focusing distance between object lens and the determinand effectively, and analyze the sweep limit that obtains interference fringe automatically.
The present invention proposes a kind of interference fringe focusing and automatic testing method automatically, comprises at first capturing a plurality of interference images.Each this interference image is made of a plurality of pixel, and each above-mentioned these pixel has corresponding pixel intensity.Write down the position of each above-mentioned these interference image, sharpness and contrast.This sharpness and this contrast are to calculate acquisition according to sharpness computation and contrast, wherein, according to the calculation of cross focusing method automatically, contrast is calculated above-mentioned these pixels that comprise each above-mentioned these interference image to sharpness meter, makes an intensity histogram at last; And analyze this intensity histogram, draw the dispersion of distribution, as this contrast of corresponding each above-mentioned these interference image.Analyze above-mentioned these sharpness of above-mentioned these interference images, obtain the sharpness distribution curve, to draw the best sharpness position, as the focal position of this optical interference system.Analyze above-mentioned these contrasts of above-mentioned these interference images, obtain the contrast distribution curve.Decision interferential scanning scope, wherein this contrast distribution curve is by independent analysis, or this contrast distribution curve and the mixed analysis of this sharpness distribution curve, to determine this interferential scanning scope.
According to preferred embodiment of the present invention, automatically focus and automatic testing method in described interference fringe, this step of above-mentioned these interference images of acquisition and this position of each above-mentioned these interference image of record, this step of this sharpness and this contrast, be the one of above-mentioned these interference images of every acquisition, just calculate this sharpness and this contrast.Again or, acquisition all just calculates above-mentioned these sharpness and above-mentioned these contrasts behind above-mentioned these interference images.
According to preferred embodiment of the present invention, this cross method of focusing automatically, for example comprise the mechanism of using cross shade, to be selected by above-mentioned these pixels that this cross shade is contained, the centre position of the intermediate pixel of this cross shade wherein is positioned at the one of above-mentioned these pixels that constitute this interference image.Pixel intensity according to above-mentioned these pixels that contained by this cross shade that are selected draws calculated value according to computation rule.Move this cross shade,, draw the total value of corresponding above-mentioned these calculated values according to this identical computation rule, as this sharpness of each above-mentioned these interference image of correspondence with above-mentioned these pixels of the zone line of scanned at least this interference image.
According to preferred embodiment of the present invention, automatically focus and automatic testing method in described interference fringe, the shape of this cross shade is to be the center with this intermediate pixel, 90 degree constitute in two pixels of first, second, third, fourth direction extension at least at interval, and each location of pixels is set with weighted value respectively to take advantage of this pixel intensity in correspondence.
According to preferred embodiment of the present invention, automatically focus and automatic testing method in described interference fringe, the weighted value of the intermediate pixel of this cross shade is zero, above-mentioned these weighted values of other above-mentioned these directions distribute, first direction and third direction symmetry be opposite in sign still, but second direction and four directions are to the symmetry opposite in sign.
According to preferred embodiment of the present invention, automatically focus and automatic testing method in described interference fringe, this calculated value for this pixel that is contained by this cross shade of correspondence, comprise: calculate this first direction to should third direction, and this second direction to should the four directions to, pairing each pixel intensity is multiplied by a plurality of pixel calculated values of corresponding above-mentioned these weighted values respectively.Then, with after the pixel calculated value addition of each pixel calculated value and its symmetry direction square, again the square value totalling is obtained this calculated value.
According to preferred embodiment of the present invention, automatically focus and automatic testing method in described interference fringe, in this step of above-mentioned these sharpness of analyzing above-mentioned these interference images, comprise the peak location that analyzes this sharpness distribution curve, be this best sharpness position.
According to preferred embodiment of the present invention, automatically focus and automatic testing method in described interference fringe, wherein analyze this step of this intensity histogram, comprise this contrast distribution curve of the picture frame order of making corresponding above-mentioned these interference images of above-mentioned these contrasts.Determine this contrast distribution scope according to this contrast distribution curve, and then determine this interferential scanning scope.Wherein this contrast distribution scope is to determine according to the curve distribution width of above-mentioned these contrasts greater than the contrast threshold value, or determines according to the maximum distribution width of a plurality of spikes of this contrast distribution curve.
