CN106709501B - Scene matching area selection and reference image optimization method of image matching system - Google Patents
Scene matching area selection and reference image optimization method of image matching system Download PDFInfo
- Publication number
- CN106709501B CN106709501B CN201510784714.0A CN201510784714A CN106709501B CN 106709501 B CN106709501 B CN 106709501B CN 201510784714 A CN201510784714 A CN 201510784714A CN 106709501 B CN106709501 B CN 106709501B
- Authority
- CN
- China
- Prior art keywords
- matching
- image
- scene
- edge
- optimization method
- 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
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/0002—Inspection of images, e.g. flaw detection
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/10—Image acquisition modality
- G06T2207/10032—Satellite or aerial image; Remote sensing
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/30—Subject of image; Context of image processing
- G06T2207/30168—Image quality inspection
Abstract
The invention relates to a scene matching area selection and reference image optimization method of an image matching system, which is characterized in that blocks are initialized aiming at a remote sensing satellite picture, edge characteristics of each area are extracted according to an edge characteristic extraction algorithm, and block areas with concentrated edge gradient points are screened; calculating the autocorrelation degree of each region by using a repetitive pattern index measurement method, sequencing the regions from small to large according to the inner correlation index, and outputting an optimal candidate region; establishing a spatial feature vector of a reference map of the candidate region by adopting a spatial distribution description method; establishing a scene matchability measurement evaluation index set, analyzing the relevance between the evaluation index set and the matching probability, and outputting a matching evaluation index set; and statistically analyzing the matching correlation surface of the reference image and the satellite image, and optimizing the reference image according to the reference image optimization method. According to the invention, when the matching system works in an unknown environment, the optimized reference image is utilized to realize high matching performance and high reliability, and the image matching system is ensured to successfully complete tasks.
Description
Technical Field
The invention relates to the technical field of performance evaluation of an image processing system, in particular to a scene matching area selection and reference image optimization method of an image matching system.
Background
In the scene matching system, the matching process is essentially to calculate the similarity between the real-time image and the reference image, so that the quality of the reference image and the scene matching precision are closely related. The reference image is taken as advance planning information and mainly comes from remote sensing images, time and imaging system differences exist, quality of the reference image is seriously affected, and if structural characteristics of a matching area can be analyzed in a reference image preparation link, evaluation indexes and criteria based on high matching performance measurement are constructed, and the matching area is selected and the reference image is optimized according to the indexes, the method has important significance for improving performance of the scene matching system.
The scene region suitability selection and analysis is a specific process, namely a region of interest (ROI) which accords with the basic indexes of region suitability is determined essentially. In the areas, certain characteristic attributes are different from adjacent areas, a characteristic set capable of comprehensively reflecting area adaptation performance is searched, characteristic indexes are quantized, various characteristic indexes are subjected to information fusion to form comprehensive characteristic quantities, and a scene area adaptation evaluation method of various comprehensive characteristic quantities is provided, wherein the characteristic index design principle comprises the following steps:
(1) the scene matching area can be matched only when the scene matching area contains enough information, and the richer the information is, the more the successful matching is facilitated, for example, a scene matching area selecting method based on the information entropy is provided by Zhang Xiao and the like;
(2) the image quality is poor, so that the scene features of the ground objects become fuzzy or even disappear, and the matching fails, so that the characteristic indexes need to ensure that the matching area has stable characteristics, and the stable curve characteristics of the target are extracted by the method R, and the like to estimate the stable boundary of the target;
(3) reflecting the uniqueness of features in the scene, if a plurality of similar obvious objects exist in the selected matching area, the matching success probability can be greatly reduced, and the feature indexes can reflect the uniqueness of the features, such as introducing a digital map into a Caves R G, calculating a high matching area between the digital map and an imaging image, and taking the high matching area as an index for feature selection;
(4) reflecting the obvious characteristics in the scene, in order to achieve high matching precision, obviously distinguishing the matching position from the non-matching position, enabling the relevant peak to be large enough and enabling the relevant peak to be sharp enough, such as Jucinor and the like, calculating the number of suspected targets of images with different scales by utilizing wavelet domain multi-scale images, weighting all levels of images, and designing a target dominant index;
the suitability evaluation of the comprehensive characteristic quantity is mostly contained in two basic theoretical systems of multi-attribute decision and mode classification:
(1) scene region suitability evaluation based on multi-attribute decision theory
The basic idea of the evaluation method is to abstract the adaptive evaluation process into a decision-making overstatement, take each characteristic index as the basic attribute of the decision-making, construct and select functions to form comprehensive characteristic quantities through a specific decision-making model, for example, Zhao kuwei and the like adopt simple weighting of each characteristic extreme value in the area to realize the decision-making process, Cao and Country and the like use a modified D-S theory to complete multi-attribute decision, obtain each global characteristic index weighting coefficient through orthogonal experimental design, and obtain the weighting coefficient through feedback modification to realize the decision-making process.
