CN107358587A - Image mending method and system - Google Patents
Image mending method and system Download PDFInfo
- Publication number
- CN107358587A CN107358587A CN201710565449.6A CN201710565449A CN107358587A CN 107358587 A CN107358587 A CN 107358587A CN 201710565449 A CN201710565449 A CN 201710565449A CN 107358587 A CN107358587 A CN 107358587A
- Authority
- CN
- China
- Prior art keywords
- cavity
- left view
- pixel
- view
- image
- 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
Links
- 238000000034 method Methods 0.000 title claims abstract description 32
- 238000004422 calculation algorithm Methods 0.000 claims abstract description 22
- 230000015572 biosynthetic process Effects 0.000 claims description 7
- 238000003786 synthesis reaction Methods 0.000 claims description 6
- 230000009466 transformation Effects 0.000 claims description 6
- 238000006243 chemical reaction Methods 0.000 claims description 5
- 230000009471 action Effects 0.000 claims description 2
- 230000000694 effects Effects 0.000 abstract description 3
- 238000012545 processing Methods 0.000 abstract description 3
- 238000009877 rendering Methods 0.000 abstract description 3
- 230000000007 visual effect Effects 0.000 description 14
- 238000013507 mapping Methods 0.000 description 10
- 230000008859 change Effects 0.000 description 5
- 150000001875 compounds Chemical class 0.000 description 4
- 238000010586 diagram Methods 0.000 description 3
- 230000008569 process Effects 0.000 description 3
- 238000013461 design Methods 0.000 description 2
- 238000006073 displacement reaction Methods 0.000 description 2
- 238000005516 engineering process Methods 0.000 description 2
- 238000002474 experimental method Methods 0.000 description 2
- 230000002194 synthesizing effect Effects 0.000 description 2
- 241000208340 Araliaceae Species 0.000 description 1
- 235000005035 Panax pseudoginseng ssp. pseudoginseng Nutrition 0.000 description 1
- 235000003140 Panax quinquefolius Nutrition 0.000 description 1
- 238000007796 conventional method Methods 0.000 description 1
- 238000009432 framing Methods 0.000 description 1
- 230000006870 function Effects 0.000 description 1
- 235000008434 ginseng Nutrition 0.000 description 1
- 238000003384 imaging method Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000008447 perception Effects 0.000 description 1
- 230000008439 repair process Effects 0.000 description 1
- 241000894007 species Species 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
- 239000011800 void material Substances 0.000 description 1
Classifications
-
- G06T5/77—
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T5/00—Image enhancement or restoration
- G06T5/50—Image enhancement or restoration by the use of more than one image, e.g. averaging, subtraction
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/50—Depth or shape recovery
- G06T7/55—Depth or shape recovery from multiple images
- G06T7/593—Depth or shape recovery from multiple images from stereo images
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/10—Image acquisition modality
- G06T2207/10004—Still image; Photographic image
- G06T2207/10012—Stereo images
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/20—Special algorithmic details
- G06T2207/20228—Disparity calculation for image-based rendering
Abstract
Provided herein image mending method and system, the application are related to technical field of image processing, wherein, the image mending method specifically includes：First, by being scanned in left view, to judge which of corresponding right view pixel is cavity, which secondly, after obtaining pixel and being cavity, cavity is filled up using the foreground point in left view and background dot, so, by the concrete operation step of above-mentioned image mending method, the caused empty reparation during left view is mapped to right view to DIBR algorithms can be completed, it is achieved thereby that showing the effect of good 3D rendering to spectators.
Description
Technical field
The present invention relates to technical field of image processing, more particularly to image mending method and system.
Background dot technology
In image processing process, it will usually using DIBR algorithms come compound stereoscopic image.DIBR full name
DepthImageBasedRendering, DIBR algorithm are a kind of virtual perspective mapping techniques, and the core of DIBR technologies is virtual
Mapping algorithm, its principle are the pixel position in color reference image to be passed through and the combination of depth information rearranges void
Intend the corresponding position in visual angle.It is come the object distance camera position in phenogram picture with the size of gray value in depth map
Distance, the span of the subject image vegetarian refreshments in depth map is set as 0~255, the point set depth nearer apart from camera
Value is bigger, and unlimited distance assumption value is 0.And it is a width Two-dimensional Color Image to refer to image, if reference picture plane definition
The plane formed for the X-axis in rectangular coordinate system and Y-axis, then, the Two-dimensional Color Image just constitutes the plane.And depth map
Z axis is then represented, and then a three-dimensional stereo model can be constructed by X-axis, Y-axis and Z axis, and then is shown most to people
Whole design sketch.
It can be seen that DIBR algorithms map to obtain other using Two-dimensional Color Image and its gray-scale map by similar triangle theory
The virtual image at visual angle, then further according to display feature compound stereoscopic image, further, it is possible to reference to demand change parallax
Method adjusts depth parameter, and so as to be adjusted the space or depth perception of 3D rendering, be finally reached spectators most preferably views and admires easypro
Appropriateness.Different from the method for changing a certain object's position in image in conventional method, DIBR algorithms will not change pair in scene
The proportionate relationship of far and near position as between.But during the virtual perspective image at other visual angles is synthesized, script visual angle
In the region that is blocked can be exposed and then form image cavity, i.e., the process of right view is being synthesized by left view and depth map
In can produce cavity, analyze its Producing reason and be, in left view point by depth map mapping as right view mistake
Cheng Zhong, two points or more points in left view can be mapped to same point, cause to produce cavity in right view, lead to
Often, the frequency occurred in foreground point and the alternate place cavity of background dot is higher, and foreground point is moved right by the mapping of depth map
A dynamic segment distance, background dot can also move right a segment distance, still, because the depth value of foreground point is big, the depth of background dot
It is worth small, so as to cause mobile distance different, produces cavity, and then be difficult to form natural appreciation effect.
To sum up, based on DIBR algorithm compound stereoscopic images during the problem of producing cavity, there is no at present effective
Solution.
The content of the invention
In view of this, the purpose of the embodiment of the present invention is the provision of image mending method and system, by using left view
Foreground point and background dot in figure are filled up to cavity, realize the reparation to cavity, improve the visual experience of spectators.
In a first aspect, the embodiments of the invention provide image mending method, including：By being scanned in left view,
It is cavity to judge which of right view pixel；
Cavity is filled up using the foreground point in left view and background dot.
With reference in a first aspect, the embodiments of the invention provide the possible embodiment of the first of first aspect, wherein, lead to
Cross and scanned in left view, judge which of right view pixel also includes before being cavity：
Obtain the left view of image；
Go out the depth map of image by left view transformation；
According to left view and the right view of depth map composograph.
With reference in a first aspect, the embodiments of the invention provide the possible embodiment of second of first aspect, wherein, lead to
Cross and scanned in left view, judge which of right view pixel is that cavity includes：
If the coordinate of pixel is (i, j) in left view, the conversion expression formula of DIBR algorithms is f (m, n), then in right view
The coordinate of respective pixel point is y=f (i, j)；
The coordinate of pixel scans in the range of being 2 points of (i-10, j) and (i+10, j) in left view；
The coordinate that pixel in right view is can not find in the range of is corresponding to y=f (l, k) during pixel, by right view
Middle coordinate is that y=f (l, k) pixel is determined as cavity.
With reference in a first aspect, the embodiments of the invention provide the possible embodiment of the third of first aspect, wherein,
Also include before being filled up using the foreground point in left view and background dot to cavity：
The gradient of the prospect of left view is sought, to obtain the gradient map of whole prospect；
Ask for the gray value of gradient map；
When gray value is more than threshold value set in advance, the cavity in right view corresponding to left view is determined as the first kind
Type cavity；
When gray value is less than or equal to threshold value set in advance, the cavity in right view corresponding to left view is judged
For Second Type cavity.
With reference to the third possible embodiment of first aspect, the embodiments of the invention provide the 4th of first aspect kind
Possible embodiment, wherein, also include after being filled up in the foreground point in using left view and background dot to cavity：
First kind cavity is filled up with the background dot in left view；
First kind cavity is filled up with the foreground point in left view and background dot.
Second aspect, the embodiments of the invention provide image mending system, including：Empty determination module, for by
Scanned in left view, it is cavity to judge which of right view pixel；
Hole-filling module, for being filled up using the foreground point in left view and background dot to cavity.
With reference to second aspect, the embodiments of the invention provide the possible embodiment of the first of second aspect, wherein, figure
As patch system also includes：
Left view acquisition module, for obtaining the left view of image；
Depth map acquisition module, for going out the depth map of image by left view transformation；
Synthesis module, for the right view according to left view and depth map composograph.
With reference to second aspect, the embodiments of the invention provide the possible embodiment of second of second aspect, wherein, figure
As patch system also includes：
Relation setting module, for setting the coordinate of pixel in left view as (i, j), the conversion expression formula of DIBR algorithms is
F (m, n), then the coordinate of respective pixel point is y=f (i, j) in right view；
Search module, in the range of the coordinate for the pixel in left view is 2 points of (i-10, j) and (i+10, j)
Scan for；
Empty determination module, for can not find the coordinate of pixel in right view in the range of as corresponding to y=f (l, k)
During pixel, the pixel that coordinate in right view is y=f (l, k) is determined as cavity.
With reference to second aspect, the embodiments of the invention provide the possible embodiment of the third of second aspect, wherein, figure
As patch system also includes：
Gradient map acquisition module, the gradient of the prospect for seeking left view, to obtain the gradient map of whole prospect；
Gray value asks for module, for asking for the gray value of gradient map；
First kind cavity determination module, for when gray value is more than threshold value set in advance, by corresponding to left view
Cavity in right view is determined as first kind cavity；
Second Type cavity determination module, for when gray value is less than or equal to threshold value set in advance, by left view
Cavity in right view corresponding to figure is determined as Second Type cavity.
With reference to the third possible embodiment of second aspect, the embodiments of the invention provide the 4th of second aspect kind
Possible embodiment, wherein, image mending system also includes：
First fills up module, for being filled up with the background dot in left view to first kind cavity；
Second fills up module, for being filled up with the foreground point in left view and background dot to first kind cavity.
Image mending method and system provided in an embodiment of the present invention, wherein, the image mending method includes：First, lead to
Cross and scanned in left view, it is cavity to judge which of right view pixel, completes the positioning to cavity, afterwards, profit
Cavity is filled up with the foreground point in left view and background dot, to realize to reparation empty in image, so, Guan Zhongneng
Good visual experience is obtained from the image after reparation.
Other features and advantages of the present invention will illustrate in the following description, also, partly become from specification
Obtain it is clear that or being understood by implementing the present invention.The purpose of the present invention and other advantages are in specification, claims
And specifically noted structure is realized and obtained in accompanying drawing.
To enable the above objects, features and advantages of the present invention to become apparent, preferred embodiment cited below particularly, and coordinate
Appended accompanying drawing, is described in detail below.
Brief description of the drawings
, below will be to specific in order to illustrate more clearly of the specific embodiment of the invention or technical scheme of the prior art
The required accompanying drawing used is briefly described in embodiment or description of the prior art, it should be apparent that, in describing below
Accompanying drawing is some embodiments of the present invention, for those of ordinary skill in the art, before creative work is not paid
Put, other accompanying drawings can also be obtained according to these accompanying drawings.
Fig. 1 shows the schematic diagram for the DIBR algorithms that prior art is provided；
Fig. 2 shows the mapping schematic diagram for the DIBR algorithms that prior art is provided；
Fig. 3 shows the flow chart for the image mending method that the embodiment of the present invention is provided；
Fig. 4 shows the connection figure for the image mending system that the embodiment of the present invention is provided；
Fig. 5 shows the structural framing figure for the image mending system that the embodiment of the present invention is provided；
Fig. 6 shows the structure connection figure for the image mending system that the embodiment of the present invention is provided.
Icon：1- cavities determination module；2- hole-filling modules；3- left view acquisition modules；4- depth map acquisition modules；
5- synthesis modules；6- relation setting modules；7- search modules；8- searches determination module in cavity.
Embodiment
Below in conjunction with accompanying drawing in the embodiment of the present invention, the technical scheme in the embodiment of the present invention is carried out clear, complete
Ground describes, it is clear that described embodiment is only part of the embodiment of the present invention, rather than whole embodiments.Generally exist
The component of the embodiment of the present invention described and illustrated in accompanying drawing can be configured to arrange and design with a variety of herein.Cause
This, the detailed description of the embodiments of the invention to providing in the accompanying drawings is not intended to limit claimed invention below
Scope, but it is merely representative of the selected embodiment of the present invention.Based on embodiments of the invention, those skilled in the art are not doing
The every other embodiment obtained on the premise of going out creative work, belongs to the scope of protection of the invention.
At present, during stereopsis is obtained using DIBR algorithms, will not change between the object in scene
The proportionate relationship of far and near position, it mainly adjusts depth parameter by using the method for changing parallax, so that 3D rendering regards
Feel that depth is adjusted, and reach the optimal of spectators and view and admire comfort level.But synthesizing the virtual perspective image at other visual angles
During, the region being blocked in script visual angle can be exposed and then form image cavity, i.e., by left view and depth map
Cavity can be produced during synthesis right view, because mistake of the point in left view by the mapping of depth map as right view
Cheng Zhong, two points or more points in left view can be mapped to same point, cause to produce cavity in right view, sternly
Ghost image rings the visual effect of spectators.
The core of DIBR algorithms is virtual map algorithm, and its principle is that the pixel position in color reference image is passed through
And the combination of depth information is rearranged on the correspondence position in virtual perspective.Depth map be a width with the size of gray value come
Object in the phenogram picture two dimensional image far and near from camera position, the span of its subject image vegetarian refreshments are set as 0~255,
The point set depth value nearer apart from camera is bigger, and unlimited distance assumption value is 0.It is a width Two-dimensional Color Image with reference to image,
If it is the plane that the X-axis in rectangular coordinate system is formed with Y-axis reference picture plane definition, then, depth map then represents Z
Axle, a three-dimensional stereo model is constructed by X-axis, Y-axis and Z axis can, it maps schematic diagram as depicted in figs. 1 and 2.In figure
P points are the destination objects of observation, and what Z was represented is the depth of P points, and Cc directions represent the visual angle of two-dimentional reference picture, Cl, Cr difference
The visual angle of left and right virtual image to be generated is represented, f is focal length, and tx is the interval of two virtual perspectives, and left and right visual angle is based on ginseng
It is symmetrical to examine visual angle.In above-mentioned three-dimensional model, image in originally (xc, y)) P points, now project (xl, the y) of imaging plane,
2 points of (xr, y).With the principle of similar triangles, horizontal displacement value xr, xl can be calculated according to formula (1), by the formula
It can draw, horizontal parallax (xr, xl) is proportional relationship between tx.Therefore, there are some successively at present according to video pictures
The method that classification is identified in feature.But the main flow three-dimensional video-frequency form species that these methods can be distinguished and identified is endless
Entirely, also, method robustness it is not strong, higher False Rate be present.
Based on this, the embodiments of the invention provide image mending method and system, it is described below by embodiment.
Embodiment 1
Referring to Fig. 3, the image mending method that the present embodiment proposes specifically includes following steps：
Step S101：By being scanned in left view, it is cavity to judge which of right view pixel.
I.e. during with DIBR algorithm compound stereoscopic images, first, the left view of image is got, secondly,
Go out the depth map of image by the left view transformation got, afterwards, closed with DIBR algorithms according to left view and depth map
Into the right view of image, to form stereo-picture eventually through above-mentioned left view, depth map and right view.
But the process that right view is synthesized by left view and depth map can produce cavity, concrete scene caused by cavity
It is as follows, during the point in left view turns into right viewpoint by the mapping of depth map, two points in left view or more
Multiple spot is mapped to same point, causes empty generation in right view.By theory analysis and many experiments exploration,
Cavity is generally present in that prospect background is alternately local, and prospect is moved right a segment distance by the mapping of depth map, background
Can be moved right a segment distance, but because the depth value of prospect is big, the depth value of background is small to cause mobile distance different,
The fracture of pixel occurs, produces cavity, i.e., does not have corresponding mapping point in right view.
In this image mending method, by being scanned in left view, judge which of right view pixel is
Cavity includes：
First, if the coordinate of pixel is (i, j) in left view, the conversion expression formula of DIBR algorithms is f (m, n), then right
The coordinate of respective pixel point is y=f (i, j) in view, and after above-mentioned definition is completed, the coordinate of pixel is in left view
Scanned in the range of 2 points of (i-10, j) and (i+10, j), secondly, pixel in right view is can not find in the range of
Coordinate is corresponding to y=f (l, k) during pixel, and the pixel that coordinate in right view is y=f (l, k) is determined as into cavity.
Step S102：Cavity is filled up using the foreground point in left view and background dot.
In the method, also include before being filled up using the foreground point in left view and background dot to cavity：First,
The gradient of the prospect of left view is sought, to obtain the gradient map of whole prospect, i.e., whole prospect is stated by two-dimensional discrete function,
Secondly, the gray value of gradient map is asked for, because each pixel only has the image of a sample color in gray level image, therefore, leads to
Cross and ask for gray value to be judged the most accurately, afterwards, when gray value is more than threshold value set in advance, here, it is necessary to say
Bright, the setting of threshold value is the empirical value obtained according to the exploration of many experiments, and cavity can be effectively divided as boundary
Type, the cavity in right view corresponding to left view is determined as first kind cavity, i.e., foreground point and background dot be all to the right
Mobile, simply displacement is different, and this empty origin cause of formation is caused by mainly due to the movement of background dot, when gray value is less than
Or during equal to threshold value set in advance, the cavity in right view corresponding to left view is determined as Second Type cavity, it is this kind of
Cavity is due to that the depth value of the corresponding depth map of foreground point causes greatly mapping forward movement of changing the time very much, and background dot is moved rearwards,
So form cavity.
Action when being filled up to cavity specifically includes：First, for first kind cavity mainly due to background dot
Movement cause, just take the background of the relevant position point of left view to fill up, i.e., it is empty to the first kind with the background dot in left view
Hole is filled up.It is very big mainly due to the depth value of the corresponding depth map of foreground point for Second Type cavity, cause the point to reflect
Forward movement is penetrated, background dot is moved rearwards what is formed, foreground point in the left view of cavity in this case and background dot
First kind cavity is filled up.
In summary, the image mending method that the present embodiment provides includes：Firstly, it is necessary to caused sky during mapping
Hole is positioned, i.e., is scanned in left view, and it is cavity to judge which of right view pixel, afterwards, utilizes left view
Foreground point and background dot in figure are filled up to cavity, further, it is possible to use background dot respectively according to the particular type in cavity
And foreground point and background dot carry out the reparation in cavity, the stereopsis complete display obtained using DIBR algorithms, to spectators
Form natural appreciation effect.
Embodiment 2
Referring to Fig. 4, Fig. 5 and Fig. 6, present embodiments providing image mending system includes：The cavity being sequentially connected judges mould
Block 1 and hole-filling module 2, during work, empty determination module 1 is used to, by scanning in left view, judge right view
Which of pixel be cavity, hole-filling module 2 is used for using the foreground point in left view and background dot to cavity progress
Fill up.
In addition, image mending system also includes：Left view acquisition module 3, depth map acquisition module 4 and the conjunction being sequentially connected
Into module 5, left view acquisition module 3 is used for the left view for obtaining image, and depth map acquisition module 4 is used to be gone out by left view transformation
The depth map of image, synthesis module 5 are used for the right view according to left view and depth map composograph.
In addition, image mending system also includes：Relation setting module 6, search module 7 and the cavity being sequentially connected judge mould
Block 1, during work, as (i, j), the map table of DIBR algorithms reaches the coordinate that relation setting module 6 is used to set pixel in left view
Formula is f (m, n), then the coordinate of respective pixel point is y=f (i, j) in right view, and search module 7 is used for the pixel in left view
The coordinate of point scans in the range of being 2 points of (i-10, j) and (i+10, j), and cavity searches determination module 8 and is used to work as model
The coordinate that pixel in right view is can not find in enclosing is corresponding to y=f (l, k) during pixel, is y=f by coordinate in right view
The pixel of (l, k) is determined as cavity.
In addition, image mending system also includes：Gradient map acquisition module, the gray value being sequentially connected ask for module, first
Type cavity determination module and Second Type cavity determination module, during work, gradient map acquisition module is used for before seeking left view
The gradient of scape, to obtain the gradient map of whole prospect, gray value asks for the gray value that module is used for asking for gradient map, the first kind
Empty determination module 1 is used for when gray value is more than threshold value set in advance, and the cavity in right view corresponding to left view is sentenced
It is set to first kind cavity, Second Type cavity determination module 1 is used for when gray value is less than or equal to threshold value set in advance
When, the cavity in right view corresponding to left view is determined as Second Type cavity.
In addition, image mending system also includes：First be sequentially connected fills up module and second and fills up module, in use,
First to fill up module be that first kind cavity is filled up using the background dot in left view, and second to fill up module be to utilize a left side
Foreground point and background dot in view are filled up to first kind cavity.
In summary, the image mending system that the present embodiment provides includes：The empty determination module 1 being sequentially connected and cavity
Fill up module 2, during work, empty determination module 1 is used to, by scanning in left view, judge which of right view picture
Vegetarian refreshments is cavity, and hole-filling module 2 is used to fill up cavity using the foreground point in left view and background dot, by upper
State the setting of modules, can to using DIBR algorithms carry out image synthesizing procedure in caused cavity be accurately positioned and
Repair, so that good visual effect is presented in the image of synthesis.
Finally it should be noted that：Embodiment described above, it is only the embodiment of the present invention, to illustrate the present invention
Technical scheme, rather than its limitations, protection scope of the present invention is not limited thereto, although with reference to the foregoing embodiments to this hair
It is bright to be described in detail, it will be understood by those within the art that：Any one skilled in the art
The invention discloses technical scope in, it can still modify to the technical scheme described in previous embodiment or can be light
Change is readily conceivable that, or equivalent substitution is carried out to which part technical characteristic；And these modifications, change or replacement, do not make
The essence of appropriate technical solution departs from the spirit and scope of technical scheme of the embodiment of the present invention, should all cover the protection in the present invention
Within the scope of.Therefore, protection scope of the present invention described should be defined by scope of the claims.
Claims (10)
1. image mending method, it is characterised in that including：
By being scanned in left view, it is cavity to judge which of right view pixel；
The cavity is filled up using the foreground point in the left view and background dot.
2. image mending method according to claim 1, it is characterised in that it is described by being scanned in left view,
Judge which of right view pixel also includes before being cavity：
Obtain the left view of image；
Go out the depth map of described image by the left view transformation；
According to the left view and the right view of the depth map composograph.
3. image mending method according to claim 1, it is characterised in that it is described by being scanned in left view,
Judge which of right view pixel is that cavity includes：
If the coordinate of pixel is (i, j) in the left view, the conversion expression formula of DIBR algorithms is f (m, n), then the right side regards
The coordinate of respective pixel point is y=f (i, j) in figure；
The coordinate of pixel scans in the range of being 2 points of (i-10, j) and (i+10, j) in the left view；
When the coordinate that pixel in the right view is can not find in the scope is pixel corresponding to y=f (l, k), by institute
It is that y=f (l, k) pixel is determined as cavity to state coordinate in right view.
4. image mending method according to claim 1, it is characterised in that in the foreground point using in left view and
Background dot also includes before being filled up to the cavity：
The gradient of the prospect of the left view is sought, to obtain the gradient map of whole prospect；
Ask for the gray value of the gradient map；
When the gray value is more than threshold value set in advance, the cavity in the right view corresponding to the left view is judged
For first kind cavity；
When the gray value is less than or equal to threshold value set in advance, by the right view corresponding to the left view
Cavity is determined as Second Type cavity.
5. image mending method according to claim 4, it is characterised in that in the foreground point using in left view and
Background dot also includes after being filled up to the cavity：
The first kind cavity is filled up with the background dot in the left view；
The first kind cavity is filled up with the foreground point in the left view and background dot.
6. image mending system, it is characterised in that including：
Empty determination module, for by being scanned in left view, it to be cavity to judge which of right view pixel；
Hole-filling module, for being filled up using the foreground point in the left view and background dot to the cavity.
7. image mending system according to claim 6, it is characterised in that also include：
Left view acquisition module, for obtaining the left view of image；
Depth map acquisition module, for going out the depth map of described image by the left view transformation；
Synthesis module, for the right view according to the left view and the depth map composograph.
8. image mending system according to claim 6, it is characterised in that also include：
Relation setting module, for setting the coordinate of pixel in the left view as (i, j), the conversion expression formula of DIBR algorithms is
F (m, n), then the coordinate of respective pixel point is y=f (i, j) in the right view；
Search module, in the range of the coordinate for the pixel in the left view is 2 points of (i-10, j) and (i+10, j)
Scan for；
Determination module is searched in cavity, for when can not find in the scope coordinate of pixel in the right view for y=f (l,
K) corresponding to during pixel, the pixel that coordinate in the right view is y=f (l, k) is determined as cavity.
9. image mending system according to claim 6, it is characterised in that also include：
Gradient map acquisition module, the gradient of the prospect for seeking the left view, to obtain the gradient map of whole prospect；
Gray value asks for module, for asking for the gray value of the gradient map；
First kind cavity determination module, for when the gray value is more than threshold value set in advance, by the left view pair
Cavity in the right view answered is determined as first kind cavity；
Second Type cavity determination module, described in when the gray value is less than or equal to threshold value set in advance, inciting somebody to action
Cavity in the right view corresponding to left view is determined as Second Type cavity.
10. image mending system according to claim 9, it is characterised in that also include：
First fills up module, for being filled up with the background dot in the left view to the first kind cavity；
Second fills up module, for being filled out with the foreground point in the left view and background dot to the first kind cavity
Mend.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710565449.6A CN107358587A (en) | 2017-07-12 | 2017-07-12 | Image mending method and system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710565449.6A CN107358587A (en) | 2017-07-12 | 2017-07-12 | Image mending method and system |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN107358587A true CN107358587A (en) | 2017-11-17 |
Family
ID=60292007
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201710565449.6A Pending CN107358587A (en) | 2017-07-12 | 2017-07-12 | Image mending method and system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN107358587A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111679777A (en) * | 2020-06-12 | 2020-09-18 | 贝壳技术有限公司 | Panorama generating method and system and object model generating method and system |
Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101808251A (en) * | 2010-04-06 | 2010-08-18 | 浙江大学 | Method for extracting blocking information in stereo image pair |
| CN103369342A (en) * | 2013-08-05 | 2013-10-23 | 重庆大学 | Method for inpainting and restoring processing of vacancy of DIBR (Depth Image Based Rendering) target image |
| CN104954780A (en) * | 2015-07-01 | 2015-09-30 | 南阳师范学院 | DIBR (depth image-based rendering) virtual image restoration method applicable to high-definition 2D/3D (two-dimensional/three-dimensional) conversion |
| KR101575997B1 (en) * | 2015-01-02 | 2015-12-11 | 한국과학기술원 | Method and device for detection of depth-image-based rendering 3d image |
| CN105163104A (en) * | 2015-08-31 | 2015-12-16 | 四川虹微技术有限公司 | Intermediate view synthesis method without generating cavity |
| KR20160024419A (en) * | 2014-08-25 | 2016-03-07 | 국방과학연구소 | System and Method for identifying stereo-scopic camera in Depth-Image-Based Rendering |
| CN106028020A (en) * | 2016-06-21 | 2016-10-12 | 电子科技大学 | Multi-direction prediction based virtual visual-angle image cavity filling method |
| CN106651871A (en) * | 2016-11-18 | 2017-05-10 | 华东师范大学 | Automatic filling method for cavities in depth image |
-
2017
- 2017-07-12 CN CN201710565449.6A patent/CN107358587A/en active Pending
Patent Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101808251A (en) * | 2010-04-06 | 2010-08-18 | 浙江大学 | Method for extracting blocking information in stereo image pair |
| CN103369342A (en) * | 2013-08-05 | 2013-10-23 | 重庆大学 | Method for inpainting and restoring processing of vacancy of DIBR (Depth Image Based Rendering) target image |
| KR20160024419A (en) * | 2014-08-25 | 2016-03-07 | 국방과학연구소 | System and Method for identifying stereo-scopic camera in Depth-Image-Based Rendering |
| KR101575997B1 (en) * | 2015-01-02 | 2015-12-11 | 한국과학기술원 | Method and device for detection of depth-image-based rendering 3d image |
| CN104954780A (en) * | 2015-07-01 | 2015-09-30 | 南阳师范学院 | DIBR (depth image-based rendering) virtual image restoration method applicable to high-definition 2D/3D (two-dimensional/three-dimensional) conversion |
| CN105163104A (en) * | 2015-08-31 | 2015-12-16 | 四川虹微技术有限公司 | Intermediate view synthesis method without generating cavity |
| CN106028020A (en) * | 2016-06-21 | 2016-10-12 | 电子科技大学 | Multi-direction prediction based virtual visual-angle image cavity filling method |
| CN106651871A (en) * | 2016-11-18 | 2017-05-10 | 华东师范大学 | Automatic filling method for cavities in depth image |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111679777A (en) * | 2020-06-12 | 2020-09-18 | 贝壳技术有限公司 | Panorama generating method and system and object model generating method and system |
| CN111679777B (en) * | 2020-06-12 | 2021-09-07 | 贝壳技术有限公司 | Panorama generating method and system and object model generating method and system |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US9288476B2 (en) | System and method for real-time depth modification of stereo images of a virtual reality environment | |
| KR102653808B1 (en) | Methods and systems for detecting and combining structural features in 3d reconstruction | |
| US9407904B2 (en) | Method for creating 3D virtual reality from 2D images | |
| EP3242275B1 (en) | Using photo collections for three dimensional modeling | |
| CN104780355B (en) | Empty restorative procedure based on the degree of depth in a kind of View Synthesis | |
| US7505623B2 (en) | Image processing | |
| WO2015188684A1 (en) | Three-dimensional model reconstruction method and system | |
| CN105262958B (en) | A kind of the panorama feature splicing system and its method of virtual view | |
| CN102325259A (en) | Method and device for synthesizing virtual viewpoints in multi-viewpoint video | |
| CN103971408A (en) | Three-dimensional facial model generating system and method | |
| CN109147027B (en) | Monocular image three-dimensional rebuilding method, system and device based on reference planes | |
| CN111047709B (en) | Binocular vision naked eye 3D image generation method | |
| CN109769109A (en) | Method and system based on virtual view synthesis drawing three-dimensional object | |
| CN113253845A (en) | View display method, device, medium and electronic equipment based on eye tracking | |
| Pacheco et al. | Reconstruction of high resolution 3D objects from incomplete images and 3D information | |
| CN112102504A (en) | Three-dimensional scene and two-dimensional image mixing method based on mixed reality | |
| CN107358587A (en) | Image mending method and system | |
| CN102186093B (en) | Stereo image generation method and system | |
| CN117061720B (en) | Stereo image pair generation method based on monocular image and depth image rendering | |
| Eggert et al. | Multi-layer visualization of mobile mapping data | |
| CN117115274B (en) | Method, device, equipment and storage medium for determining three-dimensional information | |
| Futragoon et al. | 3D reconstruction of a collapsed historical site from sparse set of photographs and photogrammetric map | |
| CN115205493A (en) | Single-key-generated multi-view three-dimensional reconstruction method and device | |
| Lin et al. | Complete 3D model reconstruction from multiple views | |
| Yaopu et al. | Research on Construction Technology of 3D Scene Point Cloud Model Based on Improved PMVS Algorithm |
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 | ||
| RJ01 | Rejection of invention patent application after publication | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20171117 |