CN104392453A - Method of matching and optimizing ransac features based on polar-line insertion image - Google Patents
Method of matching and optimizing ransac features based on polar-line insertion image Download PDFInfo
- Publication number
- CN104392453A CN104392453A CN201410720152.9A CN201410720152A CN104392453A CN 104392453 A CN104392453 A CN 104392453A CN 201410720152 A CN201410720152 A CN 201410720152A CN 104392453 A CN104392453 A CN 104392453A
- Authority
- CN
- China
- Prior art keywords
- image
- polar curve
- matching
- ransac
- polar
- 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.)
- Granted
Links
- 238000000034 method Methods 0.000 title claims abstract description 63
- 238000003780 insertion Methods 0.000 title abstract description 4
- 230000037431 insertion Effects 0.000 title abstract description 4
- 239000011159 matrix material Substances 0.000 claims abstract description 24
- 238000012360 testing method Methods 0.000 claims abstract description 8
- 238000012545 processing Methods 0.000 claims abstract description 4
- 230000008878 coupling Effects 0.000 claims description 52
- 238000010168 coupling process Methods 0.000 claims description 52
- 238000005859 coupling reaction Methods 0.000 claims description 52
- 230000008569 process Effects 0.000 claims description 25
- 238000005457 optimization Methods 0.000 claims description 15
- 238000005070 sampling Methods 0.000 claims description 4
- 230000006978 adaptation Effects 0.000 claims description 3
- 238000001514 detection method Methods 0.000 claims description 3
- HUTDUHSNJYTCAR-UHFFFAOYSA-N ancymidol Chemical compound C1=CC(OC)=CC=C1C(O)(C=1C=NC=NC=1)C1CC1 HUTDUHSNJYTCAR-UHFFFAOYSA-N 0.000 claims description 2
- 239000000284 extract Substances 0.000 claims description 2
- 239000012141 concentrate Substances 0.000 claims 1
- 230000000694 effects Effects 0.000 abstract description 3
- 238000012216 screening Methods 0.000 abstract description 2
- 238000010586 diagram Methods 0.000 description 4
- 238000005516 engineering process Methods 0.000 description 4
- 238000012549 training Methods 0.000 description 2
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000004364 calculation method Methods 0.000 description 1
- 230000000052 comparative effect Effects 0.000 description 1
- 230000013011 mating Effects 0.000 description 1
- 230000000717 retained effect Effects 0.000 description 1
- 230000011218 segmentation Effects 0.000 description 1
- 239000007787 solid Substances 0.000 description 1
- 238000009827 uniform distribution Methods 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/30—Determination of transform parameters for the alignment of images, i.e. image registration
Landscapes
- Engineering & Computer Science (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Image Analysis (AREA)
Abstract
The invention discloses a method of matching and optimizing ransac features based on a polar-line insertion image. The method has the advantages of improving matching quality and increasing excellent matching quantity of feature points and solving problems that the image feature points of ransac algorithm has few excellent matching and low accuracy. The method comprises the steps: firstly, detecting, describing and matching feature points of the image to be matched, primarily screening matching collections according to rate and symmetry testing to remove error matching; secondly, utilizing the ransac algorithm to obtain an original excellent matching collection and a supported basic matrix; thirdly, utilizing the basic matrix to calculate the polar-lines of the feature matching points and choosing the obtained polar-lines so as to guarantee that the polar-lines are uniformly distributed on the image as possible; fourthly, processing the thickness and the quantity of the chosen polar-lines and then inserting into the image; and finally, obtaining excellent matching to the processed image based on the ransac algorithm. The method has good effect and is applicable to various images.
Description
Technical field
The present invention relates to computer image processing technology field, particularly a kind of ransac Feature Points Matching optimization method inserting image based on polar curve.
Background technology
In computer vision, the concept of images match is widely used in object identification, and vision is followed the tracks of, the problems such as three-dimensional reconstruction.It depends on such idea, namely some special point in property detector (the present invention uses surf property detector) detected image is first utilized, again they are described, finally utilize the content of description to mate unique point, thus realize mating between image with image.