(2) Scene region suitability evaluation based on mode classification theory
The basic idea of the evaluation method is to use each characteristic index value as perception information, design a classifier according to a pre-selected classification criterion, so that a matching probability estimation problem is converted into a classification problem of pixels or regions, such as a Mumford-Shah model proposed by Lijun and the like, optimal division of two sets of a matching stable local region and an unstable local region is obtained through level set curve evolution, and a Fisher classifier is used for designing a fault rate minimum threshold value for Yang Xin and the like to predict a classification result.
Therefore, the selection of the current reference image is mainly focused on the analysis of image gray scale information or the evaluation of characteristic quality, and does not have the relationship with matching performance to perform the prior optimization of the reference image, especially in an image matching system, most of the image matching systems work on an infrared imaging system, and for the preparation of the prior reference image of the image matching system, a homologous test image is adopted to calibrate, namely an infrared image sequence. However, in an actual unknown environment, homologous infrared image sequences are difficult to acquire, and heterologous images of visible light are very easy to acquire, such as satellite images and aerial images, and analysis errors caused by the imaging system are less concerned, so that the matching probability is reduced. Therefore, a reference image selection and evaluation method facing to a heterogeneous image needs to be established, the method can adapt to the preparation conditions of the heterogeneous image, and meanwhile, the method has the function of optimizing the reference image in advance, so that the preparation of the reference image is close to a high-quality target, and related results of the method for preparing and optimizing the reference image are not published.
Disclosure of Invention
Aiming at the defects of the prior art, the invention provides a method for establishing an internal correlation and external correlation evaluation index set, supervising the preparation process of a reference map and finishing region selection and reference map optimization by utilizing the edge feature similarity of a heterogeneous image and using a remote sensing satellite image.
The technical scheme adopted by the invention for realizing the purpose is as follows:
a scene matching area selection and reference map optimization method of an image matching system comprises the following steps:
step 1: initializing blocks aiming at a remote sensing satellite picture, extracting edge characteristics of each region according to an edge characteristic extraction algorithm, and screening the block regions with concentrated edge gradient points;
and step 3: establishing a spatial feature vector of a reference map of the candidate region by adopting a spatial distribution description method;
and 4, step 4: establishing a scene matchability measurement evaluation index set, analyzing the relevance between the evaluation index set and the matching probability, and outputting a matching evaluation index set;
and 5: and statistically analyzing the matching correlation surface of the reference image and the satellite image, and optimizing the reference image according to the reference image optimization method.
The edge feature extraction algorithm is a Canny operator edge feature extraction algorithm.
The screening of the blocking region in the edge gradient point set comprises the following processes:
step 1: extracting edge features by using a Canny operator aiming at each block subregion;
step 2: calculating the gradient change histogram distribution of each subregion;
and step 3: and comparing the distribution of the gradient point sets of each sub-region, and selecting one or more regions in the point sets as candidate regions according to the gradient change from small to large.