According to preferred embodiment of the present invention, automatically focus and automatic testing method in described interference fringe, in this step of above-mentioned these contrasts of analyzing above-mentioned these interference images, comprise actual value or the slope value of analyzing this contrast distribution curve, obtain distribution range, as this interferential scanning scope greater than the reference threshold value.
According to preferred embodiment of the present invention, to focus automatically and automatic testing method in described interference fringe, this contrast threshold value is to obtain according to statistical mechanism.
According to preferred embodiment of the present invention, automatically focus and automatic testing method in described interference fringe, wherein in this step of above-mentioned these contrasts of analyzing above-mentioned these interference images, comprise this contrast distribution curve and this sharpness distribution curve, do a subdifferential respectively, obtain contrast differential curve and sharpness differential curve.In addition, with this contrast differential curve and this sharpness differential curve addition, become and interfere the focusing differential curve.This is interfered the focusing differential curve, carry out the noise filtering computing.Utilize the method for moving average of computing length, change this interference focusing differential cloth curve and become to interfere the focusing curve for people having the same aspiration and interest length.Wherein, the computing length of the method for moving average can be random length, is not limited to people having the same aspiration and interest length.According to this distribution range of interfering the focusing curve, determine this interferential scanning scope.
State with other purpose, feature and advantage and can become apparent on the present invention for allowing, preferred embodiment cited below particularly, and conjunction with figs. are described in detail below.
Description of drawings
Fig. 1 represents an embodiment of optical interference used in the present invention system.
Fig. 2 represents according to the embodiment of the invention, the steps flow chart of automatic focusing of interference fringe and automatic testing method.
Fig. 3 represents the contrast analysis process according to the embodiment of the invention.
Fig. 4 represents the decision contrast method according to the embodiment of the invention.
Fig. 5 represents according to one embodiment of the invention, decides the mechanism of interferential scanning scope according to contrast.
Fig. 6 is according to one embodiment of the invention, the structural representation of cross shade.
Fig. 7 represents that according to one embodiment of the invention cross shade is at the move mode of interference image.
Fig. 8 represents the sharpness analysis process according to the embodiment of the invention.
Fig. 9 represents according to one embodiment of the invention, according to contrast and sharpness analysis-by-synthesis, decides the mechanism of interferential scanning scope.
The main element description of symbols
100: light source
102: optical system
104: interference objective
106: sample
108: the beam split reflector element
110：CCD
112: the capturing images card
114: computer system
116: the capture distance
S200～S232: step
600: cross shade
700: interference image
702: pel array 702
804: peak location
Embodiment
How the special consideration of the present invention discerns the scope that interference fringe occurs, and then decision interferential scanning scope.Simultaneously, in the interferential scanning scope, carry out the analysis of sharpness,, thereby improve the searching efficiency of focus point, increase the degree of robotization in the hope of preferable focal position.
In the detection and measuring method of optical profile type, utilize the technology of the interference of light to detect the feature of determinand at large, in order to obtain higher measurement resolution, the light source of use is necessary for low same light modulation, at this moment, the low people having the same aspiration and interest length because of light source makes that the recognizable set of interference fringe is limited.Based on the detection method of image-type, be example with the light path framework of white light interference object lens, interference can occur in the focus place, and the scope that interference image occurs may only be the distance of a few μ m to tens μ m short on the light path.If adjust the focus place of determinand, be easy to because regulate the speed too fast and the slow influence of human eye reaction velocity, and miss the scope that interference image occurs to object lens with manual method.
Therefore, with the arrange in pairs or groups information of interference fringe in the automatic focus algorithm arithmograph picture of computer vision, can mouthful low degree of difficulty of seeking interference image, and improve the search speed of interference image, improve the accuracy of identification interference image.
The optical microscope framework that generally comprises optical interference circuit is subjected to the influence of interference fringe, is easy to generate the erroneous judgement of automatic focusing.Therefore can carry out in the mode that face detects.Except capturing the interference image with the set of diagrams image-position sensor, and utilize another group pattern formula light intensity sensor, judge on the sensor each pixel because of the variation of the light intensity that interference fringe caused, and then calculate a pairing contrast of interference image (contrast index).In addition, also can calculate sharpness (sharpness index).Through the acquisition and the analysis of a plurality of interference images of order of single pass, can determine the sweep limit of interference fringe, and can determine the position that focuses on simultaneously.