Images match can be roughly divided into the coupling based on pixel grey scale and the coupling based on characteristics of image, and wherein characteristics of image is divided into again provincial characteristics, edge feature and point patterns.The basic demand of the images match that the image matching technology of distinguished point based can reach effectively, quick and robustness is high, is therefore widely used.At present, distinguished point based (ORB, BRISK, FAST, SURF, SIFT etc.) image matching technology be applied between different images, because these images exist yardstick, the anglec of rotation, brightness, block and the change such as mirror-reflection, therefore, after preliminary matches, there is a large amount of poor quality couplings.The inaccurate of matching result will cause the problems such as such as BREAK TRACK, image mosaic deformity, three-dimensional reconstruction weak effect.Introduce stochastic sampling unification algorism (that is: ransac) in the above-described techniques, can more reliably matching image feature, improve quality of match.Its basic thought is that to choose some couplings randomly right, calculates their corresponding relation according to the polarity constraint of Double-visual angle.Afterwards, in coupling set, remaining coupling subset is all used for supporting such relation, the final coupling set obtaining this relation of maximum support.Ransac algorithm has the characteristic supporting that the more large possibility that is more accurate, acquisition correct result of set is larger.But the number of matches that the method obtains is on the low side, application demand can not be met.And the present invention can solve problem above well.
Summary of the invention
The object of the invention is to solve the problem that ransac algorithm image characteristic point high-quality is mated less, accuracy is not high.Need, containing more this requirement of multielement in Matching supporting set, to calculate ratio juris according to ransac, propose a kind of Feature Points Matching optimization method for ransac algorithm.The high-quality this method increased based on surf unique point mates right quantity, be applicable to various view data, such as: to large scene, part, Noise image etc., the result obtained by the method can be applicable to three-dimensional reconstruction, target following, in the technical fields such as recognition of face.
The present invention solves the technical scheme that its technical matters takes: a kind of ransac characteristic matching optimization method inserting image based on polar curve.First the method utilizes ransac algorithm to obtain basis matrix, on this basis feature based match point, obtains accurate polar curve set by polarity geometric relationship.Then from polar curve set, suitable polar curve is selected.The standard that polar curve is selected is that the polar curve picked out can be uniformly distributed as much as possible on two width figure to be matched.The polar curve selected is inserted in image, it is different from other parts that the pixel of line region on image is set.Finally treat matching image and re-start surf feature point detection, describe, arest neighbors mates, ratio testing, symmetrical test, obtains the set of more high-quality coupling through ransac algorithm.
In above process, the principle of surf characteristic matching be according to surf Feature Descriptor between Euclidean distance express; It is random choose 8 coupling (that is: calculating the quantity of basis matrix) in multiple times that ransac calculates ratio juris, calculate their basis matrix, in set, remaining set of matches is used for supporting this basis matrix, finally retain the basis matrix with maximum Matching supporting set, and return its Matching supporting set; The polarity geometrical principle of Double-visual angle is that the unique point in present image forms straight line through space projection in correspondence image, and this straight line is exactly the polar curve of current signature point.