The repetitive pattern index measurement method comprises the following steps:
step 1: selecting a subgraph i aiming at each block area, wherein a calculation formula of a repeated mode index is as follows:
wherein s is the number of the image blocks participating in matching calculation when the subgraph i is matched point by point on the candidate region; pi is the number of image blocks with edge point correlation coefficient between s image blocks and the sub-image i larger than threshold in the matching process;
step 2: the formula for calculating the reference graph repeat pattern is:
wherein cf is the degree of pattern measurement for the entire tile; n is the number of selected sub-picture blocks.
The method for establishing the spatial feature vector of the candidate region reference map by adopting the spatial distribution description method comprises the following processes:
step 1: extracting features of the reference graph, expressing the features as direction features and scale features, and establishing a feature sensitivity analysis test;
step 2: dividing the reference image of the reference image area into intervals according to the minimum resolvable interval of the directional characteristic;
and step 3: counting the number of gauge points in the interval, and establishing the form of { X1,X2,X3,X4,X5,X6The multidimensional feature vector of.
The reference map optimization method comprises the following steps:
step 1: completing a matching process by using a Harsdoff distance matching method, and calculating a matching performance measurement index;
step 2: after judging the quality of the matching metric index, calculating a main influence interval of a main peak and a secondary peak;
and step 3: and correcting the edge number of the interval, optimizing the reference graph, and recalculating the matching metric index until the matching threshold requirement is met.
The invention has the following beneficial effects and advantages:
1. the method adopts an image characteristic analysis method, aims at image edge characteristic information, utilizes self-similarity indexes to search and evaluate a most matchable candidate area, utilizes spatial distribution description of a reference image to establish a characteristic vector and an evaluation index, and corrects and optimizes the characteristic vector by measuring a matching correlation surface of a matching system;
2. the method is particularly suitable for preparing the reference diagram of a heterogeneous matching system, improves the matching performance through the analysis and the evaluation of an internal correlation layer and an external correlation layer, establishes a supervision index system, measures the quality of the reference diagram in real time, guides and optimizes the reference diagram, and realizes the preparation of the high-quality reference diagram.
Drawings
FIG. 1 is an overall flow chart of the present invention;
FIG. 2 is a histogram of the sub-regions of the present invention;
FIG. 3 is a diagram of analysis of spatial distribution of a reference map, (a) is a diagram of analysis of directional feature sensitivity, and (b) is a diagram of analysis of scale feature sensitivity;
FIG. 4 is a graph of the matching index after the optimization by the method of the present invention.
Detailed Description
The present invention will be described in further detail with reference to the accompanying drawings and examples.
As shown in fig. 1, the scene matching area selection and reference map optimization method of the image matching system of the present invention comprises the following steps:
(1) extracting the edge characteristics of each region in a blocking manner, and screening blocking regions with stable edge gradient points;
(2) measuring the autocorrelation degree of each region by using the repetitive pattern index, sequencing the regions according to the internal correlation index, and outputting an optimal candidate region;
(3) establishing a spatial feature vector of a reference map of the candidate region by adopting a spatial distribution description method;
(4) establishing a scene matchability measurement evaluation index set, and analyzing the relevance between the index set and the matching probability;
(5) and (4) statistically analyzing the matching correlation surface of the reference image and the satellite image, calculating a matching performance measurement index, correcting the characteristic vector, guiding and optimizing the reference image, and ending the reference image preparation and optimization process.
The process of extracting the edge characteristics of each region and screening the blocked regions with stable edge gradient points is as follows:
according to the satellite picture of the target area, the satellite picture is divided into a plurality of sub-areas, as shown in fig. 2, the edge detection of the image is to use a discretization gradient approximation function to find the gray level jump position of the image gray level matrix according to the gradient vector of the two-dimensional gray level matrix, and then connect the points of the positions in the image to form the so-called image edge. The selection strategy is to select a sub-region in which gradient edges are concentrated and stable, and the selection mode is also to ensure that stable edge information can be obtained in heterogeneous infrared system imaging in a large probability and to ensure the similarity of heterogeneous images.