The present invention is applied to interference fringe appearance place and the sweep limit that determines interference fringe in the universe formula recognition image sequence, and reaches the function of automatic focusing.The method of using is directly to add up the histogram distribution width of every image, this width is considered as contrast, wider distribution person is interference fringe appearance place, write down the also histogram distribution width of statistical picture sequence simultaneously, make contrast index curve map, judge the scope that interference fringe occurs with this contrast index curve map.In addition, the automatic focus algorithm of cross of arranging in pairs or groups again, obtaining the articulation index curve map, and interference image and best focusing position the most clearly.The computing of again contrast index curve and articulation index curve being compared can obtain interfering the focusing curve, and the distribution range of curve can determine the scope of interferential scanning thus.
The direct image data processing of the present invention does not need to install additional in addition sensor, escapable cost.When judging interference image, the information of adding up whole image has the effect of filtered noise, improves accuracy.Cooperate the automatic focus algorithm of cross again, can obtain interference image and its position more clearly.
Below as an illustration, for a plurality of embodiment but the present invention not only be subject to for those embodiment.
Fig. 1 represents an embodiment of optical interference used in the present invention system.With reference to Fig. 1, the optical interference system for example comprises light source 100.This light source 100 may be the broadband or the light source of multifrequency.When light passes through optical system 102, by spectrophotometric unit 108, after reflection enters interference objective 104, spectrophotometric unit 122 beam split through interference objective inside, part light is projected to the surface of sample 106, be reflected by interference objective 104, and pass spectrophotometric unit 108, another part light is reflexed to standard reflection unit 120 by interference objective inside by spectrophotometric unit 122.Standard reflection unit 120 reflexes to spectrophotometric unit 122 to the light of this part again, produce with the sample reflected light and interfere so that reflected by spectrophotometric unit 122 again and pass spectrophotometric unit 108 according to same paths, and by image capture unit 110, for example CCD (chargecoupled device) absorbs.Having between interference objective 104 and the sample 106 can be by self-adjusting capture distance 116.The interference image that capturing images card (Frame Grabber) 112 acquisition image capture unit 110 are absorbed, and send the interference image of institute's acquisition to computer system 114, calculate sharpness and contrast, and sharpness and contrast done holistic approach, obtain best capture distance 116, and the interferential scanning scope of corresponding capture distance 116.Computer system 114 also according to the capture of the best distance 116 as focusing distance, the focus of adjusting interference objective 104 can drop on the sample 106, to obtain interference image the most clearly.
The present invention proposes for example to cooperate the system of Fig. 1, proposes automatic operating process, as shown in Figure 2.Fig. 2 represents according to the embodiment of the invention, the steps flow chart of automatic focusing of interference fringe and automatic testing method.In step S200, corresponding capture distance 116, acquisition interference image.In addition, for making more efficient operation,, can write down the position of interference image simultaneously in step 216.In addition, the position of interference image also can be according to the reference position, according to the order of interference image, push away its position.Therefore can get any suitable mode of the position of interference image, all go for the present invention.Then, in step S202, utilize cross focusing method automatically to be displaced into the pixel coverage of interference image.In step S204, calculate the image definition of present interference image.In step S206, the image definition that recording gauge is calculated.About the analytical calculation of contrast,, add up in the interference image number of respective pixel value on the other hand in step S208.In step S210, the size of the distribution range of calculating pixel value is to determine the contrast of corresponding this interference image.In step S212, the contrast that recording gauge calculates.In step S214, if also will continue to capture interference image, then get back to step S200, otherwise continue subsequent step.
In this, the analysis that is noted that interference image is just analyzed and record after also can waiting whole capturing images to store with record.Its actual operating process needn't be limited to the flow process of Fig. 1.About the computational details of sharpness and contrast, can describe in addition in the back.
In step S218,, make the sharpness distribution curve, and analyze position with optimized image sharpness according to the sharpness of resulting sequence.In step S220, also analysis determines between the lane place with big image definition in addition.In addition, in step S222,, make the contrast distribution curve, and analyze position with maximum image contrast according to the contrast of resulting sequence.In step S224, also analysis determines between the lane place with big picture contrast in addition.