Method flow:
Step 1: read two images (image 1 and image 2) to be matched, obtain the initial matching set of two width images to be matched;
Step 1-1: the unique point detecting two width images with surf property detector respectively;
Step 1-2: the descriptor calculating unique point on two width images with surf describer respectively;
Step 1-3: utilize adaptation to carry out bi-directional matching to descriptor, finds each unique point of image 1 to two optimum matching of image 2, finds each unique point two optimum matching in the image 1 in image 2;
Step 1-4: ratio testing, (that is: image 1 is to the coupling set of image 2 to process two coupling set respectively, and image 2 is to the coupling set of image 1), calculate the distance ratio that Optimum Matching is mated with suboptimum, remove the coupling that ratio is greater than given threshold value;
Step 1-5: symmetry is tested, when the index value in two coupling set is symmetrical, extracts this coupling set, removes the set of asymmetric coupling, return symmetrical coupling set;
Step 2: use ransac stochastic sampling unification algorism, calculates the maximum basis matrix supporting coupling set, returns the high-quality coupling set meeting this basis matrix and the basis matrix supporting this characteristic matching collection;
Step 3: utilize the basis matrix that former algorithm obtains, calculates the polar curve of match point in correspondence image;
Step 4: in polar curve set obtain can on image equally distributed polar curve;
Step 4-1: calculate limit, judges that limit is in image or outside image;
Step 4-2: limit, when image is outer, according to the intersection point relation of polar curve and image border, is chosen 1,2,3 or 4 polar curve respectively and made a search;
Step 4-3: when limit is in image, according to the angled relationships between polar curve, chooses 1,2,3 or 4 polar curve respectively and makes a search;
Step 5: utilize the polar curve selected to process image;
Step 5-1: 1,2,3 or 4 polar curve is inserted on image 1 to be matched and image 2;
Step 5-2: the image pixel at polar curve place is arranged to 0, and the thickness of polar curve is set to the different-thickness of 1,5,10 or 15 pixel respectively;
Step 6: the calculating image after process being re-started to high-quality set of matches;
Step 6-1: carry out above-mentioned steps 1 operating process, obtains initial matching set;
Step 6-2: carry out above-mentioned steps 2 operating process, obtains the set of high-quality coupling.
Beneficial effect:
1, the present invention can obtain more, that quality is more excellent, accuracy is higher coupling set.
2, the present invention can be applicable to any image.
3, when the present invention selects polar curve set, using simple step to pick out can the polar curve of near uniform distribution on image; No matter limit is in image or outside image, and whether polar curve exists, and whether polar curve is uniformly distributed, and can carry out good screening to it; Be different from former ransac algorithm, the present invention is directed to the polar curve of varying number, at polar curve place to different-thickness region, can both analyze and obtain the set of high-quality coupling.
Accompanying drawing explanation
Fig. 1 is method flow diagram of the present invention.
Wherein, dotted arrow represents the former algorithm flow comprising part; Solid arrow represents the flow process that new algorithm newly increases.
Fig. 2 is algorithm details process flow diagram involved in the present invention.
Fig. 3 is the details process flow diagram of the process flow diagram that proposes of the present invention about polar curve process.
Embodiment
The optimization method realized for a better understanding of the present invention, below in conjunction with accompanying drawing, is further described the specific embodiment of the present invention.The language that the example implemented uses in the de-scription does not cause limiting to the claimed invention.
Embodiment one
As shown in Figure 1, applying the present invention, to carry out the idiographic flow of characteristics of image Optimized Matching as follows:
Step 1: ask basis matrix, obtain the high-quality set of matches of former algorithm, details flow process is as shown in Figure 2;
Step 1-1: input two width image image1 to be matched, image2;
Step 1-2: the key point keypoints detecting two width images respectively with surf detecting device, use surf describer to calculate the 64 dimension descriptors of keypoints;
Step 1-3: utilize adaptation to carry out bi-directional matching to descriptor.That is: find each unique point of image1 to two optimum matching of image2, then find two optimum matching of each unique point in image1 in image2.