The gradient was divided into 4 levels and compared to the sample statistics of the 4 levels as shown in table 1. And selecting areas with obvious concentration, wherein the candidate areas selected by the target area are an area 5 and an area 7.
TABLE 1 subregion gradient feature histogram Table
| Subarea | Gradient grade | 1 | |
|
|
| 1 | 401695 | 210796 | 120432 | 77159 | |
| 2 | 456190 | 199114 | 109368 | 69573 | |
| 3 | 611629 | 146493 | 68119 | 41373 | |
| 4 | 505308 | 172192 | 93492 | 61521 | |
| 5 | 710924 | 146338 | 41584 | 22355 | |
| 6 | 517751 | 171202 | 83510 | 51910 | |
| 7 | 782992 | 91640 | 30488 | 18005 | |
| 8 | 583435 | 151043 | 68775 | 43351 | |
| 9 | 419121 | 187095 | 111749 | 74036 |
Measuring the autocorrelation degree of each region by using the repetitive pattern index, sequencing the regions according to the internal correlation index, and outputting an optimal candidate region;
step 1: selecting a subgraph i aiming at each block area, wherein a calculation formula of a repeated mode is as follows:
wherein S is the number of the image blocks participating in matching calculation when the subgraph i is matched point by point on the candidate region;
pi is the number of image blocks with edge point correlation coefficient between s image blocks and the sub image i larger than threshold in the matching process.
Step 2: the formula for calculating the reference graph repeat pattern is:
and step 3: and sorting the repeated mode indexes from small to large, and selecting the area with the minimum repeated index as the final output area of the reference image.
TABLE 2 correlation table in subregion
| | Region | 5 | |
| |
0 | 0.020117191 | |
| |
0 | 0 | |
| |
0 | 0.017970387 | |
| |
0.013828491 | 0 | |
| |
0.017665928 | 0.030609096 | |
| |
0.02223386 | 0.02458329 | |
| |
0.005933324 | 0.030570649 | |
| |
0.015193883 | 0.02530859 | |
| |
0.016019976 | 0.000337711 | |
| Comprehensive repetitive pattern | 0.010097273 | 0.016610768 |
According to the principle of lowest repetition pattern, the sub-region 5 is determined as the final reference map preparation region.
The process of establishing the spatial feature vector of the candidate region reference map by adopting the spatial distribution description method comprises the following steps:
and (3) performing mode extraction on the reference diagram, expressing the reference diagram as a direction mode and a scale mode, selecting grades according to quadrants on the direction attribute, selecting a plurality of grades according to points of the contour length on the scale attribute, designing a sensitivity analysis test of the mode, and proving that the direction and scale attributes and the matching result of the reference diagram are high in sensitivity as shown in fig. 3.
Through actual scene analysis, interested targets are all artifacts, edge included angles of the artifacts in all directions are over 30 degrees, the distribution of edge directions is too large, the distribution of template features cannot be achieved, and if the intervals are too thin, information redundancy is possible, so that the template standardization process is complicated. Therefore, here, the division precision of the selected direction attribute is selected as follows: [0,30), [30,60), [60,90), [90,120), [120,150, [150,180) according to eachThe projection length of each section contour line is calculated in scale, and finally the form can be established as { X1,X2,X3,X4,X5,X6The six-dimensional feature vector of. The six-dimensional feature vector of this target is: {45, 349, 52, 56, 4545, 167}.
The selected scene matching capability measurement evaluation index set verifies the correlation analysis process between the index set and the matching probability:
referring to the matching performance metric method, three indexes for measuring the matching correlation surface are selected from the local indexes and the global indexes, as shown in table 3.
TABLE 3 set of significance indicators
| Serial number | Index (I) | Type of |
| 1 | Peak to side lobe ratio | |
| 2 | Degree of sharpness | |
| 3 | Global significance | Global index |
And designing an index monotonicity verification experiment, decomposing the contour lines of the reference graph, sequencing the matching score condition of each contour line, sequentially deleting the contour lines where the secondary peaks of the matching correlation surfaces are located, verifying the monotonicity of the indexes, and obtaining the result shown in table 4.
TABLE 4 index monotonicity verification results
| Serial number | Peak to side lobe ratio | Degree of sharpness | |
| 1 | 3.34 | 0.42 | 37.8 |
| 2 | 3.34 | 0.31 | 35.7 |
| 3 | 3.3 | 0.41 | 65.3 |
| 4 | 4.55 | 0.47 | 58.5 |
| 5 | 5.64 | 0.48 | 72.8 |
| 6 | 6.82 | 0.47 | 71.6 |
| 7 | 6.81 | 0.47 | 75.6 |
| 8 | 7.58 | 0.43 | 82.2 |
| 9 | 7.84 | 0.4 | 83.3 |
| 10 | 8.09 | 0.38 | 83.5 |
| 11 | 8.23 | 0.38 | 85.3 |
| 12 | 8.41 | 0.4 | 79.5 |
| 13 | 8.54 | 0.41 | 74.5 |
| 14 | 8.59 | 0.44 | 70 |
| 15 | 9.67 | 0.45 | 69.5 |
| 16 | 9.76 | 0.46 | 69 |
| 17 | 9.86 | 0.5 | 68 |
| 18 | 10 | 0.51 | 67 |
| 19 | 9.98 | 0.51 | 65 |
| 20 | 10.12 | 0.52 | 66 |
| 21 | 9.98 | 0.53 | 66 |
| 22 | 10.12 | 0.54 | 66 |
| 23 | 10.42 | 0.55 | 65 |
| 24 | 10.65 | 0.56 | 62 |
| 25 | 9.87 | 0.67 | 59 |
| 26 | 8.96 | 0.75 | 0 |
The monotonicity verification result shows that the local index has obvious monotonicity, the global index is non-monotonicity, and the peak-to-side lobe ratio and the sharpness are finally selected as supervision indexes for optimizing the reference graph.
As shown in fig. 4, as a result of matching the test image, the matching uses two indexes, i.e., a peak-to-side lobe ratio and a sharpness degree, to iterate, modify the reference map of the corresponding quadrant, and after the iteration, compare histograms of the primary peak and the secondary peak, and reduce the contour until the matchability metric meets the threshold requirement, so that the final optimized reference map improves the matching probability, and completes the scene matching region selection and reference map optimization process of the matching system.
Claims (7)
1. A scene matching area selection and reference map optimization method of an image matching system is characterized in that: the method comprises the following steps:
step 1: initializing blocks aiming at a remote sensing satellite picture, extracting edge characteristics of each region according to an edge characteristic extraction algorithm, and screening the block regions with concentrated edge gradient points;
step 2, calculating the autocorrelation degree of each region by using a repetitive pattern index measurement method, sequencing the regions from small to large according to the inner correlation index, and outputting an optimal candidate region;
and step 3: establishing a spatial feature vector of a reference map of the candidate region by adopting a spatial distribution description method;
and 4, step 4: establishing a scene matchability measurement evaluation index set, analyzing the relevance between the evaluation index set and the matching probability, and outputting a matching evaluation index set;
and 5: and statistically analyzing the matching correlation surface of the reference image and the satellite image, and optimizing the reference image according to the reference image optimization method.
2. The scene matching area selection and reference map optimization method of the image matching system according to claim 1, wherein: the edge feature extraction algorithm is a Canny operator edge feature extraction algorithm.
3. The scene matching area selection and reference map optimization method of the image matching system according to claim 1, wherein: the screening of the blocking region in the edge gradient point set comprises the following processes:
step 1: extracting edge features by using a Canny operator aiming at each block subregion;
step 2: calculating the gradient change histogram distribution of each subregion;
and step 3: and comparing the distribution of the gradient point sets of each sub-region, and selecting one or more regions in the point set from small to large according to the gradient change as candidate regions.
4. The scene matching area selection and reference map optimization method of the image matching system according to claim 1, wherein: the repetitive pattern index measurement method comprises the following steps:
step 1: selecting a subgraph i aiming at each block area, wherein a calculation formula of a repeated mode index is as follows:
wherein s is the number of the image blocks participating in matching calculation when the subgraph i is matched point by point on the candidate region; pi is the number of image blocks with edge point correlation coefficient between s image blocks and the sub-image i larger than threshold in the matching process;
step 2: the formula for calculating the reference graph repeat pattern is:
wherein cf is the degree of pattern measurement for the entire tile; n is the number of selected sub-picture blocks.
5. The scene matching area selection and reference map optimization method of the image matching system according to claim 1, wherein: the method for establishing the spatial feature vector of the candidate region reference map by adopting the spatial distribution description method comprises the following processes:
step 1: extracting features of the reference graph, expressing the features as direction features and scale features, and establishing a feature sensitivity analysis test;
step 2: dividing the reference image of the reference image area into intervals according to the minimum resolvable interval of the directional characteristic;
and step 3: and (5) counting the number of gauge points in the interval and establishing a multi-dimensional feature vector.
6. The scene matching area selection and reference map optimization method of the image matching system according to claim 5, wherein: the multi-dimensional feature vector is { X1,X2,X3,X4,X5,X6In which XiIs a single interval point number.
7. The scene matching area selection and reference map optimization method of the image matching system according to claim 1, wherein: the reference map optimization method comprises the following steps:
step 1: completing a matching process by using a Harsdoff distance matching method, and calculating a matching performance measurement index;
step 2: after judging the quality of the matching metric index, calculating a main influence interval of a main peak and a secondary peak;
and step 3: and correcting the edge number of the interval, optimizing the reference graph, and recalculating the matching metric index until the matching threshold requirement is met.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510784714.0A CN106709501B (en) | 2015-11-16 | 2015-11-16 | Scene matching area selection and reference image optimization method of image matching system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510784714.0A CN106709501B (en) | 2015-11-16 | 2015-11-16 | Scene matching area selection and reference image optimization method of image matching system |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN106709501A CN106709501A (en) | 2017-05-24 |
| CN106709501B true CN106709501B (en) | 2020-03-10 |
Family
ID=58931542
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201510784714.0A Active CN106709501B (en) | 2015-11-16 | 2015-11-16 | Scene matching area selection and reference image optimization method of image matching system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN106709501B (en) |
Families Citing this family (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109063731B (en) * | 2018-06-26 | 2020-11-10 | 北京航天自动控制研究所 | Scene adaptability criterion training sample set generation method |
| CN110503679A (en) * | 2019-08-29 | 2019-11-26 | 四川轻化工大学 | A kind of preparation of infrared reference figure and evaluation method |
| CN111444985B (en) * | 2020-04-26 | 2023-04-07 | 南京大学 | Image matching method based on histogram matching |
| CN113392255B (en) * | 2021-05-21 | 2022-02-08 | 广东国地规划科技股份有限公司 | Automatic matching method and device for pattern spots in ecological restoration project |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US5027422A (en) * | 1988-08-29 | 1991-06-25 | Raytheon Company | Confirmed boundary pattern matching |
| CN102708386A (en) * | 2012-04-18 | 2012-10-03 | 中国电子科技集团公司第十研究所 | Optical/SAR (synthetic aperture radar) heterogeneous image matching method |
| CN105659906B (en) * | 2010-04-30 | 2013-02-13 | 北京理工大学 | Scene matching area automatically selecting method based on image information entropy |
| CN103353988A (en) * | 2013-06-17 | 2013-10-16 | 苏州科技学院 | Method for evaluating performance of heterogeneous SAR (synthetic aperture radar) image feature matching algorithm |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20150199573A1 (en) * | 2014-01-10 | 2015-07-16 | Mitsubishi Electric Research Laboratories, Inc. | Global Scene Descriptors for Matching Manhattan Scenes using Edge Maps Associated with Vanishing Points |
-
2015
- 2015-11-16 CN CN201510784714.0A patent/CN106709501B/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US5027422A (en) * | 1988-08-29 | 1991-06-25 | Raytheon Company | Confirmed boundary pattern matching |
| CN105659906B (en) * | 2010-04-30 | 2013-02-13 | 北京理工大学 | Scene matching area automatically selecting method based on image information entropy |
| CN102708386A (en) * | 2012-04-18 | 2012-10-03 | 中国电子科技集团公司第十研究所 | Optical/SAR (synthetic aperture radar) heterogeneous image matching method |
| CN103353988A (en) * | 2013-06-17 | 2013-10-16 | 苏州科技学院 | Method for evaluating performance of heterogeneous SAR (synthetic aperture radar) image feature matching algorithm |
Non-Patent Citations (1)
| Title |
|---|
| 基于边缘相似性的异源图像匹配;李壮等;《飞行器测控学报》;20110430;第30卷(第2期);全文 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN106709501A (en) | 2017-05-24 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN113689428B (en) | Mechanical part stress corrosion detection method and system based on image processing | |
| CN106709501B (en) | Scene matching area selection and reference image optimization method of image matching system | |
| CN103632363B (en) | Object level high-resolution remote sensing image change detecting method based on Multiscale Fusion | |
| CN107945215B (en) | High-precision infrared image tracker and target rapid tracking method | |
| CN108205645B (en) | Reference image quality evaluation method of heterogeneous image matching system | |
| CN101697229A (en) | Method for extracting region of interest of medical image | |
| CN104899892A (en) | Method for quickly extracting star points from star images | |
| CN105447488B (en) | SAR image target detection method based on sketch line segment topological structure | |
| CN111814597A (en) | Urban function partitioning method coupling multi-label classification network and YOLO | |
| CN105205816A (en) | Method for extracting high-resolution SAR image building zone through multi-feature weighted fusion | |
| CN108537169A (en) | A kind of high-resolution remote sensing image method for extracting roads based on center line and detection algorithm of having a lot of social connections | |
| CN113985406B (en) | Target track splicing method for marine radar | |
| CN101770583A (en) | Template matching method based on global features of scene | |
| CN104637060A (en) | Image partition method based on neighbor-hood PCA (Principal Component Analysis)-Laplace | |
| CN113936214A (en) | Karst wetland vegetation community classification method based on fusion of aerospace remote sensing images | |
| CN111882573B (en) | Cultivated land block extraction method and system based on high-resolution image data | |
| CN106204596B (en) | Panchromatic waveband remote sensing image cloud detection method based on Gaussian fitting function and fuzzy mixed estimation | |
| CN111488907B (en) | Robust image recognition method based on dense PCANet | |
| CN108205683B (en) | Infrared target identification system threshold value optimization method based on visible light satellite picture | |
| CN109272534B (en) | SAR image change detection method based on multi-granularity cascade forest model | |
| CN109886133B (en) | Ship detection method and system based on optical remote sensing image | |
| CN112927201A (en) | Curve detection method and device | |
| CN112163454A (en) | High-frequency ground wave radar clutter intelligent classification and positioning method based on RD spectrum enhancement | |
| CN112014821A (en) | Unknown vehicle target identification method based on radar broadband characteristics | |
| CN111104922A (en) | Feature matching algorithm based on ordered sampling |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |