CN103345642B - A kind of image matching method based on dotted line antithesis - Google Patents
A kind of image matching method based on dotted line antithesis Download PDFInfo
- Publication number
- CN103345642B CN103345642B CN201310268597.3A CN201310268597A CN103345642B CN 103345642 B CN103345642 B CN 103345642B CN 201310268597 A CN201310268597 A CN 201310268597A CN 103345642 B CN103345642 B CN 103345642B
- Authority
- CN
- China
- Prior art keywords
- prime
- target
- antithesis
- point set
- theta
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Expired - Fee Related
Links
Landscapes
- Apparatus For Radiation Diagnosis (AREA)
Abstract
The invention discloses a kind of image matching method based on dotted line antithesis, comprise: from reference image R and target image S, extract respectively straight line, to obtain consult straight line set and target line set, be mapped to dual spaces with reference to all straightways in straight line set and target line set from image space respectively, to obtain with reference to antithesis point setTotal m dual points, and target antithesis point setTotal n dual points, concentrate contiguous dual points to merge with reference to antithesis point set and target dual points respectively, to obtain new reference antithesis point set and new target antithesis point set, estimate the rotation transformation parameter between new reference antithesis point set and new target antithesis point set, estimate the translation transformation parameter between new reference antithesis point set and new target antithesis point set. The present invention many straightways of the image space Fracture fragmentation bar straightway that again permeates, has improved matching efficiency and stability equivalently.
Description
Technical field
The invention belongs to computer vision and image processing field, more specifically, relate to a kind of based on pointThe image matching method of line antithesis.
Background technology
Images match is a very important problem in computer vision, is to realize graphical analysis and figureA basic step of senior visual performances such as picture understanding etc., production assembly automation, quality testing,The crowds such as target identification location, safety monitoring, super-resolution rebuilding, panorama splicing, Medical image fusionMulti-field extensive application.
In images match, a crucial problem is to select which kind of feature, and suitable feature can makeQuick, the matching result of matching process is stable and accurately. The gray scale that the most basic feature of digital picture is,Images match based on gray scale be propose the earliest, one of the image matching technology of most study, but depositTo the shortcoming such as illumination variation, noise-sensitive. In order to improve stability and the speed of images match, angleThe geometric primitive such as point, edge, straight line, circular arc are extracted and are used as matching characteristic from image,Images match based on geometric primitive has been carried out higher level analysis to image, has got rid of many uselessInformation, the feature using has more directly perceived, clear and definite meaning, is that current image matching technology grindsThe focus of studying carefully.
Point is the most frequently used geometric primitive, and How to choose point has material impact to the result of images match.The number of point is many, and matching precision is higher, but speed is slower; Otherwise precision reduces, speed is carriedHigh. Straight line can be undertaken after various processing obtaining by edge, further removed outlier andNoise spot is more senior geometric primitive. The quantity of the straight line in general, extracting from imageWill be far less than the quantity of the point extracting, so use the image matching algorithm meeting tool of straight line as featureThere is higher computational efficiency. First some straight line matching process utilizes part straight line to generate that some are possibleCoupling hypothesis, then utilizes remaining straight line to verify. Some straight line matching process uses straight lineBetween intersection point complete coupling. Some straight line matching process is according to the Pixel Information structure in straight line neighborhoodMake high dimensional feature vector, then complete coupling by the similitude of comparative feature vector. But large at presentMost straight line matching process all carry out at image space, straight to what cause due to noise and pollutionLine segment fracture is broken, and the situation such as the straightway disconnection causing due to partial occlusion is difficult to carry out intuitivelyAnd effectively process, so had a strong impact on the high efficiency of matching process and the stability of matching result.
Summary of the invention
For above defect or the Improvement requirement of prior art, the invention provides a kind of based on dotted line pairEven image matching method, its object is by first straight line being transformed into dual-space from image spaceBetween become a little, thereby straight line matching problem is converted into Point set matching problem, then enter at dual spacesRow point merges, and finally in dual spaces, mates two point sets, thereby realizes images match, with theseEffect ground, many straightways of the image space Fracture fragmentation bar straightway that again permeates, has improvedMatching efficiency and stability.
For achieving the above object, according to one aspect of the present invention, provide a kind of based on dotted line antithesisImage matching method, comprise the following steps:
(1) from reference image R and target image S, extract respectively straight line, to obtain consult straight line collectionClose and target line set, the consult straight line set of wherein extracting from R is designated asAltogetherHave m bar straightway, the target line set of extracting from S is designated asTotal n bar is straightLine segment;
(2) respectively with reference to straightways all in straight line set and target line set from image spaceBe mapped to dual spaces, to obtain with reference to antithesis point setTotal m dual points, withAnd target antithesis point setTotal n dual points;
(3) concentrate contiguous dual points to merge with reference to antithesis point set and target dual points respectively,To obtain new reference antithesis point set and new target antithesis point set;
(3-1) counter i=1 is set, counter cnt=1, state array [St1,St2,…,Stm]=1;
(3-2) judge whether i is less than or equal to m, if proceed to step (3-3), otherwise obtain newReference antithesis point setTotal u dual points, and proceed to step (3-13);
(3-3) judge StiWhether equal 1, if proceed to step (3-4), otherwise proceed to step(3-12);
(3-4) antithesis point set G to be merged being set is empty set, from concentrating and take out i with reference to dual pointsWith reference to dual pointsIts coordinate is (θi,ρi), added G, and St is seti=0;
(3-5) counter j=1 is set;
(3-6) judge whether j is less than or equal to m, if proceed to step (3-7), otherwise proceed to stepSuddenly (3-11);
(3-7) judge whether j equals i, or StjWhether equal 0, if enter step (3-10),Otherwise directly enter step (3-8);
(3-8) judgement | θj-θi|<TθAnd | ρj-ρi|<TρWhether set up, enter step (3-9) if set up,Otherwise enter step (3-10), wherein TθAnd TρFor default threshold value;
(3-9) by j with reference to dual pointsAdd antithesis point set G to be merged, and St is setj=0;
(3-10) j=j+1 is set, then returns to step (3-6);
(3-11) all in antithesis point set G to be merged are merged with reference to dual points, to obtainNew reference dual pointsIts coordinate is (θ ', ρ '), and cnt=cnt+1 is set;
(3-12) i=i+1 is set, then returns to step (3-2);
(3-13) all target dual points of concentrating for target dual points, adopt and above-mentioned steps(3-1) to (3-12) identical step, to obtain new target antithesis point setTotal v dual points;
(4) estimate the rotation transformation ginseng between new reference antithesis point set and new target antithesis point setNumber;
(5) estimate the translation transformation ginseng between new reference antithesis point set and new target antithesis point setNumber.
Preferably, in step (3-11) with reference to dual pointsCoordinate (θ ', ρ ') specifically adopt belowFormula:
Wherein, k is the quantity with reference to dual points in antithesis point set G to be merged; wzWeight coefficient, itsValue equal with reference to dual pointsCorresponding straightwayLength, wherein z be between 1 to k appointMeaning integer.
Preferably, step (4) comprises following sub-step:
(4-1) counter cti=1 is set, initializes cumulative array [A1,A2,…A180]=0;
(4-2) judge whether cti is less than or equal to u, if proceed to step (4-3), otherwise enterStep (4-9);
(4-3) counter ctj=1 is set;
(4-4) judge whether ctj is less than or equal to v, if proceed to step (4-5), otherwise enterStep (4-8);
(4-5) judge whether ctj equals cti, if enter step (4-7), otherwise directly enterEnter step (4-6);
(4-6) calculate its corresponding φ according to following formula, and make Aφ=Aφ+1;
Wherein, [] represents to round.
(4-7) ctj=ctj+1 is set, then returns to step (4-4);
(4-8) cti=cti+1 is set, then returns to step (4-2);
(4-9) find out [A1,A2,…A180] in the element of numerical value maximum, be designated as Aφ, the φ that it is correspondingmax=φIt is exactly rotation parameter.
Preferably, step (5) comprises following sub-step:
(5-1) counter a=1 is set, cumulative matrixItsMiddle height is the height of target image S, and width is the width of target image S, maximum max=0,Translation parameters txmax=0,tymax=0;
(5-2) judge whether a is less than or equal to u, if proceed to step (5-3), else process knotBundle;
(5-3) counter b=1 is set;
(5-4) judge whether b is less than or equal to v, if proceed to step (5-5), otherwise enter stepSuddenly (5-18);
(5-5) counter c=1 is set;
(5-6) judge whether c is less than or equal to u, if proceed to step (5-7), otherwise enter stepSuddenly (5-17);
(5-7) judge whether c equals a, if enter step (5-13), otherwise directly enter stepSuddenly (5-8);
(5-8) counter d=1 is set;
(5-9) judge whether d is less than or equal to v, if proceed to step (5-10), otherwise enter stepSuddenly (5-13);
(5-10) judge whether d equals b, if enter step (5-12), otherwise directly enterStep (5-11);
(5-11) calculate its corresponding (t according to following formulax,ty), and make Btx,ty=Btx,ty+1;
Wherein, [] represents to round.
(5-12) d=d+1 is set, then returns to step (5-9);
(5-13) c=c+1 is set, then returns to step (5-6);
(5-14) find outThe element of middle numerical value maximum, is designated asBtxab,tyab, and record possible translation parameters txab,tyab;
(5-15) judge whether max is greater than Btxab,tyabIf, enter step (5-17), otherwise straightTap into into step (5-16);
(5-16) max=B is settxab,tyab,txmax=txab,tymax=tyab;
(5-17) b=b+1 is set, then returns to step (5-4);
(5-18) a=a+1 is set, then returns to step (5-2).
In general, the above technical scheme of conceiving by the present invention compared with prior art, canObtain following beneficial effect:
1, the present invention becomes a little by straight line is mapped to dual spaces from image space, and at antithesisA fusion is carried out in space, has solved the low and stability of matching efficiency that the broken straightway that ruptures causesPoor problem;
2, the present invention utilizes the method for Point set matching to solve the problem of straight line coupling, has realized having and has revolvedTurn, efficient stable coupling between translation, illumination variation and the image of partial occlusion.
Brief description of the drawings
Fig. 1 is the flow chart that the present invention is based on the image matching method of dotted line antithesis.
Fig. 2 is the schematic diagram that straightway is mapped to dual spaces and obtains dual points from image space.
Fig. 3 is the flow chart of dual points fusion method.
Fig. 4 is that dual points merge the schematic diagram in dual spaces and image space effect.
Fig. 5 is the flow chart of estimating the rotation transformation parametric technique between template image and target image.
Fig. 6 is the flow chart of estimating the translation transformation parametric technique between template image and target image.
Detailed description of the invention
In order to make object of the present invention, technical scheme and advantage clearer, below in conjunction with accompanying drawingAnd embodiment, the present invention is further elaborated. Should be appreciated that described herein concreteEmbodiment only, in order to explain the present invention, is not intended to limit the present invention. In addition described,Involved technical characterictic in each embodiment of the present invention just can as long as do not form to conflict each otherMutually to combine.
The present invention proposes a kind of new straight line matching process and complete images match, utilize dotted line antithesisStraightway in image space is mapped as to dual points, utilizes dual points to merge the fragmentation of rupturing equivalentlyStraightway be again fused to same straightway, utilize antithesis point set to carry out images match.
The present invention is based on the image matching method of dotted line antithesis as shown in Figure 1, it has used two width images(but the present invention is not limited to two width images), i.e. reference picture and target image, to locating of each imageReason flow process is identical, and this method comprises the following steps:
(1) from reference picture and target image, extract respectively straight line, with obtain consult straight line set andTarget line set; Particularly, in the time carrying out straight line extraction, first from image, extract edge,Then edge is cut apart and matching, thereby obtains straight line; For example, reference picture is designated as to R,Target image is designated as S, and the consult straight line set of extracting from R is designated asTotal m barStraightway, the target line set of extracting from S is designated asTotal n bar straightway.
(2) respectively with reference to straightways all in straight line set and target line set from image spaceBe mapped to dual spaces, can obtain with reference to antithesis point setTotal m dual points,And target antithesis point setTotal n dual points; Particularly, with reference to Fig. 2,Any straight line in image space is designated as to L, takes up an official post and get two point (x at L1,y1) and (x2,y2),L is mapped to dual spaces, can obtains the coordinate (θ, ρ) of its dual points p according to formula (1):
(3) concentrate contiguous dual points to merge with reference to antithesis point set and target dual points respectively,To obtain new reference antithesis point set and new target antithesis point set; As shown in Figure 3, this step comprisesFollowing sub-step:
(3-1) counter i=1 is set, counter cnt=1, state array [St1,St2,…,Stm]=1;Each corresponding with an element in state array with reference to dual points;
(3-2) judge whether i is less than or equal to m, if proceed to step (3-3), otherwise obtain newReference antithesis point setTotal u dual points, and proceed to step (3-13);
(3-3) judge StiWhether equal 1, if proceed to step (3-4), otherwise proceed to step(3-12);
(3-4) antithesis point set G to be merged being set is empty set, from concentrating and take out i with reference to dual pointsWith reference to dual pointsIts coordinate is (θi,ρi), added G, and St is seti=0;
(3-5) counter j=1 is set;
(3-6) judge whether j is less than or equal to m, if proceed to step (3-7), otherwise proceed to stepSuddenly (3-11);
(3-7) judge whether j equals i, or StjWhether equal 0, if enter step (3-10),Otherwise directly enter step (3-8);
(3-8) judgement | θj-θi|<TθAnd | ρj-ρi|<TρWhether set up, enter step (3-9) if set up,Otherwise enter step (3-10), wherein TθAnd TρFor default threshold value, TθSpan be 0 to 20Degree, TρSpan be 0 to 50 pixel;
(3-9) by j with reference to dual pointsAdd antithesis point set G to be merged, and St is setj=0;
(3-10) j=j+1 is set, then returns to step (3-6);
(3-11) all in antithesis point set G to be merged are merged with reference to dual points, to obtainNew reference dual pointsIts coordinate is (θ ', ρ '), and cnt=cnt+1 is set; Public below concrete employingFormula:
Wherein, k is the quantity with reference to dual points in antithesis point set G to be merged; wzWeight coefficient, itsValue equal with reference to dual pointsCorresponding straightwayLength, wherein z be between 1 to k appointMeaning integer;
(3-12) i=i+1 is set, then returns to step (3-2);
(3-13) all target dual points of concentrating for target dual points, adopt and above-mentioned steps(3-1) to (3-12) identical step, to obtain new target antithesis point setTotal v dual points;
Particularly, with reference to Fig. 4, in dual spaces, with dual points p1Centered by, shown in dotted line frameIts neighborhood in have two dual points p2And p3, dual points p1,p2And p3Distance each other veryApproach, therefore these three dual points are fused into a new dual points p', correspondingly, and at image skyBetween in, with dual points p1,p2And p3Corresponding straightway L1,L2And L3Be fused into one new straightLine segment L'.
The advantage of this step is, by having solved by a fusion, the broken straightway that ruptures causesThe problem of the low and poor stability of matching efficiency.
(4) estimate the rotation transformation ginseng between new reference antithesis point set and new target antithesis point setNumber; As shown in Figure 5, this step comprises following sub-step:
(4-1) counter cti=1 is set, initializes cumulative array [A1,A2,…A180]=0;
(4-2) judge whether cti is less than or equal to u, if proceed to step (4-3), otherwise enterStep (4-9);
(4-3) counter ctj=1 is set;
(4-4) judge whether ctj is less than or equal to v, if proceed to step (4-5), otherwise enterStep (4-8);
(4-5) judge whether ctj equals cti, if enter step (4-7), otherwise directly enterEnter step (4-6);
(4-6) calculate its corresponding φ according to formula (3), and make Aφ=Aφ+1;
Wherein, [] represents to round.
(4-7) ctj=ctj+1 is set, then returns to step (4-4);
(4-8) cti=cti+1 is set, then returns to step (4-2);
(4-9) find out [A1,A2,…A180] in the element of numerical value maximum, be designated as Aφ, the φ that it is correspondingmax=φIt is exactly rotation parameter;
Particularly, in order to obtain rotation transformation parameter, need to set up a cumulative array [A1,A2,…A180],Then each point of new reference dual points being concentratedBe cti=1,2 ..., u,With each concentrated point of new target dual pointsBe ctj=1,2 ..., v,Calculate its corresponding φ and to A according to formula (3)φUpgrade, after traversal all-pair, cumulative numberThe elements A of numerical value maximum in groupφCorresponding φ is exactly the rotation transformation parameter phi of finally obtainingmax。
(5) estimate the translation transformation ginseng between new reference antithesis point set and new target antithesis point setNumber; As shown in Figure 6, this step comprises following sub-step:
(5-1) counter a=1 is set, cumulative matrixItsMiddle height is the height of target image S, and width is the width of target image S, maximum max=0,Translation parameters txmax=0,tymax=0;
(5-2) judge whether a is less than or equal to u, if proceed to step (5-3), else process knotBundle;
(5-3) counter b=1 is set;
(5-4) judge whether b is less than or equal to v, if proceed to step (5-5), otherwise enter stepSuddenly (5-18);
(5-5) counter c=1 is set;
(5-6) judge whether c is less than or equal to u, if proceed to step (5-7), otherwise enter stepSuddenly (5-17);
(5-7) judge whether c equals a, if enter step (5-13), otherwise directly enter stepSuddenly (5-8);
(5-8) counter d=1 is set;
(5-9) judge whether d is less than or equal to v, if proceed to step (5-10), otherwise enter stepSuddenly (5-13);
(5-10) judge whether d equals b, if enter step (5-12), otherwise directly enterStep (5-11);
(5-11) calculate its corresponding (t according to formula (4)x,ty), and make Btx,ty=Btx,ty+1;
Wherein, [] represents to round.
(5-12) d=d+1 is set, then returns to step (5-9);
(5-13) c=c+1 is set, then returns to step (5-6);
(5-14) find outThe element of middle numerical value maximum, is designated asBtxab,tyab, and record possible translation parameters txab,tyab;
(5-15) judge whether max is greater than Btxab,tyabIf, enter step (5-17), otherwise straightTap into into step (5-16);
(5-16) max=B is settxab,tyab,txmax=txab,tymax=tyab;
(5-17) b=b+1 is set, then returns to step (5-4);
(5-18) a=a+1 is set, then returns to step (5-2);
Particularly, in order to obtain translation transformation parametric txAnd ty, need to set up a cumulative matrixThen concentrate an optional point from new reference dual pointsConcentrate an optional point from new reference dual points againAnd hypothesisThey are corresponding.
Selected corresponding dual points pairWithAfterwards, new reference dual points are concentrated and removedOutsideEach pointBe c=1,2 ..., u and c ≠ a, and new target antithesisPoint is concentrated and is removedOutside each pointBe d=1,2 ..., v and d ≠ b,Calculate its corresponding t according to formula (4)xAnd tyAnd to Btx,tyUpgrade, after traversal all-pair,The element B of numerical value maximum in cumulative matrixtxab,tyabCorresponding (txab,tyab) be exactly one of translation transformation parameterIndividual possible estimated value.
To every a pair of dual points pair that may be correspondingWithAll carry out taking turns calculating, all can obtain oneThe element B of numerical value maximumtxab,tyabWith a possible translation transformation estimates of parameters (txab,tyab). Then,From the element of all numerical value maximums, select that of numerical value the maximum, (t of its correspondencexmax,tymax) justIt is the translation transformation parameter of finally obtaining.
Those skilled in the art will readily understand, the foregoing is only preferred embodiment of the present invention,Not in order to limit the present invention, all any amendments of doing within the spirit and principles in the present invention, etc.With replacement and improvement etc., within all should being included in protection scope of the present invention.
Claims (4)
1. the image matching method based on dotted line antithesis, is characterized in that, comprises the following steps:
(1) from reference image R and target image S, extract respectively straight line, to obtain consult straight line collectionClose and target line set, the consult straight line set of wherein extracting from R is designated asAltogetherHave m bar straightway, the target line set of extracting from S is designated asTotal n bar is straightLine segment;
(2) respectively with reference to straightways all in straight line set and target line set from image spaceBe mapped to dual spaces, to obtain with reference to antithesis point setTotal m dual points, withAnd target antithesis point setTotal n dual points;
(3) concentrate contiguous dual points to merge with reference to antithesis point set and target dual points respectively,To obtain new reference antithesis point set and new target antithesis point set;
(3-1) counter i=1 is set, counter cnt=1, state array [St1,St2,…,Stm]=1;
(3-2) judge whether i is less than or equal to m, if proceed to step (3-3), otherwise obtain newReference antithesis point setTotal u dual points, and proceed to step (3-13);
(3-3) judge StiWhether equal 1, if proceed to step (3-4), otherwise proceed to step(3-12);
(3-4) antithesis point set G to be merged being set is empty set, from concentrating and take out i with reference to dual pointsWith reference to dual pointsIts coordinate is (θi,ρi), added G, and St is seti=0;
(3-5) counter j=1 is set;
(3-6) judge whether j is less than or equal to m, if proceed to step (3-7), otherwise proceed to stepSuddenly (3-11);
(3-7) judge whether j equals i, or StjWhether equal 0, if enter step (3-10),Otherwise directly enter step (3-8);
(3-8) judgement | θj-θi|<TθAnd | ρj-ρi|<TρWhether set up, enter step (3-9) if set up,Otherwise enter step (3-10), wherein TθAnd TρFor default threshold value;
(3-9) by j with reference to dual pointsAdd antithesis point set G to be merged, and St is setj=0;
(3-10) j=j+1 is set, then returns to step (3-6);
(3-11) all in antithesis point set G to be merged are merged with reference to dual points, to obtainNew reference dual pointsIts coordinate is (θ ', ρ '), and cnt=cnt+1 is set;
(3-12) i=i+1 is set, then returns to step (3-2);
(3-13) all target dual points of concentrating for target dual points, adopt and above-mentioned steps(3-1) to (3-12) identical step, to obtain new target antithesis point setTotal v dual points;
(4) estimate the rotation transformation ginseng between new reference antithesis point set and new target antithesis point setNumber;
(5) estimate the translation transformation ginseng between new reference antithesis point set and new target antithesis point setNumber.
2. image matching method according to claim 1, is characterized in that, step (3-11)Middle with reference to dual pointsCoordinate (θ ', ρ ') specifically adopt following formula:
Wherein, k is the quantity with reference to dual points in antithesis point set G to be merged; Wz is weight coefficient, itsValue equal with reference to dual pointsCorresponding straightwayLength, wherein z be between 1 to k appointMeaning integer.
3. image matching method according to claim 1, is characterized in that, step (4) bagDraw together following sub-step:
(4-1) counter cti=1 is set, initializes cumulative array [A1,A2,…A180]=0;
(4-2) judge whether cti is less than or equal to u, if proceed to step (4-3), otherwise enterStep (4-9);
(4-3) counter ctj=1 is set;
(4-4) judge whether ctj is less than or equal to v, if proceed to step (4-5), otherwise enterStep (4-8);
(4-5) judge whether ctj equals cti, if enter step (4-7), otherwise directly enterEnter step (4-6);
(4-6) calculate its corresponding φ according to following formula, and make Aφ=Aφ+1;
Wherein, [] represents to round;
(4-7) ctj=ctj+1 is set, then returns to step (4-4);
(4-8) cti=cti+1 is set, then returns to step (4-2);
(4-9) find out [A1,A2,…A180] in the element of numerical value maximum, be designated as Aφ, the φ that it is correspondingmax=φIt is exactly rotation parameter.
4. image matching method according to claim 1, is characterized in that, step (5) bagDraw together following sub-step:
(5-1) counter a=1 is set, cumulative matrixItsMiddle height is the height of target image S, and width is the width of target image S, maximum max=0,Translation parameters txmax=0,tymax=0;
(5-2) judge whether a is less than or equal to u, if proceed to step (5-3), else process knotBundle;
(5-3) counter b=1 is set;
(5-4) judge whether b is less than or equal to v, if proceed to step (5-5), otherwise enter stepSuddenly (5-18);
(5-5) counter c=1 is set;
(5-6) judge whether c is less than or equal to u, if proceed to step (5-7), otherwise enter stepSuddenly (5-17);
(5-7) judge whether c equals a, if enter step (5-13), otherwise directly enter stepSuddenly (5-8);
(5-8) counter d=1 is set;
(5-9) judge whether d is less than or equal to v, if proceed to step (5-10), otherwise enter stepSuddenly (5-13);
(5-10) judge whether d equals b, if enter step (5-12), otherwise directly enterStep (5-11);
(5-11) calculate its corresponding (t according to following formulax,ty), and make Btx,ty=Btx,ty+1;
Wherein, [] represents to round;
(5-12) d=d+1 is set, then returns to step (5-9);
(5-13) c=c+1 is set, then returns to step (5-6);
(5-14) find outThe element of middle numerical value maximum, is designated asBtxab,tyab, and record possible translation parameters txab,tyab;
(5-15) judge whether max is greater than Btxab,tyabIf, enter step (5-17), otherwise straightTap into into step (5-16);
(5-16) max=B is settxab,tyab,txmax=txab,tymax=tyab;
(5-17) b=b+1 is set, then returns to step (5-4);
(5-18) a=a+1 is set, then returns to step (5-2).
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201310268597.3A CN103345642B (en) | 2013-06-28 | 2013-06-28 | A kind of image matching method based on dotted line antithesis |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201310268597.3A CN103345642B (en) | 2013-06-28 | 2013-06-28 | A kind of image matching method based on dotted line antithesis |
Publications (2)
Publication Number | Publication Date |
---|---|
CN103345642A CN103345642A (en) | 2013-10-09 |
CN103345642B true CN103345642B (en) | 2016-05-25 |
Family
ID=49280437
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201310268597.3A Expired - Fee Related CN103345642B (en) | 2013-06-28 | 2013-06-28 | A kind of image matching method based on dotted line antithesis |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN103345642B (en) |
Families Citing this family (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN104537663B (en) * | 2014-12-26 | 2018-01-02 | 广东中科遥感技术有限公司 | A kind of method for quickly correcting of flating |
CN105957074B (en) * | 2016-04-27 | 2018-12-14 | 深圳积木易搭科技技术有限公司 | Line match method and system based on the description of V-type intersection point and local homography matrix |
Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP4205760B1 (en) * | 2007-12-27 | 2009-01-07 | 株式会社ファースト | Image matching method, program and application apparatus |
CN100590658C (en) * | 2008-07-11 | 2010-02-17 | 北京航空航天大学 | Method for matching two dimensional object point and image point with bilateral constraints |
CN102385750A (en) * | 2011-06-22 | 2012-03-21 | 清华大学 | Line matching method and line matching system on basis of geometrical relationship |
Family Cites Families (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP2003067739A (en) * | 2001-08-28 | 2003-03-07 | Monolith Co Ltd | Method and device for image transformation |
-
2013
- 2013-06-28 CN CN201310268597.3A patent/CN103345642B/en not_active Expired - Fee Related
Patent Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP4205760B1 (en) * | 2007-12-27 | 2009-01-07 | 株式会社ファースト | Image matching method, program and application apparatus |
CN100590658C (en) * | 2008-07-11 | 2010-02-17 | 北京航空航天大学 | Method for matching two dimensional object point and image point with bilateral constraints |
CN102385750A (en) * | 2011-06-22 | 2012-03-21 | 清华大学 | Line matching method and line matching system on basis of geometrical relationship |
Non-Patent Citations (2)
Title |
---|
MSLD: a robust descriptor for line matching;J. Matas et al.;《Pattern Recognition》;20090531;第42卷(第3期);941-953 * |
Robust wide-baseline stereo from maximally stable extremal regions;J. Matas et al.;《Image and Vision Computing》;20041031;第22卷(第10期);761-767 * |
Also Published As
Publication number | Publication date |
---|---|
CN103345642A (en) | 2013-10-09 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN104169965B (en) | For system, the method and computer program product adjusted during the operation of anamorphose parameter in more filming apparatus systems | |
CN103106688B (en) | Based on the indoor method for reconstructing three-dimensional scene of double-deck method for registering | |
Deserno | Fluid lipid membranes: From differential geometry to curvature stresses | |
US8743119B2 (en) | Model-based face image super-resolution | |
CN101877143B (en) | Three-dimensional scene reconstruction method of two-dimensional image group | |
US20180137611A1 (en) | Novel View Synthesis Using Deep Convolutional Neural Networks | |
CN103226822B (en) | Medical imaging joining method | |
EP3131060A1 (en) | Method and apparatus for constructing three dimensional model of object | |
CN110246147A (en) | Vision inertia odometer method, vision inertia mileage counter device and mobile device | |
CN103971378B (en) | A kind of mix the three-dimensional rebuilding method of panoramic picture in visual system | |
CN103460705B (en) | The real-time deep utilizing stereoscopic correspondence extracts | |
CN103761721B (en) | One is applicable to space rope system machine human stereo vision fast image splicing method | |
CN104616247B (en) | A kind of method for map splicing of being taken photo by plane based on super-pixel SIFT | |
CN102750697A (en) | Parameter calibration method and device | |
EP4174786A1 (en) | High-precision map generation method and apparatus, and device and computer storage medium | |
CN104850847B (en) | Image optimization system and method with automatic thin face function | |
CN108470324A (en) | A kind of binocular stereo image joining method of robust | |
CN104867106B (en) | A kind of depth map super-resolution method | |
EP2423870B1 (en) | Image correcting device, method for creating corrected image, correction table creating device, method for creating correction table, program for creating correction table, and program for creating corrected image | |
CN106296825A (en) | A kind of bionic three-dimensional information generating system and method | |
CN103955954A (en) | Reconstruction method for high-resolution depth image in combination with space diagram pairs of same scene | |
CN107895377A (en) | A kind of foreground target extracting method, device, equipment and storage medium | |
CN103345642B (en) | A kind of image matching method based on dotted line antithesis | |
CN105957005B (en) | Bridge image splicing method based on characteristic point and structure lines | |
CN103260008B (en) | A kind of image position is to the projection conversion method of physical location |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
C06 | Publication | ||
PB01 | Publication | ||
C10 | Entry into substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
C14 | Grant of patent or utility model | ||
GR01 | Patent grant | ||
CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20160525 Termination date: 20170628 |
|
CF01 | Termination of patent right due to non-payment of annual fee |