Step 1-4: carry out ratio test.That is: process two coupling set respectively, to two matching values existed in each coupling set, calculate the distance ratio of two values, when ratio is greater than rate threshold ratioTH (0.5f), delete this coupling, the coupling being less than ratioTH retains;
Step 1-5: carry out Symmetry Detection.Namely two set will meet: the current matching value of (1) each coupling set has two; Reference key value and another of (2) coupling set mate gather in corresponding training index value equal, while one mate the training index value gathered and another mate gather in corresponding reference key value equal.The coupling that condition all meets is retained, as matched well collection;
Step 1-6: use stochastic sampling unification algorism ransac to calculate.Repeatedly random choose 8 matching primitives basis matrixs in coupling set, once calculate basis matrix, the polarity corresponding with matrix constraint is all tested by all couplings remaining in set, the maximum support set that basis matrix obtains namely as final high-quality set of matches, and returns this basis matrix;
Step 2: as shown in Figure 3, computed image Feature point correspondence polar curve, selects the polar curve of varying number to the treatment of details flow process that polar curve is selected respectively, and these polar curves require to be uniformly distributed on image as far as possible;
Step 2-1: calculate polar curve.The basis matrix of trying to achieve according to step 1 and Feature Points Matching set, calculate the polar curve polars0 of match point on image image2 on image1 with formula (1), the polar curve polars1 of the match point on image2 on image image1;
The expression equation of polar curve as shown in the formula:
polars[n][0]·x+polars[n][1]·y+polars[n][2]＝0 (1)
Wherein y=image.rows, x=image.cols, n are polar curve quantity.
Step 2-2: calculate polar curve limit, the intersection point at polar curve and edge;
Polar curve limit (x0, y0) is calculated, wherein lines [j]=polars0 [0], lines [k]=polars0 with formula (2)] [1].
The intersection point of polar curve and image border x=0 is asked with formula (3).
The intersection point with edge y=0 is calculated with formula (4).
Step 2-3: if limit x0>0 and y0>0, then limit is in image, otherwise limit is outside polar curve;
Step 2-4: limit, when image is outer, judges polar curve quantity i (1,2,3,4), does following selection respectively: during i=1 to the polar curve of two width images, output line s_line [0] in all polar curves with the intersection point y at edge
l(or x
c) near 1/2 edge; During i=2, output line s_line [0] in all polar curves with the intersection point y at edge
l(or x
c) at the polar curve near 1/3 edge, s_line [1] is 2 times of distant places close to s_line [0]; During i=3, output line s_line [0] in all polar curves with the intersection point y at edge
l(or x
c) near 1/4 edge, s_line [1] is 2 times of distant places close to s_line [0], and s_line [2] is 3/2 times of distant place close to s_line [1]; During i=4, output line s_line [0] in all polar curves with the intersection point y at edge
l(or x
c) near 1/5 edge, s_line [1] is 2 times of distant places close to s_line [0], and s_line [2] is 3/2 times of distant place close to s_line [1], and s_line [3] is 4/3 times of distant place close to s_line [1];
Step 2-5: when limit is in image, judges polar curve quantity i (1,2,3,4), the polar curve angle calcu-lation formula according to formula (5) does following selection to the polar curve of two width images respectively: during i=1, output line s_line [0] in all polar curves with the intersection point y at edge
l(or x
c) near 1/2 edge; During i=2, output line s_line [0] in all polar curves with the intersection point y at edge
l(or x
c) at the polar curve near 1/2 edge, s_line [1] is with the angle of s_line [0] nearly 90 °; During i=3, output line s_line [0] in all polar curves with the intersection point y at edge
l(or x
c) at the polar curve near 1/2 edge, s_line [1] is with the angle of s_line [0] nearly 60 °, s_line [2] is with the angle of s_line [0] nearly 120 °; During i=4, output line s_line [0] in all polar curves with the intersection point y at edge
l(or x
c) at the polar curve near 1/2 edge, s_line [1] is with the angle of s_line [0] nearly 45 °, s_line [2] is with the angle of s_line [0] nearly 90 °, and s_line [3] is with the angle of s_line [2] nearly 135 °;
Step 3: image is handled as follows with the polar curve chosen:
Step 3-1: polar curve is inserted image, is set to 0 for pixel image being in polar curve place, operates the polar curve of different-thickness.When polar curve thickness is 1 pixel, the pixel value on polar curve is 0; When polar curve thickness is 5 pixels, 5 pixel values up and down comprising polar curve are set to 0; When polar curve thickness is 10 pixels, 10 pixel values up and down comprising polar curve are set to 0;
Step 4: the operation image after process being re-started to step 1, obtains final high-quality coupling set;
So far, the present invention completes the images match optimization method based on polar curve segmentation, obtains the set of more high-quality coupling.
Superiority for a more clear understanding of the present invention, concrete steps in conjunction with the embodiments, below list the present invention and existing former method uses one group of image log according to the comparative result in Image Feature Point Matching.
Embodiment two
When fixing polar curve quantity is 2, the present invention comprised with comparing of former method:
Result shown in table 1 represents to be inserted after 2 polar curves in the picture, when polar curve thickness is 1,5,10,15 or 20 pixel, the unique point quantity detected and final high-quality number of matches (thickness is the primal algorithm that polar curve is not introduced in 0 expression).
As shown in Table 1, in the primal algorithm situation not introducing polar curve, the feature of Fig. 1 and Fig. 2 is counted and is respectively 1478 and 1452, and coupling number is 15.When inserting polar curve, along with the increase of thickness, unique point and coupling number all present the trend first increasing and reduce.When thickness is 10, high-quality coupling can reach 22, improves 47% than primal algorithm.
Table 1: when insertion polar curve thickness is different, high-quality matching result compares
Note: it is 2 that data fix the quantity inserting polar curve; Thickness is the polar curve thickness inserted; Thickness=0 represents the matching result of former method, does not namely insert polar curve, directly operates entire image.
Embodiment three
When fixing polar curve thickness is 5 pixels, the present invention comprised with comparing of former method:
Table 2 is the comparison of matching result under different situations when fixing polar curve thickness is 5 pixels.Can find out, high-quality that former method (number=0) obtains coupling is 15, and between polar curve number of the present invention is 1 to 4, and when to insert polar curve number be 4, effect is best, and the high-quality coupling obtained can reach 23, improves 53%.
Table 2: when the quantity of insertion polar curve is different, high-quality matching result compares
Note: it is 5 pixels that data fix the thickness inserting polar curve; Number is the polar curve quantity inserted; Number=0 represents the matching result of former method, does not namely insert polar curve, directly operates entire image.
From above-mentioned table 1 and table 2: by the introducing of polar curve of the present invention, when choosing suitable polar curve quantity, and during suitable polar curve thickness, the result significantly improved than primal algorithm can be obtained.
Claims (9)
1. insert a ransac characteristic matching optimization method for image based on polar curve, it is characterized in that, described method comprises step as described below:
Step 1: read two images to be matched, that is: image 1 and image 2, obtains the initial matching set of two width images to be matched;
Step 2: use ransac stochastic sampling unification algorism, calculates the maximum basis matrix supporting coupling set, returns the high-quality coupling set meeting this basis matrix and the basis matrix supporting this characteristic matching collection;
Step 3: utilize the basis matrix that former algorithm obtains, calculates the polar curve of match point in correspondence image;
Step 4: in polar curve set obtain can on image equally distributed polar curve;
Step 5: utilize the polar curve selected to process image;
Step 6: the calculating image after process being re-started to high-quality set of matches.
2. a kind of ransac characteristic matching optimization method inserting image based on polar curve according to claim 1, it is characterized in that, described step 1 is the process to Image Acquisition initial matching, comprises the steps:
Step 1-1: the unique point detecting two width images with surf property detector respectively;
Step 1-2: the descriptor calculating unique point on two width images with surf describer respectively;
Step 1-3: utilize adaptation to carry out bi-directional matching to descriptor, finds each unique point of image 1 to two optimum matching of image 2, finds each unique point two optimum matching in the image 1 in image 2;
Step 1-4: ratio testing, process two coupling set respectively, that is: image 1 is to the coupling set of image 2, and image 2 is to the coupling set of image 1, calculates the distance ratio that Optimum Matching is mated with suboptimum, removes the coupling that ratio is greater than given threshold value;
Step 1-5: symmetry is tested, when the index value in two coupling set is symmetrical, extracts this coupling set, removes the set of asymmetric coupling, return symmetrical coupling set.
3. a kind of ransac characteristic matching optimization method inserting image based on polar curve according to claim 1, is characterized in that, described step 2 obtains the image characteristic point high-quality coupling set of former algorithm and the process of basis matrix.
4. a kind of ransac characteristic matching optimization method inserting image based on polar curve according to claim 1, is characterized in that, described step 3 is the acquisition processs utilizing basis matrix to calculate the polar curve collection of match point in correspondence image.
5. according to claim 1ly a kind ofly insert the ransac characteristic matching optimization method of image based on polar curve, it is characterized in that, described step 4 concentrates at polar curve to select suitable polar curve and can be evenly distributed on image, and its process comprises the steps:
Step 4-1: calculate limit, judges that limit is in image or outside image;
Step 4-2: limit, when image is outer, according to the intersection point relation of polar curve and image border, is chosen 1,2,3 or 4 polar curve respectively and made a search;
Step 4-3: when limit is in image, according to the angled relationships between polar curve, chooses 1,2,3 or 4 polar curve respectively and makes a search.
6. a kind of ransac characteristic matching optimization method inserting image based on polar curve according to claim 1, is characterized in that, described step 5 inserts in image to the polar curve selected, the processing procedure following steps of its polar curve:
Step 5-1: 1,2,3 or 4 polar curve is inserted on image to be matched;
Step 5-2: the image pixel at polar curve place is arranged to 0, and the thickness of polar curve is set to the different-thickness of 1,5,10 or 15 pixel respectively.
7. a kind of ransac characteristic matching optimization method inserting image based on polar curve according to claim 1, is characterized in that, described step 6 is the processes image processed being regained to the set of high-quality coupling, and the step of its processing procedure is as follows:
Step 6-1: carry out above-mentioned steps 1 operating process, obtains initial matching set;
Step 6-2: carry out above-mentioned steps 2 operating process, obtains the set of high-quality coupling.
8. a kind of ransac characteristic matching optimization method inserting image based on polar curve according to claim 1, it is characterized in that, described method is applied to three-dimensional reconstruction, target following, in technical field of face recognition.
9. a kind of ransac characteristic matching optimization method inserting image based on polar curve according to claim 1, it is characterized in that, first described method utilizes ransac algorithm to obtain basis matrix, on this basis feature based Point matching, obtains accurate polar curve set by polarity geometric relationship; Then from polar curve set, select suitable polar curve, the standard that polar curve is selected is that the polar curve picked out can be uniformly distributed as much as possible on two width figure to be matched; The polar curve selected is inserted in image, it is different from other parts that the pixel of line region on image is set; Finally treat matching image and re-start surf feature point detection, describe, arest neighbors mates, ratio testing, symmetrical test, obtains the set of more high-quality coupling through ransac algorithm.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201410720152.9A CN104392453B (en) | 2014-12-01 | 2014-12-01 | The ransac characteristic matching optimization methods of image are inserted based on polar curve |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201410720152.9A CN104392453B (en) | 2014-12-01 | 2014-12-01 | The ransac characteristic matching optimization methods of image are inserted based on polar curve |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN104392453A true CN104392453A (en) | 2015-03-04 |
| CN104392453B CN104392453B (en) | 2017-09-29 |
Family
ID=52610351
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201410720152.9A Expired - Fee Related CN104392453B (en) | 2014-12-01 | 2014-12-01 | The ransac characteristic matching optimization methods of image are inserted based on polar curve |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN104392453B (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106815824A (en) * | 2016-12-08 | 2017-06-09 | 华中科技大学 | A kind of image neighbour's optimization method for improving extensive three-dimensional reconstruction efficiency |
| CN113139986A (en) * | 2021-04-30 | 2021-07-20 | 东风越野车有限公司 | Integrated environment perception and multi-target tracking system |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US8401281B1 (en) * | 2003-10-20 | 2013-03-19 | Open Invention Network, Llc | Method and system for three dimensional feature attribution through synergy of rational polynomial coefficients and projective geometry |
| CN104021542A (en) * | 2014-04-08 | 2014-09-03 | 苏州科技学院 | Optimization method of SIFT characteristic matching points based on limit restraint |
-
2014
- 2014-12-01 CN CN201410720152.9A patent/CN104392453B/en not_active Expired - Fee Related
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US8401281B1 (en) * | 2003-10-20 | 2013-03-19 | Open Invention Network, Llc | Method and system for three dimensional feature attribution through synergy of rational polynomial coefficients and projective geometry |
| CN104021542A (en) * | 2014-04-08 | 2014-09-03 | 苏州科技学院 | Optimization method of SIFT characteristic matching points based on limit restraint |
Non-Patent Citations (2)
| Title |
|---|
| 张培耘等: ""基于ＲANSAC算法的极线约束立体视觉匹配方法研究"", 《组合机床与自动化加工技术》 * |
| 李志: ""基于双目视觉的图像匹配算法研究"", 《中国优秀硕士学位论文全文数据库 信息科技辑》 * |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106815824A (en) * | 2016-12-08 | 2017-06-09 | 华中科技大学 | A kind of image neighbour's optimization method for improving extensive three-dimensional reconstruction efficiency |
| CN106815824B (en) * | 2016-12-08 | 2019-07-05 | 华中科技大学 | A kind of image neighbour's optimization method improving extensive three-dimensional reconstruction efficiency |
| CN113139986A (en) * | 2021-04-30 | 2021-07-20 | 东风越野车有限公司 | Integrated environment perception and multi-target tracking system |
Also Published As
| Publication number | Publication date |
|---|---|
| CN104392453B (en) | 2017-09-29 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN104182974B (en) | A speeded up method of executing image matching based on feature points | |
| US20160086048A1 (en) | Device and Method for Analyzing the Correlation Between an Image and Another Image or Between an Image and a Video | |
| CN109308460B (en) | Article detection method, system and computer readable storage medium | |
| JP5711387B2 (en) | Method and apparatus for comparing pictures | |
| CN103336957B (en) | A kind of network homology video detecting method based on space-time characteristic | |
| Wang et al. | Guided image contrast enhancement based on retrieved images in cloud | |
| US8983199B2 (en) | Apparatus and method for generating image feature data | |
| JP6202147B2 (en) | Curve detection method and curve detection apparatus | |
| CN106055704A (en) | Image retrieving and matching method and system | |
| US10699156B2 (en) | Method and a device for image matching | |
| CN111027343B (en) | Bar code region positioning method and device | |
| Priya et al. | Edge strength extraction using orthogonal vectors for shot boundary detection | |
| US9323989B2 (en) | Tracking device | |
| CN109919232B (en) | Image classification method based on convolutional neural network and non-local connection network | |
| CN103914819B (en) | A kind of based on the infrared image joining method improving RANSAC | |
| Pok et al. | Efficient block matching for removing impulse noise | |
| Katkar et al. | A novel approach for medical image segmentation using PCA and K-means clustering | |
| CN104392453A (en) | Method of matching and optimizing ransac features based on polar-line insertion image | |
| CN104484679A (en) | Non-standard gun shooting bullet trace image automatic identification method | |
| CN117496109A (en) | Image comparison and analysis method and device, electronic equipment and storage medium | |
| US11023781B2 (en) | Method, apparatus and device for evaluating image tracking effectiveness and readable storage medium | |
| CN109544614B (en) | Method for identifying matched image pair based on image low-frequency information similarity | |
| Liu et al. | Depth-aware layered edge for object proposal | |
| Shivapuja et al. | Wisdom of (binned) crowds: A bayesian stratification paradigm for crowd counting | |
| CN103198294B (en) | The recognition methods of a kind of video or image viewpoint type |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant | ||
| CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20170929 |