In step S226, contrast distribution curve that comprehensive income arrives and sharpness distribution curve the two, carry out analytical procedure.In this, step S226 can have multiple analysis rule, can describe in the back again.Then, in step S228, find out the interference image position of best focusing.In step S230, also determine the sweep limit of interference image in addition.In step S232, according to the interference image position of the resulting best focusing of step S228, adjust the distance between object lens and the sample, make best focal position.
Below, describe than the detail analysis rule, wherein earlier do explanation to how calculating contrast.Fig. 3 represents the contrast analysis process according to the embodiment of the invention.With reference to Fig. 3, be equivalent to step S208～S212 of Fig. 2, at different capture distances 116, can comply with the interference image or the order of picture frame (Frame) that are captured and give a numbering (step 300).For each picture frame, do the statistics of histogram (histogram) with pixel value, wherein pixel value for example is a pixel intensity, again or GTG value (step 302).With picture frame I, because away from the position that produces interference fringe, so the intensity of image is more even, so the narrow distribution of pixel value.
When capturing the position of picture frame I+K, interference fringe occurs.Because when interference fringe was arranged, the intensity of pixel no longer was evenly, the concealed wire zone is for example arranged, white portion, and the gray area between the concealed wire.Therefore, pixel three zones generally can occur in histogrammic distribution, extends into one in a big way.
In step 304, the calculating of degree of comparing value.For each histogram, the scope that needs the decision pixel to extend, it can utilize general statistical analysis technique to decide.And just define the contrast value of corresponding diagram frame for this reason to the pixel extension.So with all contrast values that calculated, be transverse axis with the picture frame sequence number for example, respective distances 116 is made the contrast distribution curve.Then analyze the contrast distribution scope that this contrast distribution curve obtains having bigger distribution.Because the sequence number of picture frame is that corresponding relation is arranged with the capture of Fig. 2 distance 116, the contrast distribution scope can be converted into the interferential scanning scope.That is to say, the contrast distribution curve also can be directly with capture distance 116 as the transverse axis coordinate.
In addition, above-mentioned statistical analysis technique for example Fig. 4 is described, but Fig. 4 is not the only method of the present invention.In Fig. 4, for example,, utilize threshold value N at the histogram of picture frame I+K with the statistical of left figure
Th, (pixel number) is lower than threshold value N with number of pixels
ThThe pixel value filtering, and obtain new histogrammic distribution range, just for example max pixel value 400b is deducted the resulting value range of minimum pixel value 400a.In addition, also can for example right figure, to analyze the position that spike occurs, when a plurality of spike appears in histogram, the pixel value that the peak location 402b of max pixel value is for example arranged deducts the pixel value of the peak location 402a of minimum pixel value, and obtains the possible distribution range size of contrast.Therefore, corresponding each interference image that is captured all has the contrast of a correspondence, can obtain the distribution curve of contrast.
Fig. 5 represents according to one embodiment of the invention, decides the mechanism of interferential scanning scope according to contrast.With reference to Fig. 5, directly mode is still and adopts general statistical analysis.For example, according to the contrast distribution curve, can estimate the scope [502a, 502b] that interference fringe occurs, it for example utilizes threshold value to decide.Again for example, can utilize the contrast distribution slope of a curve to change, cooperate the mode of slope threshold value to decide again.At this moment, scope [502a, 502b] for example can be used as the scope of interferential scanning.Yet, according to the consideration that changes than actual capabilities, also can outwards extend a scope more again according to estimated scope [502a, 502b], for example 10% expanded range becomes scope [500a, 500b].This scope [500a, 500b] promptly is the scope of the interferential scanning that draws automatically.
Note to such an extent that be, above-mentioned statistical analysis technique only is some embodiment, is not the method that only has of the present invention.Therefore particularly for example when the scope of decision interferential scanning, image definition also may influence decision, also can add the consideration of sharpness, after will being described in.
Below describe the cross mechanism of focusing automatically, it can be used to determine image definition, and then the focusing distance between decision object lens and the sample.Fig. 6 is according to one embodiment of the invention, the structural representation of cross shade.Cross shade 600 of the present invention is the virtual cross shades 600 that set at two-dimensional array.It is the center with the intermediate pixel, extends at least two pixels with interval 90 degree in first, second, third, fourth direction and is constituted, and for example does explanation with two pixels.Each location of pixels is set with a weighted value a1～a8 respectively, as shown, and in order to take advantage of 8 pixel intensity in correspondence.The weighted value of intermediate pixel is made as zero, that is to say intermediate pixel, the aligning that moves as the position only, and do not do calculating.
