CN110378940B - Aviation image feature point matching diffusion recursive calibration method - Google Patents
Aviation image feature point matching diffusion recursive calibration method Download PDFInfo
- Publication number
- CN110378940B CN110378940B CN201910521634.4A CN201910521634A CN110378940B CN 110378940 B CN110378940 B CN 110378940B CN 201910521634 A CN201910521634 A CN 201910521634A CN 110378940 B CN110378940 B CN 110378940B
- Authority
- CN
- China
- Prior art keywords
- density
- image
- matching
- reference image
- units
- 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
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/22—Matching criteria, e.g. proximity measures
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/10—Segmentation; Edge detection
- G06T7/11—Region-based segmentation
-
- 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
- G06T7/33—Determination of transform parameters for the alignment of images, i.e. image registration using feature-based methods
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- Life Sciences & Earth Sciences (AREA)
- Artificial Intelligence (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Bioinformatics & Computational Biology (AREA)
- Evolutionary Biology (AREA)
- Evolutionary Computation (AREA)
- General Engineering & Computer Science (AREA)
- Image Analysis (AREA)
Abstract
The invention discloses a method for matching diffusion recursive calibration of aviation image feature points, which comprises the following steps: s1: dividing the reference image and the matching image into density units respectively; s2: performing the following operations on both the reference image and the matching image: setting a threshold value n according to the number of the feature points in the density units, marking the density units with the number of the feature points being more than or equal to n as high-density units, and marking other density units as low-density units; s3: performing the following operations on both the reference image and the matching image: extracting the communicated high-density units to obtain a high-density area of the aerial image; s4: performing the following operations on both the reference image and the matching image: position marking is carried out on all the high-density areas; s5: and matching the high-density areas of the reference image and the matching image. The invention effectively improves the anti-interference capability and efficiency.
Description
Technical Field
The invention relates to the field of aerial photogrammetry, in particular to an aerial image feature point matching diffusion recursive calibration method.
Background
The difficulty of the aerial triangular measurement image is increased when the characteristic points are matched due to the light brightness and the influence caused by angle rotation when the aircraft shoots. In order to solve the problem, the invention provides a region calibration algorithm based on the distribution density of the feature points, so that the feature point matching algorithm has scale invariance and the matching robustness is enhanced. There are many ways to make feature point matching have scale invariance, mainly including two ways of exploring feature direction represented by Sift operator and searching image overall feature information represented by Mean Shift algorithm. The biggest interference in aerial images is brightness change caused by light, the Sift-like algorithm cannot overcome the brightness change well, and the Mean Shift algorithm has advantages but unstable effect.
Disclosure of Invention
The purpose of the invention is as follows: the invention aims to provide a matching diffusion recursion calibration method for aviation image feature points with strong anti-jamming capability.
The technical scheme is as follows: in order to achieve the purpose, the invention adopts the following technical scheme:
the invention relates to a method for matching diffusion recursive calibration of aviation image feature points, which comprises the following steps:
s1: dividing the reference image and the matching image into density units respectively;
s2: performing the following operations on both the reference image and the matching image: setting a threshold value n according to the number of the feature points in the density units, marking the density units with the number of the feature points being more than or equal to n as high-density units, and marking other density units as low-density units;
s3: performing the following operations on both the reference image and the matching image: extracting the communicated high-density units to obtain a high-density area of the aerial image;
s4: performing the following operations on both the reference image and the matching image: position marking is carried out on all the high-density areas;
s5: and matching the high-density areas of the reference image and the matching image.
Further, in step S1, the reference image is divided into density units according to equation (1):
in the formula (1), d is the side length of the density unit, t is the root of the number of pixels contained in the density unit, w is the side length of the detection window of the feature point extraction operator, and l is the side length of the reference image.
Further, in the step S2, the total number of the high density units does not exceed the total number of all the density units.
Further, the step S3 specifically includes the following steps:
s31: establishing an x coordinate system and a y coordinate system for the reference image and the matching image, and forming keys according to coordinates in the x coordinate system and the y coordinate system;
s32: inquiring the value corresponding to the key in the set formed by all the high-density units, namely inquiring whether an adjacent high-density unit exists, if so, continuing to perform the step S33, and if not, ending;
s33: judging whether the adjacent high-density unit has diffused before, if yes, ending, otherwise continuing to step S34;
s34: returning to step S32.
Further, the specific process of step S4 is: the minimum y-value component of all the high-density units in the high-density area is selected as the ordinate of the position mark point, and the minimum x-value component of all the high-density units in the high-density area is selected as the abscissa of the position mark point.
