CN103514450B - Image feature extracting method and image correcting method and equipment - Google Patents
Image feature extracting method and image correcting method and equipment Download PDFInfo
- Publication number
- CN103514450B CN103514450B CN201210222295.8A CN201210222295A CN103514450B CN 103514450 B CN103514450 B CN 103514450B CN 201210222295 A CN201210222295 A CN 201210222295A CN 103514450 B CN103514450 B CN 103514450B
- Authority
- CN
- China
- Prior art keywords
- image
- angle
- collection
- pose
- capture device
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Expired - Fee Related
Links
Abstract
The embodiment of the invention discloses an image feature extracting method. The image feature extracting method comprises the steps that an image is collected, three-dimensional attitude information of image collecting equipment when a camera collects images is acquired, the image collecting equipment comprises a camera, the collected images are corrected based on the three-dimensional attitude information, and feature extracting processing is conducted on the corrected images. Correspondingly, the embodiment of the invention further discloses image collecting equipment. The problems that the local part is not changed, and feature extracting performance is lowered under the condition that the image collecting equipment tilts, rotates and the like can be solved.
Description
Technical field
The present invention relates to image processing field, more particularly, to a kind of image characteristic extracting method and method for correcting image and
Equipment.
Background technology
Feature extraction refers to extract image information using computer, determines whether the pixel of each image belongs to one
Characteristics of image.Image characteristics extraction is an important research content in the fields such as graphical analyses, pattern recognition and computer vision,
It is the Research foundation of great number of issues.The most important characteristic of feature extraction is " repeatable ", i.e. the different figures of Same Scene
Feature as being extracted should be repeatable, but between the multiple image being located due to target, major part all has geometry change
Change, light change, influence of noise, pattern distortion etc., the invariant feature therefore how extracting image becomes grinding of association area
Study carefully emphasis.In recent years, investigated a kind of extracting method of local invariant feature, because the method is directed to all kinds of viewpoints of image
Change, illumination variation, noise etc. still have preferable repeatability, survey in image registration, target following identification, remote sensing
Paint, image-guidance, three-dimensionalreconstruction, virtual reality and the aspect such as augmented reality and image retrieval achieve and widely apply, obtain
The common concern of Chinese scholars.
The extraction of invariant features mainly includes two parts：Invariant features detection and invariant features description.Wherein, constant spy
Levy the premise that detection is feature description it is therefore an objective to independent detection goes out content identical characteristic area on multiple image.Constant spy
Levying description is on the basis of detection, the feature obtaining is described, most important characteristic is that have higher ga s safety degree.
With progressively opening of China's 3G network and developing rapidly of mobile communication cause, huge smart phone user group
Also increasing year by year, application on smart mobile phone for the image processing and analysis also gradually occupies increasingly consequence.Feature
Extract as the basic and very crucial module of one of image procossing, be applied to a lot of neck such as target recognition, augmented reality
Domain, is requisite link in the development and application of smart mobile phone.Existing image characteristic extracting method is still generally direct
Input picture is processed.And input picture is to obtain in the case that mobile phone has inclination, rotates mostly, even if feature carries
When taking, institute is using being characterized as local invariant feature, and the invariance of such geometric transformation such as it is to tilting, rotation is still limited, and this just makes
The about accuracy of feature extraction, robustness, thus have influence on the development of intelligent mobile phone enhancement application.
Content of the invention
Embodiments provide a kind of image characteristic extracting method and method for correcting image and equipment, can avoid
Under the attitudes such as image capture device run-off the straight, rotation, local invariant feature extracts the problem of hydraulic performance decline.
In order to solve above-mentioned technical problem, a kind of image characteristic extracting method provided in an embodiment of the present invention, including：
Collection image, and obtain the 3 d pose information of image capture device during camera collection image；
Based on described 3 d pose information, the image of correction collection；
Feature extraction process is carried out to the image after correction.
The embodiment of the present invention additionally provides a kind of method for correcting image, including：
The image of collection object, and obtain the 3 d pose information of image capture device during camera collection image, institute
State the described photographic head of inclusion of collecting device；
Based on described 3 d pose information, the image of correction collection, obtain described image collecting device and be in and described mesh
The non-distorted image of described object is gathered during the 3 d pose of 3 d pose equidistant vertical of mark thing.
Correspondingly, the embodiment of the present invention also provides a kind of image capture device, and this equipment includes described photographic head, this equipment
Also include：Acquisition module, correction module and extraction module, wherein：First acquisition module and the first correction module and offer module,
Wherein：
Photographic head, for gathering image；
First acquisition module, for obtaining the 3 d pose information of image capture device during camera collection image；
Correction module, for the 3 d pose information obtaining based on described first acquisition module, corrects described acquisition module
The image of collection；
Extraction module, for carrying out feature extraction process to the image after the correction of described correction module.
Correspondingly, the embodiment of the present invention also provides a kind of image capture device, and this equipment includes described photographic head, this equipment
Also include：Second acquisition module and the second correction module, wherein：
Photographic head, for gathering the image of object；
Second acquisition module, for gathering the image of object, and obtains image capture device during camera collection image
3 d pose information；
Second correction module, for based on described 3 d pose information, the image of correction collection, obtaining described image collection
Equipment gathers the non-distorted image of described object when being in the 3 d pose with the 3 d pose equidistant vertical of described object.
The embodiment of the present invention, by gathering image, and obtains the 3 d pose information of image capture device during camera collection image, base
In described 3 d pose information, the image of correction collection, feature extraction process is carried out to the image after correction, so can avoid
Under the attitudes such as image capture device run-off the straight, rotation, local invariant feature extracts the problem of hydraulic performance decline.
Brief description
In order to be illustrated more clearly that the embodiment of the present invention or technical scheme of the prior art, below will be to embodiment or existing
Have technology description in required use accompanying drawing be briefly described it should be apparent that, drawings in the following description be only this
Some embodiments of invention, for those of ordinary skill in the art, on the premise of not paying creative work, acceptable
Other accompanying drawings are obtained according to these accompanying drawings.
Fig. 1 is a kind of schematic flow sheet of the first embodiment of image characteristic extracting method provided in an embodiment of the present invention；
Fig. 2 is a kind of optional design sketch of the embodiment of the present invention；
Fig. 3 is a kind of schematic flow sheet of the second embodiment of image characteristic extracting method provided in an embodiment of the present invention；
Fig. 4 is a kind of schematic flow sheet of the first embodiment of method for correcting image provided in an embodiment of the present invention；
Fig. 5 is a kind of schematic flow sheet of the second embodiment of method for correcting image provided in an embodiment of the present invention；
Fig. 6 is a kind of structural representation of the first embodiment of image capture device provided in an embodiment of the present invention；
Fig. 7 is a kind of structural representation of the second embodiment of image capture device provided in an embodiment of the present invention；
Fig. 8 is a kind of knot of another embodiment of the second embodiment of image capture device provided in an embodiment of the present invention
Structure schematic diagram；
Fig. 9 is the structural representation of the first embodiment of another kind image capture device provided in an embodiment of the present invention；
Figure 10 is the structural representation of the second embodiment of another kind image capture device provided in an embodiment of the present invention.
Specific embodiment
Below in conjunction with the accompanying drawing in the embodiment of the present invention, the technical scheme in the embodiment of the present invention is carried out clear, complete
Site preparation description is it is clear that described embodiment is only a part of embodiment of the present invention, rather than whole embodiments.It is based on
Embodiment in the present invention, it is every other that those of ordinary skill in the art are obtained under the premise of not making creative work
Embodiment, broadly falls into the scope of protection of the invention.
Fig. 1 is a kind of schematic flow sheet of the first embodiment of image characteristic extracting method provided in an embodiment of the present invention,
Shown in Fig. 1, including：
S101, collection image, and obtain the 3 d pose information of image capture device during camera collection image.
It should be noted that the 3 d pose information of image capture device can be specifically when obtaining camera collection image
When obtaining camera collection image by three-dimensional attitude sensor, the 3 d pose information of image capture device, specifically can wrap
Include：The angle of pitch of image capture device, angle of heel and azimuth when camera collection image is obtained by three-dimensional attitude sensor
Angle information, the described angle of pitch be image capture device around conventional coordinates (such as：World coordinate system) Y-axis rotation rotation
Angle, described angle of heel is for image capture device around conventional coordinates (such as：World coordinate system) X-axis rotation the anglec of rotation, described
The anglec of rotation that azimuth rotates around direct north for image capture device.Concrete acquisition process is as follows：
Three-dimensional attitude sensor is (such as：Three-dimensional electronic compass) in three orthogonal three-dimensional magnetoresistive transducers survey respectively
Amount earth's magnetic field is in conventional coordinates (such as：World coordinate system) the vector value of X-direction and earth's magnetic field in conventional coordinates (such as：
World coordinate system) the vector value of Y direction and earth's magnetic field in conventional coordinates (such as：World coordinate system) Z-direction arrow
Value.Then the information transfer measuring three-dimensional magnetoresistive transducer is to three-dimensional attitude sensor (such as：Three-dimensional electronic compass) in
Micro-control unit (Micro Control Unit, MCU), MCU is by following relational expression：
Azimuth=arcTan (Y/X)
Wherein Azimuth is the azimuth of image capture device.
Draw azimuthal angle information of image capture device during camera collection image, this azimuth is 3 d pose
Sensor is (such as：Three-dimensional electronic compass) anglec of rotation that rotates around direct north, that is, image capture device revolves around direct north
The anglec of rotation turning.
When image capture device run-off the straight, azimuthal accuracy will be greatly affected, this error big
The little size depending on image capture device location and inclination angle.For reducing the impact of this error, using 3 d pose
Sensor is (such as：Three-dimensional electronic compass) in double-shaft tilt angle sensor measuring image capture device during camera collection image
The angle of pitch and the angle information of angle of heel, this angle of pitch is for image capture device around conventional coordinates (such as：World coordinate system) Y
The anglec of rotation of axle rotation, angle of heel is for image capture device around conventional coordinates (such as：World coordinate system) X-axis rotation rotation
Angle.Double-shaft tilt angle sensor is by the information transfer of the angle of pitch and angle of heel to MCU, the conversion Calculation formula by following standard for the MCU：
Xr=Xcos α+Y sin α sin β-Z cos β sin α
Yr=Ycos β+Z sin β
Wherein, XrFor image capture device and ground surface earth's magnetic field when parallel in conventional coordinates (such as：World coordinates
System) X-direction vector value, YrFor image capture device and ground surface earth's magnetic field when parallel in conventional coordinates (such as：Generation
Boundary's coordinate system) Y direction vector value, α be collecting device gather image when the angle of pitch, β be collecting device collection image
When angle of heel.
Three-dimensional attitude sensor is (such as：Three-dimensional electronic compass) in MCU just can calculate image capture device and earth's surface
Earth's magnetic field when face is parallel is in conventional coordinates (such as：World coordinate system) X-direction vector value XrWith earth's magnetic field in standard
Coordinate system is (such as：World coordinate system) Y-direction vector value Yr, such three-dimensional attitude sensor is (such as：Three-dimensional electronic compass) in
MCU passes through relational expression：
Azimuth=arcTan (Y/X)
Azimuthal angle information can be accurately calculated.
By the angle of pitch, angle of heel and azimuthal angle information, can get image capture device during camera collection image
In 3 d pose information.In embodiments of the present invention, the 3 d pose of image capture device when obtaining camera collection image
Information specifically or passes through image capture device when plane compass and acceleration transducer Image Acquisition camera collection image
3 d pose information.Certainly in the embodiment of the present invention, the 3 d pose letter of image capture device when obtaining camera collection image
Breath includes the above-mentioned 3 d pose by image capture device during three-dimensional attitude sensor acquisition camera collection image listed
Information, and, the 3 d pose of image capture device when camera collection image is obtained by plane compass and acceleration transducer
Information, but it is not limited to above-mentioned listed method.
It should be noted that above-mentioned image capture device can include：Mobile phone, camera, video camera etc. can gather image
Equipment.If above-mentioned image capture device does not comprise three-dimensional attitude sensor, or, not comprising plane compass and acceleration transducer
When, external three-dimensional attitude sensor can be passed through, or, external plane compass and acceleration transducer are realizing obtaining 3 d pose
Information.
S102, the image being gathered based on described 3 d pose information, correction.
Based on 3 d pose information, the image of correction collection, thus obtain described image collecting device be in described
The non-distorted image of described object is gathered during the 3 d pose of 3 d pose equidistant vertical of object.Concrete as shown in Fig. 2
The image that step S101 collects can be as, in Fig. 2 shown in a-1 and a-2, in the image such as Fig. 2 after the correction of step S102 b-1 and
Shown in b2.When so can avoid camera collection image, during because of attitudes such as image capture device run-off the straight, rotations, cause
The image fault of collection.
It should be noted that the correct image to collection for step S102, can be specifically to the whole figure collecting
As being corrected.
S103, to correction after image carry out feature extraction process.
Carry out feature extraction process to the image after correction specifically can include：
Appoint based in Moravec operator, SUSAN operator, FAST operator, Harris operator and Harris-Laplace operator
A kind of detective operators, carry out feature detection to the image after correction；
Any one description, the feature to detection in description of description based on distribution, description of filtering and square
Carry out feature description.
Features described above detection and feature description are existing public technology, do not elaborate herein.
As a kind of optional embodiment, above-mentioned image capture device includes but is not limited to mobile phone, camera, video camera etc.
The equipment of image can be gathered.
The present embodiment, by gathering image, and obtains the 3 d pose letter of image capture device during camera collection image
Breath, based on described 3 d pose information, the image of correction collection, carries out feature extraction process to the image after correction, so may be used
To avoid local invariant feature under the attitudes such as image capture device run-off the straight, rotation to extract the problem of hydraulic performance decline.
Fig. 3 is a kind of schematic flow sheet of the second embodiment of image characteristic extracting method provided in an embodiment of the present invention,
As shown in figure 3, including：
S201, collection image, and obtain the angle of pitch of image capture device, angle of heel and orientation during camera collection image
The angle information at angle.
When step S201 obtains camera collection image, the 3 d pose information of image capture device includes：Obtain photographic head
The angle of pitch of image capture device, angle of heel and azimuthal angle information during collection image, the method for acquisition can adopt upper
The method that face embodiment provides, is not repeated explanation herein.
S202, based on the described angle of pitch, angle of heel and azimuthal angle information, geometry school is carried out to the image of collection
Just.
Geometric correction is carried out to the image of collection, detailed process is as follows：
Set up the mathematical model of geometric correction.
The mathematical model of the geometric correction set up is specifically as follows following mathematical model：
Wherein K is scale factor for image capture device internal reference matrix s, and α is the angle of pitch of image capture device, and β is to adopt
The angle of heel of collection equipment, θ=γ-γ0, γ is the azimuth of image capture device, γ0During for α=β=0, image capture device
Azimuth.(x ', y ') be image capture device the angle of pitch be α, angle of heel be β, azimuth be γ under the conditions of collect
Pixel on image g (x ', y '), g (x ', y ') is specially the image that image capture device collects.(x, y) is α=0, β=
0, azimuth is γ0Under the conditions of collection image f (x, y) on pixel, f (x, y) be specially image capture device reference
Image.Rx(β)、Ry(α)、Rz(θ) it is respectively the spin matrix around X, Y, Z axis rotation β, α, θ for the image capture device.
According to the angle of pitch obtaining, angle of heel and azimuthal angle information, described mathematics is determined by above-mentioned spin matrix
The model parameter of model, model parameter is specific as follows：
Azimuth is according to the model parameter of described mathematical model and described determination, the correct image to collection.Specifically
Process is as follows：
According to the model parameter of described mathematical model and described determination, space coordinate transformation is carried out to the image of collection.Tool
Body, set up collection the pixel point coordinates of image with reference to the corresponding relation formula between the corresponding point coordinates of figure.The image of collection
For g (x ', y '), the pixel point coordinates (x ', y ') of g (x ', y ') and be f (x, y) with reference to figure, the corresponding point coordinates (x, y) of f (x, y),
Also just set up (x ', y ') corresponding relation Yu (x, y) between, this corresponding relation is specific as follows：
Wherein, Rx -1(β)、Ry -1(α)、Rz -1(θ) it is respectively Rx(β)、Ry(α)、Rz(θ) inverse matrix, wherein K are adopted for image
Collection equipment internal reference matrix, s is scale factor.
Model parameter so according to above-mentioned determination and mathematical model draw the unknown parameter of above-mentioned corresponding relation formula, then
Set out from the image g (x ', y ') of collection, once calculate the calibration coordinate of each pixel according to above-mentioned relation formula, meanwhile, to sky
Between image after coordinate transform carry out gray scale interpolation process, give corresponding pixel grey scale pixel value, so generate a width school
Positive image..The pixel distribution of the image after often correcting is irregular, pixel extruding, the unequal phenomenon of density such as, no
Meet user's request, so by gray scale interpolation process is carried out to the image after correction, can be not with the element distribution of high-ranking officers' positive after image
The image of rule, is reduced into the real image of comparison.
S203, to correction after image carry out feature extraction process.
As a kind of optional embodiment, above-mentioned image capture device includes but is not limited to mobile phone, camera, video camera etc.
The equipment of image can be gathered.
The present embodiment, by gathering image, and obtains the 3 d pose letter of image capture device during camera collection image
Breath, the 3 d pose information of image capture device when obtaining camera collection image, based on described 3 d pose information, to collection
Image carry out space coordinate transformation and the image after space coordinate transformation carried out with gray scale interpolation process, then the figure after correcting
As carrying out feature extraction process, so can accurately correct the image of collection, with avoid image capture device run-off the straight,
Under the attitudes such as rotation, local invariant feature extracts the problem of hydraulic performance decline.
Fig. 4 is a kind of schematic flow sheet of the first embodiment of method for correcting image provided in an embodiment of the present invention, such as Fig. 4
Shown, including：
S301, the image of collection object, and obtain the 3 d pose letter of image capture device during camera collection image
Breath, the described photographic head of inclusion of described collecting device.
When obtaining camera collection image in the present embodiment, the 3 d pose information of image capture device specifically can adopt
The acquisition modes of above example description, are not repeated explanation herein.
S302, the image being gathered based on described 3 d pose information, correction, are obtained described image collecting device and are in and institute
The non-distorted image of described object is gathered during the 3 d pose of 3 d pose equidistant vertical stating object.
Based on 3 d pose information, the image of correction collection, the mode of correction specifically asks for an interview the inventive method embodiment two
Described.
By step S302 can just obtain described image collecting device be in equidistant with the 3 d pose of described object
The non-distorted image of described object is gathered during vertical 3 d pose.
Wherein, equidistant vertical refers to that the plane residing for the front of the photographic head of image capture device is equidistantly hung down with object
Directly, can be specifically that any position of object is all perpendicular to residing for the front of the camera structure of image capture device
Plane；The front of described photographic head is for gathering the face of image in photographic head.Described in actual application, image is adopted
Collection equipment be in 3 d pose with the 3 d pose equidistant vertical of described object it is simply that described image collecting device with respect to
Object do not exist inclination and rotate, when image capture device with respect to object do not exist tilt and rotate when gather to figure
As just for distortionless image.Specifically as shown in Fig. 2 the image that step S101 collects can be as a-1 and a-2 institute in Fig. 2
Show, the image after step S 102 correction is as shown in b-1 and b2 in Fig. 2.That is, as in Fig. 2 b-1 and b can be specifically
The non-distorted gathering during the 3 d pose of 3 d pose equidistant vertical of object in b-1 and b2 in described image equipment and Fig. 2
Image.After the correction of step S102, when can avoid camera collection image, because of image capture device run-off the straight, rotation
Etc. attitude when, cause gather image fault.
As a kind of optional embodiment, above-mentioned image capture device is in equidistant with the 3 d pose of described object
Vertical 3 d pose concrete it is also possible that the angle phase of the angle of heel of image capture device and the angle of heel of described object
With.
As a kind of optional embodiment, above-mentioned image capture device includes but is not limited to mobile phone, camera, video camera etc.
The equipment of image can be gathered.
The present embodiment, by gathering the image of object, and obtains three of image capture device during camera collection image
Dimension attitude information, based on described 3 d pose information, the image of correction collection, can obtain described image collecting device be in
The non-distorted image of described object is gathered during the 3 d pose of 3 d pose equidistant vertical of described object.So can keep away
When exempting from camera collection image, during because of attitudes such as image capture device run-off the straight, rotations, cause the image fault gathering.
Fig. 5 is a kind of schematic flow sheet of the second embodiment of method for correcting image provided in an embodiment of the present invention, such as Fig. 5
Shown, including：
S401, the image of collection object, and obtain the angle of pitch of image capture device, inclination during camera collection image
Angle and azimuthal angle information, the described photographic head of inclusion of described collecting device.
S402, set up the mathematical model of geometric correction.
The mathematical model of the geometric correction that step S402 is set up specifically asks for an interview the inventive method embodiment two.
S403, determined according to the described angle of pitch, angle of heel and azimuthal angle information described mathematical model model ginseng
Number.
Step S403 determines that the model parameter of described mathematical model specifically asks for an interview the inventive method embodiment two.
S404, the model parameter according to described mathematical model and described determination, carry out space coordinatess change to the image of collection
Change.
S405, the image after space coordinate transformation is carried out with gray scale interpolation process, with described image collecting device side
Inclination angle gathers the image of the non-distorted of described object when identical with the angle of the angle of heel of described object.
As a kind of optional embodiment, above-mentioned image capture device includes but is not limited to mobile phone, camera, video camera etc.
The equipment of image can be gathered.
The present embodiment, by gathering image, and obtains the 3 d pose letter of image capture device during camera collection image
Breath, based on described 3 d pose information, carries out space coordinate transformation and to the figure after space coordinate transformation to the image of collection
As carrying out gray scale interpolation process, to obtain the angle with the angle of heel of described object of the angle of heel that described image collecting device is in
Gather the image of the non-distorted of described object when spending identical, so can avoid in image capture device run-off the straight, rotation
Extract the problem of hydraulic performance decline etc. local invariant feature under attitude.
Fig. 6 is a kind of structural representation of the first embodiment of image capture device provided in an embodiment of the present invention, such as Fig. 6
Shown, including：First acquisition module 51, the first correction module 52 and extraction module 53, wherein：
Photographic head, for gathering image；
First acquisition module 51, for obtaining the 3 d pose information of image capture device during camera collection image.
It should be noted that the first acquisition module 51 is specifically additionally operable to obtain photographic head collection by three-dimensional attitude sensor
The 3 d pose information of image capture device during image, the first acquisition module 51 obtains equipment and gathers 3 d pose letter during image
Breath specifically includes the angle of pitch of image capture device, angle of heel and azimuthal angle information when obtaining camera collection image,
The described angle of pitch is for image capture device around conventional coordinates (such as：World coordinate system) Y-axis rotation the anglec of rotation, described inclination
Angle is for image capture device around conventional coordinates (such as：World coordinate system) X-axis rotation the anglec of rotation, described azimuth be image
The anglec of rotation that collecting device rotates around direct north.Concrete acquisition process is as follows：
Three-dimensional attitude sensor is (such as：Three-dimensional electronic compass) in three orthogonal three-dimensional magnetoresistive transducers survey respectively
Amount earth's magnetic field is in conventional coordinates (such as：World coordinate system) the vector value of X-direction and earth's magnetic field in conventional coordinates (such as：
World coordinate system) the vector value of Y direction and earth's magnetic field in conventional coordinates (such as：World coordinate system) Z-direction arrow
Value.Then the information transfer measuring three-dimensional magnetoresistive transducer is to three-dimensional attitude sensor (such as：Three-dimensional electronic compass) in
Micro-control unit (Micro Control Unit, MCU), MCU is by following relational expression：
Azimuth=arcTan (Y/X)
Wherein, Azimuth is the azimuth of image capture device.
Draw azimuthal angle information of image capture device during camera collection image, this azimuth is 3 d pose
Sensor is (such as：Three-dimensional electronic compass) anglec of rotation that rotates around direct north, that is, image capture device revolves around direct north
The anglec of rotation turning.
When image capture device run-off the straight, azimuthal accuracy will be greatly affected, this error big
The little size depending on image capture device location and inclination angle.For reducing the impact of this error, using 3 d pose
Sensor is (such as：Three-dimensional electronic compass) in double-shaft tilt angle sensor measuring image capture device during camera collection image
The angle of pitch and the angle information of angle of heel, this angle of pitch is for image capture device around conventional coordinates (such as：World coordinate system) Y
The anglec of rotation of axle rotation, angle of heel is for image capture device around conventional coordinates (such as：World coordinate system) X-axis rotation rotation
Angle.Double-shaft tilt angle sensor is by the information transfer of the angle of pitch and angle of heel to MCU, the conversion Calculation formula by following standard for the MCU：
Xr=Xcos α+Ysin α sin β-Zcos β sin α
Yr=Ycos β+Zsin β
Wherein, XrFor image capture device and ground surface earth's magnetic field when parallel in conventional coordinates (such as：World coordinates
System) X-direction vector value, YrFor image capture device and ground surface earth's magnetic field when parallel in conventional coordinates (such as：Generation
Boundary's coordinate system) Y direction vector value, α be collecting device gather image when the angle of pitch, β be collecting device collection image
When angle of heel.
Three-dimensional attitude sensor is (such as：Three-dimensional electronic compass) in MCU just can calculate image capture device and earth's surface
Earth's magnetic field when face is parallel is in conventional coordinates (such as：World coordinate system) X-direction vector value XrWith earth's magnetic field in standard
Coordinate system is (such as：World coordinate system) Y-direction vector value Yr, such three-dimensional attitude sensor is (such as：Three-dimensional electronic compass) in
MCU passes through relational expression：
Azimuth=arcTan (Y/X)
Azimuthal angle information can be accurately calculated.
By the angle of pitch, angle of heel and azimuthal angle information, can get image capture device during camera collection image
In 3 d pose information.In embodiments of the present invention, the first acquisition module 51 is specifically additionally operable to by plane compass and acceleration
The 3 d pose information of image capture device when degree sensor image obtains camera collection image.Certainly the embodiment of the present invention
In, when the first acquisition module 51 obtains camera collection image, the 3 d pose information of image capture device includes above-mentioned listing
The 3 d pose information of image capture device when camera collection image is obtained by three-dimensional attitude sensor, and, by plane
The 3 d pose information of image capture device when compass and acceleration transducer obtain camera collection image, but it is not limited to above-mentioned
Listed method.
It should be noted that above-mentioned image capture device can include：Mobile phone, camera, video camera etc. can gather image
Equipment.If above-mentioned image capture device does not comprise three-dimensional attitude sensor, or, not comprising plane compass and acceleration transducer
When, external three-dimensional attitude sensor can be passed through, or, external plane compass and acceleration transducer are realizing obtaining 3 d pose
Information.
First correction module 52, for the 3 d pose information being obtained based on acquisition module 51, is corrected acquisition module 51 and adopts
The image of collection.
Based on 3 d pose information, the image of correction collection, thus obtain described image collecting device be in
The non-distorted image of described object is gathered during the 3 d pose of 3 d pose equidistant vertical of described object.Concrete such as Fig. 2
Shown, the image that photographic head collects can be as, in Fig. 2 shown in a-1 and a-2, the image after the first correction module 52 correction is such as
In Fig. 2 shown in b-1 and b2.When so can avoid camera collection image, because of appearances such as image capture device run-off the straight, rotations
During state, cause the image fault gathering.Extraction module 53, carries for carrying out feature to the image after the first correction module 52 correction
Take process.
Extraction module 53 specifically can be also used for based on Moravec operator, SUSAN operator, FAST operator, Harris operator
With any one detective operators in Harris-Laplace operator, feature detection is carried out to the image after correction, and based on distribution
In description of description, description of filtering and square, any one description, carries out feature description to the feature of detection.
As a kind of optional embodiment, above-mentioned image capture device includes but is not limited to mobile phone, camera, video camera etc.
The equipment of image can be gathered.
The present embodiment, the 3 d pose letter of image capture device when camera collection image is obtained by the first acquisition module
Breath, correction module is based on described 3 d pose information, the image of correction collection, and extraction module carries out feature to the image after correction
Extraction process, so can avoid local invariant feature under the attitudes such as image capture device run-off the straight, rotation to extract performance
The problem declining.
Fig. 7 is a kind of structural representation of the second embodiment of image capture device provided in an embodiment of the present invention, such as Fig. 7
Shown, including：First acquisition module 61, the first correction module 62 and extraction module 63, wherein：
Photographic head, for gathering image；
First acquisition module 61, for obtaining the 3 d pose information of image capture device during camera collection image.
First acquisition module 61 is specifically for gathering image, and obtains camera collection image by three-dimensional attitude sensor
When the image capture device angle of pitch, angle of heel and azimuthal angle information, the described angle of pitch is image capture device around standard
Coordinate system is (such as：World coordinate system) Y-axis rotation the anglec of rotation, described angle of heel be image capture device around conventional coordinates
(such as：World coordinate system) X-axis rotation the anglec of rotation, the rotation that described azimuth rotates around direct north for image capture device
Angle.Specific acquisition process can be not repeated explanation using the acquisition process described in above example herein.
First correction module 6, for the angle of pitch, angle of heel and the azimuthal angle that are obtained based on the first acquisition module 61
Information, carries out geometric correction to the image of acquisition module 61 collection.
This first correction module 62 can include：First sets up model unit 621, the first computing unit 622 and the first school
Positive unit 623, wherein：
First sets up model unit 621, for setting up the mathematical model of geometric correction.
The mathematical model of the first geometric correction setting up model unit 621 foundation is specifically as follows following mathematical model：
Wherein, K is image capture device internal reference matrix, and s is scale factor, and α is the angle of pitch of image capture device, and β is
The angle of heel of collecting device, θ=γ-γ0, γ is the azimuth of image capture device, γ0During for α=β=0, image acquisition sets
Standby azimuth.(x ', y ') be image capture device the angle of pitch be α, angle of heel be β, azimuth be γ under the conditions of collect
Image g (x ', y ') on pixel, g (x ', y ') is specially the image that collects of image capture device.(x, y) is α=0, β
=0, azimuth is γ0Under the conditions of collection image f (x, y) on pixel, f (x, y) be specially image capture device ginseng
Examine image.Rx(β)、Ry(α)、Rz(θ) it is respectively the spin matrix around X, Y, Z axis rotation β, α, θ for the image capture device.
First computing unit 622, for the angle of pitch, angle of heel and the azimuthal angle that are obtained according to the first acquisition module 61
The 3 d pose information that degree information acquisition module obtains determines the model ginseng of the first mathematical model setting up model unit 621 foundation
Number.
The specific calculating process of first computing unit 62 can be, the angle of pitch that obtained by the first acquisition module 61, inclination
Angle and azimuthal angle information obtain calculating parameter R in above-mentioned mathematical modelx(β)、Ry(α)、Rz(θ), specific as follows：
First correction unit 623, calculates list for setting up the mathematical model of model unit 621 foundation and first according to first
The model parameter of unit 622 determination, the correct image to the collection of described acquisition module.
It should be noted that in another particular embodiment of the invention, the first correction unit 623 as shown in Figure 8 specifically may be used
To include：First coordinate transformation unit 6231 and the first gray scale interpolation processing unit 6232, wherein：
First coordinate transformation unit 6231, for setting up the mathematical model and first of model unit 621 foundation according to first
The model parameter that computing unit 622 determines, carries out space coordinate transformation to the image of collection.Particularly, set up the figure of collection
The pixel point coordinates of picture with reference to the corresponding relation formula between the corresponding point coordinates of figure.Adopt
The image integrating as g (x ', y '), the pixel point coordinates (x ', y ') of g (x ', y ') and be f (x, y) with reference to figure, f (x, y)
Corresponding point coordinates (x, y), also just sets up (x ', y ') corresponding relation Yu (x, y) between, and this corresponding relation is specific as follows：
Wherein, Rx -1(β)、Ry -1(α)、Rz -1(θ) it is respectively Rx(β)、Ry(α)、Rz(θ) inverse matrix, wherein K are adopted for image
Collection equipment internal reference matrix, s is scale factor.
Model parameter so according to above-mentioned determination and mathematical model draw the unknown parameter of above-mentioned corresponding relation formula, then
Set out from the image g (x ', y ') of collection, once calculate the calibration coordinate of each pixel according to above-mentioned relation formula, simultaneously picture
Plain gray value gives corresponding pixel, so generates a width correction chart picture.
First gray scale interpolation processing unit 6232, for the figure after the first coordinate transformation unit 6232 space coordinate transformation
As carrying out gray scale interpolation process.The pixel distribution of the image after often correcting is irregular, pixel extruding, density such as not
Impartial phenomenon, is unsatisfactory for user's request, so by gray scale interpolation process is carried out to the image after correction, can be with high-ranking officers' positive after image
Element distribution is irregular image, is reduced into the real image of comparison.
Extraction module 63, for carrying out feature extraction process to the image after the first correction module 62 correction.
As a kind of optional embodiment, above-mentioned image capture device includes but is not limited to mobile phone, camera, video camera etc.
The equipment of image can be gathered.
The present embodiment, the 3 d pose letter of image capture device when camera collection image is obtained by the first acquisition module
Breath, the angle of pitch, angle of heel and azimuthal angle information that correction module is obtained based on acquisition module, to acquisition module collection
Image carries out geometric correction, and extraction module carries out feature extraction process to the image after correction module correction, so can be accurate
The image of ground correction collection, to avoid local invariant feature extractability under the attitudes such as image capture device run-off the straight, rotation
The problem that can decline.
Fig. 9 is the structural representation that the embodiment of the present invention provides another kind of image capture device, this equipment include described in take the photograph
As head, as shown in figure 9, this equipment also includes：Second acquisition module 71 and the second correction module 72, wherein：
Photographic head, for gathering the image of object；
Second acquisition module 71, for obtaining the 3 d pose information of image capture device during camera collection image.
The first acquisition module that second acquisition module 71 specifically can be provided using above example.
Second correction module 72, for based on described 3 d pose information, the image of correction collection, obtaining described image and adopting
Collection equipment gathers the non-distorted figure of described object when being in the 3 d pose with the 3 d pose equidistant vertical of described object
Picture.
Described image collecting device can just be obtained by the second correction module 72 and be in the three-dimensional appearance with described object
The non-distorted image of described object is gathered during the 3 d pose of state equidistant vertical.Concrete as shown in Fig. 2 what photographic head collected
Image can be as, in Fig. 2 shown in a-1 and a-2, the image after the second correction module 72 correction is as shown in b-1 and b2 in Fig. 2.This
When sample can avoid camera collection image, during because of attitudes such as image capture device run-off the straight, rotations, cause the image gathering
Distortion.
As a kind of optional embodiment, the second correction module 72 is additionally operable to based on the described angle of pitch, angle of heel and side
The angle information of parallactic angle, carries out geometric correction to the image of collection, with the angle of heel that is in of described image collecting device with
The angle of the angle of heel of described object gathers the image of the non-distorted of described object when identical.
As a kind of optional embodiment, above-mentioned image capture device includes but is not limited to mobile phone, camera, video camera etc.
The equipment of image can be gathered.
The present embodiment, the 3 d pose letter of image capture device when camera collection image is obtained by the second acquisition module
Breath, the second correction module is based on described 3 d pose information, the image of correction collection, can obtain at described image collecting device
The non-distorted image of described object is gathered when the 3 d pose of the 3 d pose equidistant vertical with described object.So may be used
During avoiding camera collection image, during because of attitudes such as image capture device run-off the straight, rotations, the image gathering is caused to lose
Very.
Figure 10 is the structural representation of the second embodiment of another kind image capture device provided in an embodiment of the present invention, should
Equipment includes described photographic head, and as shown in Figure 10, this equipment also includes：Second acquisition module 81 and the second correction module 82, its
In, the second correction module 82 includes：Second sets up unit 821, the second computing unit 822, the second coordinate transformation unit 823 and
Two gray scale interpolation processing units 824, wherein：
Photographic head, for gathering the image of object；
Second acquisition module 81, for obtain the angle of pitch of image capture device during camera collection image, angle of heel and
Azimuthal angle information.
Second sets up unit 821, for setting up the mathematical model of geometric correction.
Second set up that unit 821 specifically can gather above example offer first set up unit.
Second computing unit 822, for determining described number according to the described angle of pitch, angle of heel and azimuthal angle information
Learn the model parameter of model.
Second computing unit 822 specifically can gather the first computing unit of above example offer.
Second coordinate transformation unit 823, for the model parameter according to described mathematical model and described determination, to collection
Image carries out space coordinate transformation.
Second coordinate transformation unit 822 specifically can gather the first coordinate transformation unit of above example offer.
Second gray scale interpolation processing unit 824, for the image after space coordinate transformation is carried out with gray scale interpolation process, with
Described image collecting device angle of heel identical with the angle of the angle of heel of described object when gather the non-of described object
The image of distortion.
As a kind of optional embodiment, above-mentioned image capture device includes but is not limited to mobile phone, camera, video camera etc.
The equipment of image can be gathered.
The present embodiment, the 3 d pose letter of image capture device when camera collection image is obtained by the second acquisition module
Breath, the second correction module is based on described 3 d pose information, carries out space coordinate transformation and to space seat to the image of collection
Image after mark conversion carries out gray scale interpolation process, with the angle of heel that is in of described image collecting device with described object
Angle of heel angle identical when gather described object non-distorted image, so can avoid sending out in image capture device
Under the attitudes such as raw inclination, rotation, local invariant feature extracts the problem of hydraulic performance decline.
One of ordinary skill in the art will appreciate that realizing all or part of flow process in above-described embodiment method, it is permissible
Instruct related hardware to complete by computer program, described program can be stored in a computer read/write memory medium
In, this program is upon execution, it may include as the flow process of the embodiment of above-mentioned each method.Wherein, described storage medium can be magnetic
Dish, CD, read-only memory (Read-Only Memory, ROM) or random access memory (Random Access
Memory, RAM) etc..
Above disclosed be only present pre-ferred embodiments, certainly the right model of the present invention can not be limited with this
Enclose, the equivalent variations therefore made according to the claims in the present invention, still belong to the scope that the present invention is covered.
Claims (18)
1. a kind of image characteristic extracting method is it is characterised in that include：
The image of collection object, and obtain the 3 d pose information of image capture device during camera collection image, described figure
As collecting device includes described photographic head；
Based on described 3 d pose information, the image rectification of collection is in and described object for described image collecting device
The non-distorted image of described object is gathered during the 3 d pose of 3 d pose equidistant vertical；
Feature extraction process is carried out to the non-distorted image of described object.
2. the method for claim 1 is it is characterised in that image capture device during described acquisition camera collection image
3 d pose information includes：
The 3 d pose information of image capture device when camera collection image is obtained by three-dimensional attitude sensor；Or,
The 3 d pose information of image capture device when camera collection image is obtained by plane compass and acceleration transducer.
3. method as claimed in claim 1 or 2 it is characterised in that during described acquisition camera collection image image acquisition set
Standby 3 d pose information includes：
The angle information of the angle of pitch of image capture device when obtaining camera collection image, the described angle of pitch sets for image acquisition
The anglec of rotation of the standby Y-axis rotation around conventional coordinates；
The angle information of the angle of heel of image capture device when obtaining camera collection image, described angle of heel sets for image acquisition
The anglec of rotation of the standby X-axis rotation around conventional coordinates；
Azimuthal angle information of image capture device when obtaining camera collection image, described azimuth sets for image acquisition
The standby anglec of rotation rotating around direct north.
4. method as claimed in claim 3 it is characterised in that described based on described 3 d pose information, will collection image
It is corrected to described image collecting device to be in described in collection during the 3 d pose with the 3 d pose equidistant vertical of described object
The non-distorted image of object includes：
Based on the described angle of pitch, described angle of heel and described azimuthal angle information, geometric correction is carried out to the image of collection,
The image rectification of collection is in the three-dimensional appearance with the 3 d pose equidistant vertical of described object for described image collecting device
The non-distorted image of described object is gathered during state.
5. method as claimed in claim 4 it is characterised in that described based on the described angle of pitch, described angle of heel and described side
The angle information of parallactic angle, carries out geometric correction to the image gathering and includes：
Set up the mathematical model of geometric correction；
Determine the model ginseng of described mathematical model according to the described angle of pitch, described angle of heel and described azimuthal angle information
Number；
According to the model parameter of described mathematical model and described determination, the correct image to collection.
6. method as claimed in claim 5 is it is characterised in that according to the model parameter of described mathematical model and described determination,
The correct image of collection is included：
According to the model parameter of described mathematical model and described determination, space coordinate transformation is carried out to the image of collection；
Image after space coordinate transformation is carried out with gray scale interpolation process.
7. a kind of method for correcting image is it is characterised in that include：
The image of collection object, and obtain the 3 d pose information of image capture device during camera collection image, described figure
As collecting device includes described photographic head；
Based on described 3 d pose information, the image rectification of collection is in and described object for described image collecting device
The non-distorted image of described object is gathered during the 3 d pose of 3 d pose equidistant vertical.
8. method as claimed in claim 7 it is characterised in that described based on described 3 d pose information, will collection image
It is corrected to described image collecting device to be in described in collection during the 3 d pose with the 3 d pose equidistant vertical of described object
The non-distorted image of object includes：
Based on the angle of pitch, angle of heel and azimuthal angle information, geometric correction is carried out to the image of collection, so that described figure
As gathering the non-distorted of described object when the angle with the angle of heel of described object of the angle of heel of collecting device is identical
Image.
9. method as claimed in claim 8 is it is characterised in that described believed based on the angle of pitch, angle of heel and azimuthal angle
Breath, carries out geometric correction to the image gathering and includes：
Set up the mathematical model of geometric correction；
Determine the model parameter of described mathematical model according to the described angle of pitch, angle of heel and azimuthal angle information；
According to the model parameter of described mathematical model and described determination, space coordinate transformation is carried out to the image of collection；
Image after space coordinate transformation is carried out with gray scale interpolation process, so that the angle of heel of described image collecting device and institute
State the angle of heel of object angle identical when gather described object non-distorted image.
10. a kind of image capture device, described image collecting device includes photographic head it is characterised in that described image collection sets
Standby also include：First acquisition module, the first correction module and extraction module, wherein：
Photographic head, for gathering the image of object；
First acquisition module, for obtaining the 3 d pose information of described image collecting device during camera collection image；
First correction module, for the 3 d pose information being obtained based on described first acquisition module, described photographic head is gathered
Image rectification adopt when being in the 3 d pose with the 3 d pose equidistant vertical of described object for described image collecting device
Collect the non-distorted image of described object；
Extraction module, the non-distorted image for the described object to described first correction module correction is carried out at feature extraction
Reason.
11. equipment as claimed in claim 10 are it is characterised in that described first acquisition module is additionally operable to pass by 3 d pose
The 3 d pose information of described image collecting device when sensor obtains camera collection image；Or, described first acquisition module
The three-dimensional appearance of described image collecting device when being additionally operable to obtain camera collection image by plane compass and acceleration transducer
State information.
12. equipment as claimed in claim 10 are it is characterised in that described first acquisition module is additionally operable to obtain photographic head collection
The angle of pitch of described image collecting device, angle of heel and azimuthal angle information during image, the described angle of pitch is image acquisition
Around the anglec of rotation of the Y-axis rotation of conventional coordinates, described angle of heel is that image capture device revolves around the X-axis of conventional coordinates to equipment
The anglec of rotation turning, the anglec of rotation that described azimuth rotates around direct north for image capture device.
13. equipment as claimed in claim 12 are it is characterised in that described first correction module is additionally operable to obtain based on described first
The angle of pitch, angle of heel and azimuthal angle information that delivery block obtains, carry out geometry school to the image of described photographic head collection
Just.
14. equipment as claimed in claim 13 are it is characterised in that described first correction module includes：
First sets up model unit, for setting up the mathematical model of geometric correction；
First computing unit, for the angle of pitch, angle of heel and the azimuthal angle letter that are obtained according to described first acquisition module
Breath determines the model parameter of the described first mathematical model setting up model unit foundation；
First correction unit, for setting up the mathematical model of model unit foundation and described first computing unit according to described first
The model parameter determining, the correct image to the collection of described photographic head.
15. equipment as claimed in claim 14 are it is characterised in that described first correction unit includes：
First coordinate transformation unit, for setting up the mathematical model of model unit foundation and described first calculating according to described first
The model parameter that unit determines, carries out space coordinate transformation to the image of collection；
First gray scale interpolation processing unit, for carrying out ash to the image after described first coordinate transformation unit space coordinate transformation
Degree interpolation is processed.
A kind of 16. image capture devices, described image collecting device includes photographic head it is characterised in that described image collection sets
Standby also include：Second acquisition module and the second correction module, wherein：
Photographic head, for gathering the image of object；
Second acquisition module, for obtaining the 3 d pose information of image capture device during camera collection image；
Second correction module, for based on described 3 d pose information, the image rectification of collection being described image collecting device
The non-distorted image of described object is gathered when being in the 3 d pose with the 3 d pose equidistant vertical of described object.
17. equipment as claimed in claim 16 are it is characterised in that described second correction module is additionally operable to based on the angle of pitch, side
Inclination angle and azimuthal angle information, carry out geometric correction to the image of collection, so that the inclination of described image collecting device
The angle with the angle of heel of described object at angle gathers the image of the non-distorted of described object when identical.
18. equipment as claimed in claim 17 are it is characterised in that described second correction module includes：
Second sets up unit, for setting up the mathematical model of geometric correction；
Second computing unit, for determining described mathematical model according to the described angle of pitch, angle of heel and azimuthal angle information
Model parameter；
Second coordinate transformation unit, for the model parameter according to described mathematical model and described determination, enters to the image of collection
Row space coordinate transformation；
Second gray scale interpolation processing unit, for gray scale interpolation process is carried out to the image after space coordinate transformation, so that institute
State image capture device angle of heel identical with the angle of the angle of heel of described object when gather the non-distorted of described object
Image.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201210222295.8A CN103514450B (en) | 2012-06-29 | 2012-06-29 | Image feature extracting method and image correcting method and equipment |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201210222295.8A CN103514450B (en) | 2012-06-29 | 2012-06-29 | Image feature extracting method and image correcting method and equipment |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN103514450A CN103514450A (en) | 2014-01-15 |
| CN103514450B true CN103514450B (en) | 2017-02-15 |
Family
ID=49897144
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201210222295.8A Expired - Fee Related CN103514450B (en) | 2012-06-29 | 2012-06-29 | Image feature extracting method and image correcting method and equipment |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN103514450B (en) |
Families Citing this family (18)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104951291B (en) * | 2014-03-25 | 2019-11-26 | 联想(北京)有限公司 | A kind of information processing method and electronic equipment |
| WO2016123813A1 (en) * | 2015-02-07 | 2016-08-11 | 华为技术有限公司 | Attitude relationship calculation method for intelligent device, and intelligent device |
| CN104902142B (en) * | 2015-05-29 | 2018-08-21 | 华中科技大学 | A kind of electronic image stabilization method of mobile terminal video |
| CN106257911A (en) * | 2016-05-20 | 2016-12-28 | 上海九鹰电子科技有限公司 | Image stability method and device for video image |
| WO2018200423A1 (en) | 2017-04-24 | 2018-11-01 | Magic Leap, Inc. | Tracking optical flow of backscattered laser speckle patterns |
| CN108364259A (en) * | 2017-09-29 | 2018-08-03 | 北京万维智能技术有限公司 | Method and system and equipment |
| CN109725117A (en) * | 2017-10-31 | 2019-05-07 | 青岛海尔智能技术研发有限公司 | The method and device that foodstuff calories detect in refrigerator |
| CN110966982A (en) * | 2018-09-28 | 2020-04-07 | 成都家有为力机器人技术有限公司 | Monocular camera ranging system and method for cleaning robot |
| CN109793576B (en) * | 2018-12-10 | 2021-09-28 | 湖北得康科技有限公司 | Visual device of intelligence and visual surgical instruments |
| CN109379536B (en) * | 2018-12-29 | 2020-07-31 | 深圳看到科技有限公司 | Picture generation method, device, terminal and corresponding storage medium |
| CN109917908B (en) * | 2019-02-01 | 2023-03-24 | 谷东科技有限公司 | Image acquisition method and system of AR glasses |
| CN109602451B (en) * | 2019-02-16 | 2021-08-10 | 河南科技大学第一附属医院（河南省显微外科研究所） | Medical ultrasonic auxiliary automatic diagnosis system |
| CN110208946A (en) * | 2019-05-31 | 2019-09-06 | 京东方科技集团股份有限公司 | A kind of wearable device and the exchange method based on wearable device |
| CN110398231B (en) * | 2019-06-18 | 2021-06-01 | 广东博智林机器人有限公司 | Wall surface parameter acquisition method and device, computer equipment and storage medium |
| CN111712857A (en) * | 2019-06-25 | 2020-09-25 | 深圳市大疆创新科技有限公司 | Image processing method, device, holder and storage medium |
| CN112702519B (en) * | 2020-12-24 | 2022-04-22 | 联想(北京)有限公司 | Image processing method and electronic equipment |
| CN112488136A (en) * | 2021-01-22 | 2021-03-12 | 山东商业职业技术学院 | Image recognition system and image recognition device |
| CN115359192B (en) * | 2022-10-14 | 2023-03-28 | 阿里巴巴（中国）有限公司 | Three-dimensional reconstruction and commodity information processing method, device, equipment and storage medium |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| EP1087339A1 (en) * | 1999-09-24 | 2001-03-28 | Ge Medical Systems Sa | Method of reconstruction of a three-dimensional image of an element of interest |
| CN102376089A (en) * | 2010-12-09 | 2012-03-14 | 深圳大学 | Target correction method and system |
| CN102436665A (en) * | 2011-08-25 | 2012-05-02 | 清华大学 | Two-dimensional plane representation method for images of alimentary tract |
-
2012
- 2012-06-29 CN CN201210222295.8A patent/CN103514450B/en not_active Expired - Fee Related
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| EP1087339A1 (en) * | 1999-09-24 | 2001-03-28 | Ge Medical Systems Sa | Method of reconstruction of a three-dimensional image of an element of interest |
| CN102376089A (en) * | 2010-12-09 | 2012-03-14 | 深圳大学 | Target correction method and system |
| CN102436665A (en) * | 2011-08-25 | 2012-05-02 | 清华大学 | Two-dimensional plane representation method for images of alimentary tract |
Also Published As
| Publication number | Publication date |
|---|---|
| CN103514450A (en) | 2014-01-15 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN103514450B (en) | Image feature extracting method and image correcting method and equipment | |
| CN112894832B (en) | Three-dimensional modeling method, three-dimensional modeling device, electronic equipment and storage medium | |
| CN106651942B (en) | Three-dimensional rotating detection and rotary shaft localization method based on characteristic point | |
| CN110211043B (en) | Registration method based on grid optimization for panoramic image stitching | |
| CN107830846A (en) | One kind utilizes unmanned plane and convolutional neural networks measurement communication tower aerial angle method | |
| CN108592950B (en) | Calibration method for relative installation angle of monocular camera and inertial measurement unit | |
| CN103514446B (en) | A kind of outdoor scene recognition methods merging sensor information | |
| CN106441138A (en) | Deformation monitoring method based on vision measurement | |
| CN103383773B (en) | The remote sensing satellite image of a kind of dynamic extraction Image Control Point is the most just penetrating framework and the method for correction | |
| CN107633536A (en) | A kind of camera calibration method and system based on two-dimensional planar template | |
| CN102622747B (en) | Camera parameter optimization method for vision measurement | |
| CN106971408B (en) | A kind of camera marking method based on space-time conversion thought | |
| CN110807815B (en) | Quick underwater calibration method based on corresponding vanishing points of two groups of mutually orthogonal parallel lines | |
| CN112254656B (en) | Stereoscopic vision three-dimensional displacement measurement method based on structural surface point characteristics | |
| CN102509304A (en) | Intelligent optimization-based camera calibration method | |
| CN106485753A (en) | Method and apparatus for the camera calibration of pilotless automobile | |
| CN107103056A (en) | A kind of binocular vision indoor positioning database building method and localization method based on local identities | |
| CN107886547A (en) | A kind of fisheye camera scaling method and system | |
| CN113012234B (en) | High-precision camera calibration method based on plane transformation | |
| CN106709500A (en) | Image feature matching method | |
| CN110146030A (en) | Side slope surface DEFORMATION MONITORING SYSTEM and method based on gridiron pattern notation | |
| CN110363179A (en) | Ground picture capturing method, device, electronic equipment and storage medium | |
| CN109448043A (en) | Standing tree height extracting method under plane restriction | |
| CN103353941B (en) | Natural marker registration method based on viewpoint classification | |
| CN105488541A (en) | Natural feature point identification method based on machine learning in augmented reality system |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C14 | Grant of patent or utility model | ||
| GR01 | Patent grant | ||
| CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20170215 Termination date: 20180629 |
|
| CF01 | Termination of patent right due to non-payment of annual fee |