In addition first direction and third direction the symmetry but opposite in sign, but second direction and four directions are to the symmetry opposite in sign.For example, the value of [a1, a2, a3, a4, a5, a6, a7, a8] is respectively [1,2 ,-2 ,-1,1,2 ,-2 ,-1], in order to be multiplied by the pixel intensity of corresponding conductively-closed.At the position of intermediate pixel, its computing method for example, eight value of taking advantage of (a
iP
i), [(a is calculated in i=1～8
1P
1)+(a
4P
4)]
2+ [(a
2P
2)+(a
3P
3)]
2+ [(a
5P
5)+(a
8P
8)]
2+ [(a
6P
6)+(a
7P
7)]
2, with this as statistical value, p wherein
iBe pixel intensity.P for example
1～p
8=10,20,33,100,55,21,73,200, corresponding this pixel, its calculated value is
[1×10+(-1)×100]
2+[2×20+(-2)×33]
2+[1×55+(-1)×200]
2+[2×21+(-2)×73]
2＝40617。
Be worth as statistical value with this.If the development length of cross shade surpasses two pixels,, change weighted value, but account form is the same then according to mentioned above principle.Above-mentioned counter system, if to fuzzy image, the pixel value of corresponding cross shade can tend to evenly, does not have obvious pixel value variation, so the pixel of symmetric position can cancel each other, and obtains smaller value.Otherwise statistical value has higher value.
For interference image, the center pixel of cross shade must be contained the central area of interference image at least.Fig. 7 represents that according to one embodiment of the invention cross shade is at the move mode of interference image.In Fig. 7, for interference image 700, be that the pel array 702 by two dimension is constituted.The center of cross shade 600 is positioned at the one pixel of the pel array 702 of two dimension.For example, move the center of cross shade 600, calculate aforesaid statistical value one by one, and then add up and draw calculated value, as the reference index (reference index) of sharpness.Fig. 7 only is an example.In addition, also can utilize process of heterodyning, with interference image two pixels that stretch out.So, the center of cross shade 600 just can be moved, and contains all pixels.Certainly, at the analysis of sharpness, other mode can be arranged also.All can use as long as can analyze the mode of sharpness, and be not limited to above-mentioned analysis mode.
Then, Fig. 8 represents the sharpness analysis process according to the embodiment of the invention.With reference to Fig. 8, the same with the step 300 of Fig. 3, need to analyze each picture frame.In step 800, set desired cross shade, and, give suitable weighted value at the pixel that cross shade covered.In step 802, then carry out as described in Figure 7 sharpness computation for each picture frame, and, be that transverse axis is made the sharpness distribution curve with the picture frame with the sharpness of each picture frame.According to this sharpness distribution curve, can analyze the position of spike (peak) 804.Because the sequence number of picture frame is that corresponding relation is arranged with the capture of Fig. 2 distance 116, therefore can obtain on sample 3, having the capture distance 116 of optimum focusing by the position of spike.That is to say, the sharpness distribution curve also can be directly with capture distance 116 as the transverse axis coordinate.In addition, the mode of analyzing spike is preferable selection, however selection that neither be only for example average, or consider again the secondary cause of actual design to obtain the optimum focusing distance through further analyzing according to spike.
Fig. 9 represents according to one embodiment of the invention, according to contrast and sharpness analysis-by-synthesis, decides the mechanism of interferential scanning scope.Because the contrast distribution curve is to be the assessment principle with the pixel intensity, the sharpness distribution curve serves as the assessment principle with the strength difference between pixel then, so both have different susceptibilitys for different sample images.In order to obtain the information of contrast and sharpness simultaneously, can utilize contrast distribution curve 902 and sharpness distribution curve 900, carry out a subdifferential 904.Afterwards, can obtain contrast differential curve 908 and sharpness differential curve 906 respectively.Both differential curves can show the information that interference fringe occurs.With two differential curve additions 910, synthesize the focusing differential curve 912 that has contrast and sharpness information simultaneously again.Because noise is introduced in the computing meeting of differential, therefore can utilize threshold value filtered noise 914, and obtain new focusing differential curve 916.For example the mean value with differential curve adds 3 times of standard deviations as threshold value, can obtain muting focusing differential curve behind this threshold value filtered noise.And for example and utilize the computing length L m of the people having the same aspiration and interest length of light source, differential curve 916 computings are become to interfere focusing curve 920 as the method for moving average 918.The scope that this interference focusing curve 920 is comprised is as the interferential scanning scope, and in the interferential scanning scope, the peak location of articulation curve 900 is for interfering focus image clearly, and its position is the position of focus point.When for example people having the same aspiration and interest length is the distribution range of 101 images, get the computing length L m=101 of the method for moving average, and the center with the method for moving average is the computing benchmark, and the scope of interfering focusing curve 920 to be comprised is that scope appears in interference image, and people having the same aspiration and interest length again stretches out half.
In sum, utilize the high contrast features of interference fringe in the present invention, at first, the histogram of pixel value in the statistical picture, promptly be the number of each pixel value, with the intensity distributions scope that obtains pixel on every image and the number on each intensity level, as the foundation of computed image contrast.Then utilize the picture contrast analysis, perhaps, utilize picture contrast and image definition hybrid analysis, interference fringe whether occurs with identification.Collocation cross automatic focus algorithm in the identification step of interference image, can correctly seek interference image clearly, and obtain good focusing position.
In addition, for example contrast curves and articulation curve separately behind the differential, can be got both differential curves again, two differential curve additions be synthesized differential curve with contrast and sharpness distributed intelligence, behind filtered noise, become with method of moving average computing again and interfere the focusing curve.Scope with reference to interfering the focusing scope that curve comprised as interferential scanning is applied to can improve automaticity on the interferential scanning instrument.
Interference fringe has the characteristic that contrast is higher than background image, utilizes that the information of contrast can be used as the information whether interference fringe occurs in the image.In the interference nidus, intensity values of pixels will have bigger distribution range in the image, and distribution range is bigger, and the presentation video contrast is bigger, so this distribution range can be considered as the contrast index of image.Therefore, the histogram of the number of all pixel values in the statistical picture of the present invention, and the distribution range of calculating pixel value, the image when being considered as interference fringe and occurring with the wider image of pixel value distribution range.
In addition, for fear of The noise, the threshold value of setting the pixel value number is as the criterion of judging distribution range.Pixel value distribution range in the record all images sequence is made contrast index curve, judges the appearance scope of interference image with contrast index curve.Interference image be distributed with a segment limit, only the interference fringe appearance of being judged with contrast place is not only possible case.The present invention's this moment adds focus algorithm again, obtains interference image clearly, the focus in the time of so obtaining interfering exactly.
Though the present invention discloses as above with preferred embodiment; right its is not in order to limit the present invention; any person of ordinary skill in the field; without departing from the spirit and scope of the invention; when can doing a little change and improvement, so the present invention's protection domain is as the criterion when looking the claim person of defining.
Claims (14)
1. an interference fringe is focused and automatic testing method automatically, is applicable to the optical interference system, it is characterized in that comprising:
Capture a plurality of interference images, wherein each this interference image is made of a plurality of pixel, and each above-mentioned these pixel has corresponding pixel intensity;
Write down the position of each above-mentioned these interference image, sharpness and contrast, wherein this sharpness and this contrast are to calculate acquisition according to sharpness computation and contrast, wherein, this sharpness meter calculates according to cross focusing method automatically at last, this contrast is calculated above-mentioned these pixels that comprise each above-mentioned these interference image, makes an intensity histogram; And analyze this intensity histogram, draw the dispersion of distribution, as this contrast of corresponding each above-mentioned these interference image;
Analyze above-mentioned these sharpness of above-mentioned these interference images, obtain the sharpness distribution curve, to draw the best sharpness position, as the focal position of this optical interference system;
Analyze above-mentioned these contrasts of above-mentioned these interference images, obtain the contrast distribution curve; And
Decision interferential scanning scope, wherein this contrast distribution curve is by independent analysis, or this contrast distribution curve and the mixed analysis of this sharpness distribution curve, to determine this interferential scanning scope.
2. focus automatically and automatic testing method according to the described interference fringe of claim 1, it is characterized in that capturing this step of above-mentioned these interference images and this position of each above-mentioned these interference image of record, this step of this sharpness and this contrast, be the one of above-mentioned these interference images of every acquisition, just calculate this sharpness and this contrast.
3. focus automatically and automatic testing method according to the described interference fringe of claim 1, it is characterized in that capturing this step of above-mentioned these interference images and this position of each above-mentioned these interference image of record, this step of this sharpness and this contrast, after being whole above-mentioned these interference images of acquisition, just calculate above-mentioned these sharpness and above-mentioned these contrasts.
4. focus automatically and automatic testing method according to the described interference fringe of claim 1, it is characterized in that the calculation of this cross focusing method automatically, comprising:
Use the mechanism of cross shade, will be selected by above-mentioned these pixels that this cross shade is contained, the centre position of the intermediate pixel of this cross shade wherein is positioned at the one of above-mentioned these pixels that constitute this interference image;
Pixel intensity according to above-mentioned these pixels that contained by this cross shade that are selected draws calculated value according to computation rule; And
Move this cross shade,, draw the total value of corresponding above-mentioned these calculated values according to this identical computation rule, as this sharpness of each above-mentioned these interference image of correspondence with above-mentioned these pixels of the zone line of scanned at least this interference image.
5. focus automatically and automatic testing method according to the described interference fringe of claim 4, the shape that it is characterized in that this cross shade is to be the center with this intermediate pixel, 90 degree constitute in two pixels of first, second, third, fourth direction extension at least at interval, and each location of pixels is set with weighted value respectively to take advantage of this pixel intensity in correspondence.
6. focus automatically and automatic testing method according to the described interference fringe of claim 5, the quantity that it is characterized in that these at least two pixels is two pixels, and wherein the weighted value of this intermediate pixel is zero.
7. focus automatically and automatic testing method according to the described interference fringe of claim 6, the weighted value that it is characterized in that this intermediate pixel is zero, above-mentioned these weighted values of other above-mentioned these directions distribute, first direction and third direction symmetry be opposite in sign still, but second direction and four directions are to the symmetry opposite in sign.
8. focus automatically and automatic testing method according to the described interference fringe of claim 7, it is characterized in that comprising for this calculated value of this pixel that is contained by this cross shade of correspondence:
Calculate this first direction to should third direction, and this second direction to should the four directions to, pairing each pixel intensity is multiplied by a plurality of pixel calculated values of corresponding above-mentioned these weighted values respectively; And
With after the pixel calculated value addition of each pixel calculated value and its symmetry direction square, square value is added up obtain this calculated value again.
9. focus automatically and automatic testing method according to the described interference fringe of claim 1, it is characterized in that in this step of above-mentioned these sharpness of analyzing above-mentioned these interference images, comprising the peak location that analyzes this sharpness distribution curve, is clear position of this best.
10. focus automatically and automatic testing method according to the described interference fringe of claim 1, it is characterized in that analyzing this step of this intensity histogram, comprising:
At this interference image, make this intensity histogram of a plurality of number of pixels of corresponding above-mentioned these pixel intensity of above-mentioned these pixels;
Determine this contrast according to this intensity histogram, and above-mentioned these contrasts of above-mentioned these interference images be made into this contrast distribution curve, and then determine this interferential scanning scope,
Wherein this contrast is to determine according to stating the curve distribution width of these number of pixels greater than the number of pixels threshold value on these histogrammic above-mentioned these intensity, or determines according to the maximum distribution width of a plurality of spikes of this intensity histogram.
11. focus automatically and automatic testing method according to the interference fringe that claim 1 is described, it is characterized in that in this step of above-mentioned these contrasts of analyzing above-mentioned these interference images, comprise actual value or the slope value of analyzing this contrast distribution curve, obtain distribution range, as this interferential scanning scope greater than the reference threshold value.
12. focus automatically and automatic testing method according to the interference fringe that claim 11 is described, it is characterized in that this contrast threshold value is to obtain according to statistical mechanism.
13. focus automatically and automatic testing method, it is characterized in that in this step of above-mentioned these contrasts of analyzing above-mentioned these interference images, comprising according to the interference fringe that claim 1 is described:
To this contrast distribution curve and this sharpness distribution curve, do a subdifferential respectively, obtain contrast differential curve and sharpness differential curve;
With this contrast differential curve and this sharpness differential curve addition, become and interfere the focusing differential curve;
This is interfered the focusing differential curve, carry out the noise filtering computing;
Under predetermined computing length, utilize the method for moving average, change this interference focusing differential curve and become to interfere the focusing curve; And
According to this distribution range of interfering the focusing curve, determine this interferential scanning scope.
14. focus automatically and automatic testing method according to the interference fringe that claim 13 is described, it is characterized in that this computing length of being scheduled to is people having the same aspiration and interest length.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CNB2005101374559A CN100491986C (en) | 2005-12-30 | 2005-12-30 | Interference stripe automatic focusing and automatic detection method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CNB2005101374559A CN100491986C (en) | 2005-12-30 | 2005-12-30 | Interference stripe automatic focusing and automatic detection method |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN1991341A CN1991341A (en) | 2007-07-04 |
| CN100491986C true CN100491986C (en) | 2009-05-27 |
Family
ID=38213751
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CNB2005101374559A Active CN100491986C (en) | 2005-12-30 | 2005-12-30 | Interference stripe automatic focusing and automatic detection method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN100491986C (en) |
Families Citing this family (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| TWI336767B (en) | 2007-07-05 | 2011-02-01 | Ind Tech Res Inst | Method for calibration of image and apparatus for acquiring image |
| CN101349540B (en) * | 2007-07-18 | 2011-06-29 | 财团法人工业技术研究院 | Image regulating method and image viewfinding apparatus |
| CN102313524B (en) * | 2010-07-05 | 2013-09-18 | 致茂电子(苏州)有限公司 | Image acquiring device and method |
| CN102595049B (en) * | 2012-03-16 | 2014-11-26 | 盛司潼 | Automatic focusing control system and method |
| CN106534659B (en) * | 2015-09-11 | 2019-07-02 | 深圳市道通智能航空技术有限公司 | A kind of camera focusing method and system |
| CN106679938B (en) * | 2015-11-06 | 2018-11-16 | 福建福光股份有限公司 | A kind of motorized zoom lens focusing curve detection system and method |
| CN107770520A (en) * | 2017-11-13 | 2018-03-06 | 凌云光技术集团有限责任公司 | A kind of line-scan digital camera auxilary focusing method and device |
| CN109883355B (en) * | 2019-03-05 | 2020-09-11 | 深圳市中图仪器股份有限公司 | Method and device for automatically searching interference fringes |
-
2005
- 2005-12-30 CN CNB2005101374559A patent/CN100491986C/en active Active
Also Published As
| Publication number | Publication date |
|---|---|
| CN1991341A (en) | 2007-07-04 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN100491986C (en) | Interference stripe automatic focusing and automatic detection method | |
| US10578554B2 (en) | Spectrum-scanned SPR imaging detection system | |
| US10480939B2 (en) | High speed stereoscopic pavement surface scanning system and method | |
| US9778206B2 (en) | Defect inspection device and defect inspection method | |
| US7929044B2 (en) | Autofocus searching method | |
| US5270794A (en) | Fine structure evaluation apparatus and method | |
| US6674572B1 (en) | Confocal microscopic device | |
| CN104964982B (en) | Glass surface true and false defect identification method and system based on OCT complex signals | |
| JP5354254B2 (en) | Shape measuring apparatus and method, and program | |
| JPH01126529A (en) | Matching of particle to be detected | |
| CN107203743B (en) | Face depth tracking device and implementation method | |
| CA3066752A1 (en) | High speed stereoscopic pavement surface scanning system and method | |
| US7362421B2 (en) | Analysis of signal oscillation patterns | |
| CN104350358B (en) | Method for measuring surface shape and device for measuring surface shape | |
| CN111369484B (en) | Rail profile detection method and device | |
| US11415500B2 (en) | Particulate observation device and particulate observation method | |
| CN114112959A (en) | Rapid high-precision hyperspectral depth imaging system | |
| CN111369533B (en) | Rail profile detection method and device based on polarization image fusion | |
| EP1611433A1 (en) | Glazing inspection | |
| CN208520339U (en) | Three-dimensional surface shape detection device without axial scan | |
| TWI357973B (en) | Apparatus and method for simulataneous confocal fu | |
| JPH09118153A (en) | Method and device for inspecting stringing | |
| Ciamberlini et al. | Weaving defect detection by Fourier imaging | |
| CN108955568A (en) | Three-dimensional surface shape detection device and its application method without axial scan | |
| WO2022267687A1 (en) | Micro-nano deep trench structure fast measurement method on basis of white light interference |
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 |