Further, the step S5 specifically includes the following steps:
s51: the following is performed for all high-density regions of the reference image: forming a high-density expanded area of the reference image by diffusing 40 multiplied by 40 pixel units to the periphery by taking a position mark point of the high-density area of the reference image as a center;
s52: forming a high-density expanded area matrix X of the reference image by using all the high-density expanded areas of the reference image obtained in the step S51;
s53: the following is performed for all high density regions of the matching image: taking a position mark point of a high-density area of the matching image as a center, and diffusing 40 x 40 pixel units to the periphery to form a high-density expansion area of the matching image;
s54: forming a high-density expanded area matrix Y of the matched image by using all the high-density expanded areas of the matched image obtained in the step S53; the correlation coefficient rho of X and Y is calculated according to the formula (2) X,Y ：
In the formula (2), cov (X, Y) is the covariance of X and Y, σ X Standard deviation of X, σ Y Is the standard deviation of Y, μ X Is the mean value of X, μ Y Is the mean value of Y;
s55: taking a matrix of correlation coefficients ρ X,Y The element in X and the element in Y corresponding to the largest element in all the elements are used as a group of homonymous high-density areas, and r is calculated according to the formula (3) x And r y ：
In the formula (3), r x Number of pixels, r, of horizontal displacement required to achieve alignment of the matching image with the reference image y Number of pixels, x, of vertical displacement required to achieve alignment of the matching image with the reference image l Is the abscissa, X, of the element of X in the set of homonymous high-density regions r Is the abscissa, Y, of the element of Y in the set of homonymous high-density regions l Is the ordinate, y, of the element of X in the set of homonymous high-density regions r D is the side length of the density unit in the reference image and the matching image, which is the ordinate of the elements of Y in the set of homonymous high density regions.
Has the advantages that: the invention discloses a method for matching, diffusing and recursively calibrating characteristic points of aerial images, which comprises the steps of finding out a high-density area of a reference image and a matching image, and then performing high-density area matching, thereby effectively improving the anti-interference capability; the division of the density units greatly simplifies the complexity of the method, converts all pixel points traversing the whole image into matrixes traversing a limited number, and improves the efficiency of the method.
Drawings
FIG. 1 is a flow chart of a method in accordance with an embodiment of the present invention.
Detailed Description
The technical solution of the present invention will be further described with reference to the following embodiments.
The specific embodiment discloses an aviation image feature point matching diffusion recursive calibration method, as shown in fig. 1, comprising the following steps:
s1: dividing the reference image and the matching image into density units respectively;
s2: performing the following operations on both the reference image and the matching image: setting a threshold value n according to the number of the feature points in the density units, marking the density units with the number of the feature points being more than or equal to n as high-density units, and marking other density units as low-density units; wherein the total number of the high-density units is not more than 20% of the sum of the numbers of the high-density units and the low-density units;
s3: performing the following operations on both the reference image and the matching image: extracting the communicated high-density units to obtain a high-density area of the aerial image;
s4: performing the following operations on both the reference image and the matching image: position marking is carried out on all the high-density areas;
s5: and matching the high-density areas of the reference image and the matching image.
In step S1, the reference image is divided into density units according to equation (1):
in the formula (1), d is the side length of the density unit, t is the root of the number of pixels contained in the density unit, w is the side length of the detection window of the feature point extraction operator, and l is the side length of the reference image.
The matching image is also divided into density units as the reference image, and the description thereof is omitted here.
In step S2, the total number of high density units does not exceed the total number of all density units.
The step S3 specifically includes the following steps:
s31: establishing an x coordinate system and a y coordinate system for the reference image and the matching image, and forming keys according to coordinates in the x coordinate system and the y coordinate system;
s32: inquiring the value corresponding to the key in the set formed by all the high-density units, namely inquiring whether an adjacent high-density unit exists, if so, continuing to perform the step S33, and if not, ending;
s33: judging whether the adjacent high-density unit is diffused before, if so, ending, otherwise, continuing to the step S34;
s34: the process returns to step S32.
The specific process of step S4 is: the minimum y-value component of all the high-density units in the high-density area is selected as the ordinate of the position mark point, and the minimum x-value component of all the high-density units in the high-density area is selected as the abscissa of the position mark point.
The step S5 specifically includes the following steps:
s51: the following is performed for all high-density regions of the reference image: taking a position mark point of a high-density area of the reference image as a center, and diffusing 40 x 40 pixel units to the periphery to form a high-density expanded area of the reference image;
s52: forming a high-density expanded area matrix X of the reference image by using all the high-density expanded areas of the reference image obtained in the step S51;
s53: the following is performed for all high density regions of the matching image: taking a position mark point of a high-density area of the matching image as a center, and diffusing 40 multiplied by 40 pixel units to the periphery to form a high-density expansion area of the matching image;
s54: forming a high-density expanded area matrix Y of the matched image by using all the high-density expanded areas of the matched image obtained in the step S53; the correlation coefficient rho of X and Y is calculated according to the formula (2) X,Y ：
In the formula (2), cov (X, Y) is the covariance of X and Y, σ X Is the standard deviation of X, σ Y Is the standard deviation of Y, μ X Is the mean value of X, μ Y Is the mean value of Y;
s55: taking a matrix of correlation coefficients ρ X,Y The element in X and the element in Y corresponding to the largest element in all the elements are used as a group of homonymous high-density areas, and r is calculated according to the formula (3) x And r y ：
In the formula (3), r x Number of pixels, r, of horizontal displacement required to achieve alignment of the matching image with the reference image y Number of pixels, x, of vertical displacement required to match the image to be aligned with the reference image l Of elements of X in the set of homonymous high-density regionsAbscissa, x r Is the abscissa, Y, of the element of Y in the set of homonymous high-density regions l Is the ordinate, y, of the element of X in the set of homonymous high-density regions r D is the side length of the density unit in the reference image and the matching image, which is the ordinate of the element of Y in the set of homonymous high density regions.
Claims (5)
1. The aviation image feature point matching diffusion recursive calibration method is characterized by comprising the following steps: the method comprises the following steps:
s1: dividing the reference image and the matching image into density units respectively;
s2: performing the following operations on both the reference image and the matching image: setting a threshold value n according to the number of the feature points in the density units, marking the density units with the number of the feature points being more than or equal to n as high-density units, and marking other density units as low-density units;
s3: performing the following operations on both the reference image and the matching image: extracting the communicated high-density units to obtain a high-density area of the aerial image;
s4: performing the following operations on both the reference image and the matching image: position marking is carried out on all the high-density areas;
s5: matching the high-density areas of the reference image and the matching image;
the step S5 specifically includes the following steps:
s51: the following is performed for all high-density regions of the reference image: taking a position mark point of a high-density area of the reference image as a center, and diffusing 40 x 40 pixel units to the periphery to form a high-density expanded area of the reference image;
s52: forming a high-density expanded area matrix X of the reference image by using all the high-density expanded areas of the reference image obtained in the step S51;
s53: the following is performed for all high density regions of the matching image: taking a position mark point of a high-density area of the matching image as a center, and diffusing 40 x 40 pixel units to the periphery to form a high-density expansion area of the matching image;
s54: all the high densities of the matching images obtained in step S53The degree expansion area forms a high-density expansion area matrix Y of the matched image; the correlation coefficient rho of X and Y is calculated according to the formula (2) X,Y ：
In the formula (2), cov (X, Y) is the covariance of X and Y, σ X Standard deviation of X, σ Y Standard deviation of Y,. Mu. X Is the mean value of X, μ Y Is the mean value of Y;
s55: taking a matrix of correlation coefficients ρ X,Y The element in X and the element in Y corresponding to the largest element in all the elements are used as a group of homonymous high-density areas, and r is calculated according to the formula (3) x And r y ：
In the formula (3), r x Number of pixels, r, of horizontal displacement required to achieve alignment of the matching image with the reference image y Number of pixels, x, of vertical displacement required to match the image to be aligned with the reference image l Is the abscissa, X, of the element of X in the set of homonymous high-density regions r Is the abscissa, Y, of the element of Y in the set of homonymous high-density regions l Is the ordinate, y, of the element of X in the set of homonymous high-density regions r D is the side length of the density unit in the reference image and the matching image, which is the ordinate of the element of Y in the set of homonymous high density regions.
2. The aerial image feature point matching diffusion recursive calibration method according to claim 1, characterized in that: in step S1, the reference image is divided into density units according to equation (1):
in the formula (1), d is the side length of the density unit, t is the root of the number of pixels contained in the density unit, w is the side length of the detection window of the feature point extraction operator, and l is the side length of the reference image.
3. The aerial image feature point matching diffusion recursive calibration method according to claim 1, characterized in that: in step S2, the total number of high density units does not exceed the total number of all density units.
4. The aerial image feature point matching diffusion recursive calibration method according to claim 1, characterized in that: the step S3 specifically includes the following steps:
s31: establishing an x coordinate system and a y coordinate system for the reference image and the matching image, and forming keys according to coordinates in the x coordinate system and the y coordinate system;
s32: inquiring the value corresponding to the key in the set formed by all the high-density units, namely inquiring whether an adjacent high-density unit exists, if so, continuing to perform the step S33, and if not, ending;
s33: judging whether the adjacent high-density unit is diffused before, if so, ending, otherwise, continuing to the step S34;
s34: the process returns to step S32.
5. The aerial image feature point matching diffusion recursive calibration method according to claim 1, characterized in that: the specific process of step S4 is: the minimum y-value component of all the high-density units in the high-density area is selected as the ordinate of the position mark point, and the minimum x-value component of all the high-density units in the high-density area is selected as the abscissa of the position mark point.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910521634.4A CN110378940B (en) | 2019-06-17 | 2019-06-17 | Aviation image feature point matching diffusion recursive calibration method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910521634.4A CN110378940B (en) | 2019-06-17 | 2019-06-17 | Aviation image feature point matching diffusion recursive calibration method |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN110378940A CN110378940A (en) | 2019-10-25 |
| CN110378940B true CN110378940B (en) | 2023-04-07 |
Family
ID=68248982
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910521634.4A Active CN110378940B (en) | 2019-06-17 | 2019-06-17 | Aviation image feature point matching diffusion recursive calibration method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110378940B (en) |
Families Citing this family (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112514970B (en) * | 2020-12-08 | 2021-11-05 | 烟台海裕食品有限公司 | Self-adaptive fish scale removing platform and method |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101122999A (en) * | 2007-04-16 | 2008-02-13 | 北京联合大学 | Method for automatically extracting stamper image from Chinese painting and calligraphy |
| US20120148164A1 (en) * | 2010-12-08 | 2012-06-14 | Electronics And Telecommunications Research Institute | Image matching devices and image matching methods thereof |
| CN109816051A (en) * | 2019-02-25 | 2019-05-28 | 北京石油化工学院 | A kind of harmful influence cargo characteristic point matching method and system |
-
2019
- 2019-06-17 CN CN201910521634.4A patent/CN110378940B/en active Active
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101122999A (en) * | 2007-04-16 | 2008-02-13 | 北京联合大学 | Method for automatically extracting stamper image from Chinese painting and calligraphy |
| US20120148164A1 (en) * | 2010-12-08 | 2012-06-14 | Electronics And Telecommunications Research Institute | Image matching devices and image matching methods thereof |
| CN109816051A (en) * | 2019-02-25 | 2019-05-28 | 北京石油化工学院 | A kind of harmful influence cargo characteristic point matching method and system |
Also Published As
| Publication number | Publication date |
|---|---|
| CN110378940A (en) | 2019-10-25 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN107665351B (en) | Airport detection method based on difficult sample mining | |
| CN101957919B (en) | Character recognition method based on image local feature retrieval | |
| CN107688806B (en) | Affine transformation-based free scene text detection method | |
| CN109460764B (en) | Satellite video ship monitoring method combining brightness characteristics and improved interframe difference method | |
| WO2017016448A1 (en) | Qr code feature detection method and system | |
| CN105205488A (en) | Harris angular point and stroke width based text region detection method | |
| CN107992856B (en) | High-resolution remote sensing building shadow detection method under urban scene | |
| CN111310768B (en) | Saliency target detection method based on robustness background prior and global information | |
| CN107862319B (en) | Heterogeneous high-light optical image matching error eliminating method based on neighborhood voting | |
| CN111461036B (en) | Real-time pedestrian detection method using background modeling to enhance data | |
| CN111783773B (en) | Correction method for angle-inclined telegraph pole signboard | |
| CN110852207A (en) | Blue roof building extraction method based on object-oriented image classification technology | |
| CN102004914B (en) | Method and device for detecting round seal position | |
| CN110378940B (en) | Aviation image feature point matching diffusion recursive calibration method | |
| US20210286970A1 (en) | Cloud detection method based on landsat 8 snow-containing image | |
| CN111160107B (en) | Dynamic region detection method based on feature matching | |
| CN112529901A (en) | Crack identification method in complex environment | |
| CN108830283B (en) | Image feature point matching method | |
| CN105303566A (en) | Target contour clipping-based SAR image target azimuth estimation method | |
| CN112329641B (en) | Form identification method, device, equipment and readable storage medium | |
| CN112396582A (en) | Mask RCNN-based equalizing ring skew detection method | |
| CN112215319B (en) | Two-dimensional code of color mark characteristic graph and identification method thereof | |
| CN111695557B (en) | Image processing method and device | |
| US11132557B2 (en) | Logo extraction device, and brightness adjusting method and device for logo | |
| CN110348286B (en) | Face fitting and matching method based on least square method |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |