CN110503678A - Navigation equipment based on topological structure constraint is infrared with the heterologous method for registering of optics - Google Patents

Navigation equipment based on topological structure constraint is infrared with the heterologous method for registering of optics Download PDF

Info

Publication number
CN110503678A
CN110503678A CN201910800263.3A CN201910800263A CN110503678A CN 110503678 A CN110503678 A CN 110503678A CN 201910800263 A CN201910800263 A CN 201910800263A CN 110503678 A CN110503678 A CN 110503678A
Authority
CN
China
Prior art keywords
infrared
topological structure
image
optics
heterologous
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.)
Pending
Application number
CN201910800263.3A
Other languages
Chinese (zh)
Inventor
徐衍胜
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Individual
Original Assignee
Individual
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by Individual filed Critical Individual
Priority to CN201910800263.3A priority Critical patent/CN110503678A/en
Publication of CN110503678A publication Critical patent/CN110503678A/en
Pending legal-status Critical Current

Links

Classifications

    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06TIMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00Image analysis
    • G06T7/30Determination of transform parameters for the alignment of images, i.e. image registration
    • G06T7/33Determination of transform parameters for the alignment of images, i.e. image registration using feature-based methods
    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06TIMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T2207/00Indexing scheme for image analysis or image enhancement
    • G06T2207/10Image acquisition modality
    • G06T2207/10048Infrared image
    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06TIMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T2207/00Indexing scheme for image analysis or image enhancement
    • G06T2207/10Image acquisition modality
    • G06T2207/10052Images from lightfield camera
    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06TIMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T2207/00Indexing scheme for image analysis or image enhancement
    • G06T2207/30Subject of image; Context of image processing
    • G06T2207/30108Industrial image inspection
    • G06T2207/30141Printed circuit board [PCB]

Landscapes

  • Engineering & Computer Science (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Theoretical Computer Science (AREA)
  • Aiming, Guidance, Guns With A Light Source, Armor, Camouflage, And Targets (AREA)

Abstract

The invention belongs to heterologous method for registering technical fields, and in particular to a kind of navigation equipment based on topological structure constraint is infrared with the heterologous method for registering of optics, including the following steps: S1, calibration optics image characteristic point;S2, Infrared Image Features point is found;S3, building are using triangle as the topological structure network of unit;S4, building three-legged structure network similarity measurements flow function;S5, infrared and optical imagery spatial transform relation is solved.The present invention forms the topological structure as unit of triangle using infrared image centroid feature point, topological structure the constraint relationship and topological structure measuring similarity function are constructed with triangle interior angle, complete optics and the matching of infrared topological structure, the matched accuracy of lifting feature point and heterologous registration accuracy.The present invention heterologous is registrated for infrared with optical.

Description

Navigation equipment based on topological structure constraint is infrared with the heterologous method for registering of optics
Technical field
The invention belongs to heterologous method for registering technical fields, and in particular to a kind of navigation equipment based on topological structure constraint The infrared and heterologous method for registering of optics.
Background technique
As navigation equipment is increasingly miniaturized, micromation, equipment caused by relevant equipment fault takes place frequently is reprocessed not in time The normal use of user, while traditional means fault detection higher cost are seriously affected, the period is longer, is also navigation equipment Popularization bring obstruction, for this purpose, improve navigation equipment fault-detecting ability have become important topic urgently to be resolved.
Contactless fault detection technique is by obtaining to electronic equipment optical imagery and infrared image fusion treatment There are the new images of electronic equipment background information and radiation information simultaneously, can be realized the effectively accurate feature extraction of thermal imagery and quickly Thermal imagery positions extremely.Since there is differences in spatial location between optical imagery for electronic equipment infrared, need by image Registration, therefore the present invention mainly studies the registration of infrared image and optical imagery based on navigation equipment.
In recent years, domestic that heterologous registration project is had conducted extensive research.Wherein, based on the method for registering of feature because it is matched Quasi velosity is fast, suitable for homologous or multi-source image registration, becomes one of method for registering of greatest concern.Such method is usually logical It crosses in selection two images and is capable of the similitude of representative image to a certain extent, and be easier to the common characteristic information extracted As foundation, the matching of two images characteristic quantity is then completed using the feature extracted, finally by the matching relationship of characteristic quantity Obtain the spatial transform relation of two images.In practical applications, generally use point feature, common point feature have marginal point and Angle point.The selection of characteristic point be whole image registration key link, feature point extraction it is accurate whether directly determine registration Success or failure.
Whether need artificially to participate according to the process for extracting characteristic point, the matching based on characteristic point can be divided into interactive mode Method for registering and self-action method for registering.
Interactive registration method refers to by operator's selection subjective on benchmark image or image subject to registration respectively Characteristic point pair, or the selected characteristic point on piece image are obtained in another piece image corresponding by certain searching method Characteristic point.
Autoegistration method is then to complete the extraction of point feature, the matching and spatial alternation ginseng of characteristic quantity by computer completely Number calculate, the basic procedure of algorithm is: first to image carry out pre-process and using certain detective operators to benchmark image and to Registration image extracts characteristic point (such as Harris, SUSAN, SIFT, SURF etc.) respectively, then uses customized characteristic area Similarity measurement method carry out characteristic point matching, complete the calculating of spatial transformation parameter, finally to image carry out space coordinate Transformation.
Since angle point has concentrated many important shape informations, it can effectively reflect the local feature of image, be image registration In one of most common feature.Wherein, robust features (Speed-Up Robust Feature, SURF) angle point is accelerated to have Rotational invariance, stability are strong, distinguishability is good and the characteristic of informative, representative in autoegistration method. But infrared image and optical imagery belong to heterogeneous more granular informations, and consistent SURF characteristic point ten is extracted from infrared/optical imagery Divide difficulty, causes successful match rate low.
Summary of the invention
In view of the above technical problems, it is red to provide a kind of high navigation equipment based on topological structure constraint of successful match rate Outside with the heterologous method for registering of optics.
In order to solve the above-mentioned technical problem, the technical solution adopted by the present invention are as follows:
Navigation equipment based on topological structure constraint is infrared with the heterologous method for registering of optics, including the following steps:
S1, calibration optics image characteristic point;
S2, Infrared Image Features point is found;
S3, building are using triangle as the topological structure network of unit;
S4, building three-legged structure network similarity measurements flow function;
S5, infrared and optical imagery spatial transform relation is solved.
The method of optical imagery characteristic point is demarcated in the S1 are as follows: determine according to the calorific value of heating element and reflectivity The position of position heating element, obtains the center position information of heating element, the characteristic point as optical imagery.
The method of Infrared Image Features point is found in the S2 are as follows: using the variation of front and back calorific value on circuit board, to red Outer difference image carries out binarization threshold segmentation, and extracts heating region mass center.
Building is using triangle as the method for the topological structure network of unit in the S3 are as follows: by infrared image and optical imagery On characteristic point respectively form several triangular networks, N number of characteristic point is extracted on infrared image, then is had accordinglyIt is a Triangular network extracts M characteristic point on optical imagery, accordingly hasA triangular network forms three tops of triangular network Point is A (x respectivelyi,yi)、B(xj,yj)、C(xk,yk), wherein i, j, k are point set rope of three points in entire infrared image respectively Draw value, these three points are joined end to end and form three vectors, wherein the angle θ of corresponding A pointiIt indicates, B point θjIt indicates, C point is used θkIt indicates, the solution mode of angle are as follows:
Then acquire three angles are arranged by sequence from small to large.
The method of three-legged structure network similarity measurements flow function is constructed in the S4 are as follows: find out the similar triangle in angle Network, similarity measurement criterion are
Wherein, θpFor the sum of the absolute value of the difference at three angles of two triangular networksθIiFor infrared image The angle of upper triangular network, θOiFor the angle of triangular network on optical imagery, wherein i=1,2,3, to all θ being calculatedp It is ranked up, therefrom chooses θpLess than the similar triangular network of certain threshold value.
The method of the spatial transform relation of infrared and optical imagery is solved in the S5 are as follows: image registration mathematically can be with It indicates are as follows: I1(x, y)=g (I2(f (x, y))), wherein I1、I2Respectively reference picture and image subject to registration, f are two-dimensional coordinate Transforming function transformation function, g are one-dimensional coordinate transforming function transformation function, determine ideal f function, the function f by the positional relationship between coordinate It is described with global change's model, so that it is determined that the corresponding relationship of image between points, according to the infrared and optical picture acquired As upper similar triangles find the coordinate of at least three corresponding points, the affine change of two images is obtained according to corresponding characteristic point Change relationship, Y=AX, whereinIndicate the characteristic point on infrared image,Indicate light Learn the characteristic point of image;The affine matrix between two images is indicated, to derive A= ((XTX)-1XTY)T, the affine matrix found out is then substituted into global change's model, obtains optical picture corresponding with infrared image Picture, to complete heterologous registration.
Global change's model uses projective transformation model, the projective transformation model can reflect from different location and Angle obtains the corresponding relationship between any two images of Same Scene, and two images are restored to the projective rejection of same position Are as follows:
Wherein w=a31x+a32y+a33For scale factor, projective transformation no longer guarantees the parallel sexual intercourse of transformation front and back, needs 9 parameters are calculated, under the premise of not conllinear, need four corresponding points in two images that can uniquely determine at any 3 points Required parameter.
Compared with prior art, the present invention having the beneficial effect that
The present invention forms the topological structure as unit of triangle using infrared image centroid feature point, with triangle interior angle Topological structure the constraint relationship and topological structure measuring similarity function are constructed, optics and the matching of infrared topological structure is completed, is promoted The accuracy of Feature Points Matching and heterologous registration accuracy.
Detailed description of the invention
Fig. 1 is flow chart of the invention;
Fig. 2 is vector triangle topology schematic diagram of the present invention;
Fig. 3 is matched triangle topological diagram in infrared image of the present invention;
Fig. 4 is matched triangle topological diagram in optical imagery of the present invention.
Specific embodiment
Following will be combined with the drawings in the embodiments of the present invention, and technical solution in the embodiment of the present invention carries out clear, complete Site preparation description, it is clear that described embodiments are only a part of the embodiments of the present invention, instead of all the embodiments.It is based on Embodiment in the present invention, it is obtained by those of ordinary skill in the art without making creative efforts every other Embodiment shall fall within the protection scope of the present invention.
Navigation equipment based on topological structure constraint is infrared with the heterologous method for registering of optics, as shown in Figure 1, including following step It is rapid:
S1, calibration optics image characteristic point, position heating element according to the calorific value of heating element and reflectivity Position, obtain the center position information of heating element, the characteristic point as optical imagery.
S2, Infrared Image Features point is found, using the variation of front and back calorific value on circuit board, infrared differential image is carried out Binarization threshold segmentation, and extract heating region mass center.
S3, building are using triangle as the topological structure network of unit, as shown in Figure 2,3, 4, by infrared image and optical picture As upper characteristic point respectively forms several triangular networks, N number of characteristic point is extracted on infrared image, then is had accordingly A triangular network extracts M characteristic point on optical imagery, accordingly hasA triangular network forms three of triangular network Vertex is A (x respectivelyi,yi)、B(xj,yj)、C(xk,yk), wherein i, j, k are point set of three points in entire infrared image respectively These three points are joined end to end and form three vectors, wherein the angle θ of corresponding A point by index valueiIt indicates, B point θjIt indicates, C point Use θkIt indicates, the solution mode of angle are as follows:
Then acquire three angles are arranged by sequence from small to large.
S4, building three-legged structure network similarity measurements flow function, find out the similar triangular network in angle, similarity measurements Measuring criterion is
Wherein, θpFor the sum of the absolute value of the difference at three angles of two triangular networks,θIiFor infrared image The angle of upper triangular network, θOiFor the angle of triangular network on optical imagery, wherein i=1,2,3, to all θ being calculatedp It is ranked up, therefrom chooses θpLess than the similar triangular network of certain threshold value.
S5, infrared and optical imagery spatial transform relation is solved, image registration is mathematically represented as: I1(x, Y)=g (I2(f (x, y))), wherein I1、I2Respectively reference picture and image subject to registration, f are two-dimensional coordinate transforming function transformation function, and g is One-dimensional coordinate transforming function transformation function determines ideal f function by the positional relationship between coordinate, function f with global change's model come Description, global change's model uses projective transformation model, so that it is determined that the corresponding relationship of image between points, projective transformation mould Type can reflect from the corresponding relationship between any two images that different location and angle obtain Same Scene, and two images are extensive The projective rejection of same position is arrived again are as follows:
Wherein w=a31x+a32y+a33For scale factor, projective transformation no longer guarantees the parallel sexual intercourse of transformation front and back, needs 9 parameters are calculated, under the premise of not conllinear, need four corresponding points in two images that can uniquely determine at any 3 points Required parameter.The coordinate that at least three corresponding points are found according to the similar triangles on the infrared and optical imagery acquired, according to Corresponding characteristic point obtains the affine transformation relationship of two images, Y=AX, whereinIt indicates on infrared image Characteristic point,Indicate the characteristic point of optical imagery;It indicates between two images Affine matrix, to derive A=((XTX)-1XTY)T, the affine matrix found out is then substituted into global change's model, is obtained Optical imagery corresponding with infrared image, to complete heterologous registration.
Workflow of the invention are as follows: the present invention acquires navigation equipment circuit board temperature rise sequence using infrared thermal imaging sensor Column extract the obvious region that generates heat according to the difference for powering on front and back component calorific value, and then extract mass center as on infrared image Characteristic point.Infrared image composition topological structure characterized by mass center;Optical imagery structure characterized by component central point At topological structure, similarity measurement is carried out to the topological structure relationship in two images, chooses the topological structure to match, is calculated Transformation model matrix is finally completed the registration of infrared image and optical imagery.
Only presently preferred embodiments of the present invention is explained in detail above, but the present invention is not limited to above-described embodiment, Within the knowledge of a person skilled in the art, it can also make without departing from the purpose of the present invention each Kind variation, various change should all be included in the protection scope of the present invention.

Claims (7)

1. the navigation equipment based on topological structure constraint is infrared with the heterologous method for registering of optics, it is characterised in that: including following step It is rapid:
S1, calibration optics image characteristic point;
S2, Infrared Image Features point is found;
S3, building are using triangle as the topological structure network of unit;
S4, building three-legged structure network similarity measurements flow function;
S5, infrared and optical imagery spatial transform relation is solved.
2. the navigation equipment according to claim 1 based on topological structure constraint is infrared with the heterologous method for registering of optics, Be characterized in that: in the S1 demarcate optical imagery characteristic point method are as follows: according to the calorific value of heating element and reflectivity come The position for positioning heating element, obtains the center position information of heating element, the characteristic point as optical imagery.
3. the navigation equipment according to claim 1 based on topological structure constraint is infrared with the heterologous method for registering of optics, It is characterized in that: finding the method for Infrared Image Features point in the S2 are as follows: using the variation of front and back calorific value on circuit board, to red Outer difference image carries out binarization threshold segmentation, and extracts heating region mass center.
4. the navigation equipment according to claim 1 based on topological structure constraint is infrared with the heterologous method for registering of optics, Be characterized in that: building is using triangle as the method for the topological structure network of unit in the S3 are as follows: by infrared image and optical picture As upper characteristic point respectively forms several triangular networks, N number of characteristic point is extracted on infrared image, then is had accordingly A triangular network extracts M characteristic point on optical imagery, accordingly hasA triangular network forms three of triangular network Vertex is A (x respectivelyi,yi)、B(xj,yj)、C(xk,yk), wherein i, j, k are point set of three points in entire infrared image respectively These three points are joined end to end and form three vectors, wherein the angle θ of corresponding A point by index valueiIt indicates, B point θjIt indicates, C point Use θkIt indicates, the solution mode of angle are as follows:
Then acquire three angles are arranged by sequence from small to large.
5. the navigation equipment according to claim 1 based on topological structure constraint is infrared with the heterologous method for registering of optics, It is characterized in that: constructing the method for three-legged structure network similarity measurements flow function in the S4 are as follows: find out similar three in angle Angle network, similarity measurement criterion are
Wherein, θpFor the sum of the absolute value of the difference at three angles of two triangular networks,θIiIt is on infrared image three The angle of angle network, θOiFor the angle of triangular network on optical imagery, wherein i=1,2,3, to all θ being calculatedpIt carries out Sequence, therefrom chooses θpLess than the similar triangular network of certain threshold value.
6. the navigation equipment according to claim 1 based on topological structure constraint is infrared with the heterologous method for registering of optics, It is characterized in that: solving the method for infrared and optical imagery spatial transform relation in the S5 are as follows: image registration mathematically may be used To indicate are as follows: I1(x, y)=g (I2(f (x, y))), wherein I1、I2Respectively reference picture and image subject to registration, f are that two dimension is sat Transforming function transformation function is marked, g is one-dimensional coordinate transforming function transformation function, determines ideal f function, the function by the positional relationship between coordinate F is described with global change's model, so that it is determined that the corresponding relationship of image between points, according to the infrared and optical picture acquired As upper similar triangles find the coordinate of at least three corresponding points, the affine change of two images is obtained according to corresponding characteristic point Change relationship, Y=AX, whereinIndicate the characteristic point on infrared image,Indicate light Learn the characteristic point of image;The affine matrix between two images is indicated, to derive A= ((XTX)-1XTY)T, the affine matrix found out is then substituted into global change's model, obtains optical picture corresponding with infrared image Picture, to complete heterologous registration.
7. the navigation equipment according to claim 6 based on topological structure constraint is infrared with the heterologous method for registering of optics, Be characterized in that: global change's model uses projective transformation model, and the projective transformation model can reflect from different location Corresponding relationship between any two images of angle acquisition Same Scene, the projection that two images are restored to same position is closed System are as follows:
Wherein w=a31x+a32y+a33For scale factor, projective transformation no longer guarantees the parallel sexual intercourse of transformation front and back, needs to count 9 parameters are calculated, under the premise of any 3 points not conllinear, are needed needed for four corresponding points can uniquely determine in two images Parameter.
CN201910800263.3A 2019-08-28 2019-08-28 Navigation equipment based on topological structure constraint is infrared with the heterologous method for registering of optics Pending CN110503678A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201910800263.3A CN110503678A (en) 2019-08-28 2019-08-28 Navigation equipment based on topological structure constraint is infrared with the heterologous method for registering of optics

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201910800263.3A CN110503678A (en) 2019-08-28 2019-08-28 Navigation equipment based on topological structure constraint is infrared with the heterologous method for registering of optics

Publications (1)

Publication Number Publication Date
CN110503678A true CN110503678A (en) 2019-11-26

Family

ID=68589760

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201910800263.3A Pending CN110503678A (en) 2019-08-28 2019-08-28 Navigation equipment based on topological structure constraint is infrared with the heterologous method for registering of optics

Country Status (1)

Country Link
CN (1) CN110503678A (en)

Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN103177444A (en) * 2013-03-08 2013-06-26 中国电子科技集团公司第十四研究所 Automatic SAR (synthetic-aperture radar) image rectification method
CN103778626A (en) * 2013-12-31 2014-05-07 北京理工大学 Quick image registration method based on visual remarkable area
CN107025648A (en) * 2017-03-20 2017-08-08 中国人民解放军空军工程大学 A kind of board failure infrared image automatic testing method
CN107179494A (en) * 2017-05-27 2017-09-19 浙江万里学院 Device fault detection method and system on a kind of general-purpose circuit board plate
CN206773147U (en) * 2017-05-27 2017-12-19 浙江万里学院 Device fault detecting system on a kind of general-purpose circuit board plate
CN109461132A (en) * 2018-10-31 2019-03-12 中国人民解放军国防科技大学 SAR image automatic registration method based on feature point geometric topological relation

Patent Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN103177444A (en) * 2013-03-08 2013-06-26 中国电子科技集团公司第十四研究所 Automatic SAR (synthetic-aperture radar) image rectification method
CN103778626A (en) * 2013-12-31 2014-05-07 北京理工大学 Quick image registration method based on visual remarkable area
CN107025648A (en) * 2017-03-20 2017-08-08 中国人民解放军空军工程大学 A kind of board failure infrared image automatic testing method
CN107179494A (en) * 2017-05-27 2017-09-19 浙江万里学院 Device fault detection method and system on a kind of general-purpose circuit board plate
CN206773147U (en) * 2017-05-27 2017-12-19 浙江万里学院 Device fault detecting system on a kind of general-purpose circuit board plate
CN109461132A (en) * 2018-10-31 2019-03-12 中国人民解放军国防科技大学 SAR image automatic registration method based on feature point geometric topological relation

Non-Patent Citations (4)

* Cited by examiner, † Cited by third party
Title
XIEYAN0811: "图像处理之_仿射变换与透视变换", 《HTTPS://WWW.JIANSHU.COM/P/44CFA0BE7C0E》 *
王鲲鹏等: "红外与可见光图像配准方法分类及现状", 《红外技术》 *
罗云林等: "基于ROI互信息的电路板红外图像配准技术", 《红外技术》 *
阿阿阿阿毛: "图像配准", 《HTTPS://WWW.JIANSHU.COM/P/693BB77E86FD》 *

Similar Documents

Publication Publication Date Title
CN104376552B (en) A kind of virtual combat method of 3D models and two dimensional image
CN108765328B (en) High-precision multi-feature plane template and distortion optimization and calibration method thereof
CN106228507A (en) A kind of depth image processing method based on light field
CN105701827A (en) Method and device for jointly calibrating parameters of visible light camera and infrared camera
CN109523595A (en) A kind of architectural engineering straight line corner angle spacing vision measuring method
CN108921864A (en) A kind of Light stripes center extraction method and device
CN108615244A (en) A kind of image depth estimation method and system based on CNN and depth filter
CN112766302B (en) Image fusion method and device, storage medium and electronic device
CN108010086A (en) Camera marking method, device and medium based on tennis court markings intersection point
CN110120071A (en) A kind of depth estimation method towards light field image
CN111862180A (en) Camera group pose acquisition method and device, storage medium and electronic equipment
CN107038714A (en) Many types of visual sensing synergistic target tracking method
CN110456904A (en) A kind of augmented reality glasses eye movement exchange method and system without calibration
Crispel et al. All-sky photogrammetry techniques to georeference a cloud field
CN104200476A (en) Method for solving intrinsic parameters of camera by circular motion in double-plane-mirror device
Yue et al. Accurate three dimensional body scanning system based on structured light
CN106500729A (en) A kind of smart mobile phone self-inspection calibration method without the need for control information
Chaochuan et al. An extrinsic calibration method for multiple RGB-D cameras in a limited field of view
CN108961151A (en) A method of the three-dimensional large scene that ball curtain camera obtains is changed into sectional view
CN109360249A (en) Calibration system is adjusted in camera
Cao et al. Quantifying visual environment by semantic segmentation using deep learning
CN110503678A (en) Navigation equipment based on topological structure constraint is infrared with the heterologous method for registering of optics
CN108447092A (en) The method and device of vision positioning marker
CN107256563A (en) Underwater 3 D reconstructing system and its method based on difference liquid level image sequence
CN111194015A (en) Outdoor positioning method and device based on building and mobile equipment

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