CN103337081B - A kind of shadowing method based on depth layer and device - Google Patents
A kind of shadowing method based on depth layer and device Download PDFInfo
- Publication number
- CN103337081B CN103337081B CN201310294784.9A CN201310294784A CN103337081B CN 103337081 B CN103337081 B CN 103337081B CN 201310294784 A CN201310294784 A CN 201310294784A CN 103337081 B CN103337081 B CN 103337081B
- Authority
- CN
- China
- Prior art keywords
- depth
- cavity
- minimum
- interlayer
- histogram
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Landscapes
- Image Analysis (AREA)
Abstract
The invention discloses a kind of shadowing method based on depth layer and device thereof, belong to computer vision field.The method comprises: carry out Histogram statistics to the depth image of original image, the number of pixels corresponding to all gray-scale values of statistics depth image; Carry out envelope detected according to above-mentioned histogram, obtain deep statistical envelope diagram; Neighbor smoothing filtering is done to above-mentioned deep statistical envelope diagram; According to level and smooth deep statistical envelope diagram, calculate minimum-depth saltus step, thus calculate minimum position translational movement on virtual view; Using above-mentioned translational movement as pore size threshold value, different empty filling Strategy is taked to different size cavity.This device comprises: degree of depth statistics with histogram module, envelope diagram computing module, depth layer computing module, empty size computing module and empty size judge module.Invention not only simplifies the complicacy of existing Rendering algorithms, accelerate the speed played up, and greatly strengthen the accuracy of empty type judgement.
Description
Technical field
The present invention relates to computer vision field, particularly a kind of shadowing method based on depth layer and device thereof.
Background technology
In recent years, along with the continuous breakthrough of free view-point technology in hardware and software research, free view-point is presented at medical domain, and the field such as Exploration Domain or show business all receives increasing favor.
The better effects if shown to make free view-point, emerged in large numbers the Rendering algorithms of multiple image, common Rendering algorithms is roughly divided into two classes:
The first kind is former figure plus depth figure Rendering algorithms (2DplusZ), and mainly corresponding in depth map according to scene point depth value utilizes the inside and outside parameter matrix computations of camera, first projects retroeflection again and plays up the virtual view view made new advances.
Equations of The Second Kind is multi views Rendering algorithms (multi-view), mainly renders the view of scene visual angle according to known multiple views.
When playing up view, cause the generation in cavity because of blocking between scene point.As shown in Figure 1, the corresponding left view viewpoint of C1, the corresponding right view viewpoint of C2.Scene point P1 and P2 is projected in same position P in left view L, like this, will be blocked from the scene point P1 away from left viewpoint C1 by a P2.Scene point P1 and P2 is projected on position P1 ' respectively again on right view R, P2 ', and like this, if when carrying out playing up of right view based on left view and depth map, the P1 ' point on right view R will occur cavity because not having corresponding information.
Solve the main method of empty problem and have two large classes: a class utilizes background, texture, and neighbor etc. are filled cavity, as Chinese patent CN200810046313.5; Another kind carries out pre-service to depth map, main smoothing process, the uncontinuity of depth of smoothness, with the cavity produced after reducing Image Rendering, as Chinese patent CN201010144951.8.
But above-mentioned existing empty type judgment technology at least has following shortcoming:
(1) when playing up image, above-mentioned two class methods all do not carry out classification process to cavity; Cavity is divided into two classes obviously, and a class is the little cavity that quantization error is brought, and is generally several pixel; One class is that depth jump namely blocks the macroscopic-void brought, and is generally tens pixels.
(2) when the empty problem of solution, the filling technique of the first kind can well process the small holes of quantization error generation, because can not inherently solve empty problem, particularly fills up the larger cavity of blocking and causing, the obvious distortion of meeting, fills up effect undesirable; Equations of The Second Kind antialiasing is out of shape object, although decrease the quantity of macroscopic-void, the visual quality of the virtual view obtained is significantly decreased.
Summary of the invention
For the defect existed in above-mentioned prior art, in order to strengthen the rendering effect of image, the object of this invention is to provide a kind of method judging virtual view cavity type, classification process can be carried out to cavity.Another object of the present invention is to provide the device realizing the method.
In order to realize foregoing invention object, the technical scheme that the inventive method adopts is as follows:
Based on a shadowing method for depth layer, comprise the steps:
(1) depth map of reference-view is chosen;
(2) number of pixels corresponding to all gray-scale values of described depth map is added up, draw degree of depth histogram;
(3) envelope detected is carried out to described degree of depth histogram, draw out initial envelope diagram;
(4) neighbor smoothing filtering is carried out to the histogrammic initial envelope diagram of the described degree of depth, then draw filtered envelope diagram;
(5) peakvalue's checking is carried out to described filtered envelope diagram, find out a point peak position;
(6) to described filtered envelope diagram, according to described point of peak position, mark each depth layer, between two difference done to described depth layer, find out minimum-depth interlayer every;
(7) obtain virtual view, determine the position of virtual view to be rendered;
(8) calculate the distance of described virtual view relative to described reference-view, according to the distance calculated, utilize described minimum-depth interlayer every calculating minimum translation size, namely minimum-depth interlayer is interposed between the value of the position translation that virtual view causes;
(9) carry out cavity to the virtual view obtained to detect, find out all cavities, calculate the catercorner length in cavity;
(10) utilize described minimum translation size as threshold value, and compare with the catercorner length in above-mentioned cavity, judge the type in cavity, thus different hole-filling methods is adopted respectively to different image cavity types.
In described step (6), setting swarming threshold value, using being greater than point peak position of swarming threshold value as a depth layer.Find out minimum-depth interlayer every rear set depth threshold value, compare minimum-depth interlayer every the size with depth threshold: if minimum-depth interlayer is every being greater than depth threshold, then get minimum-depth interlayer every as parameter; If minimum-depth interlayer is every being less than depth threshold, then get the minimum-depth interlayer that is greater than depth threshold every as parameter.The reason of employing said method is, avoid obtain true picture because entirety in there is the very little simultaneous situation very large with portion depth jump of portion change in depth, cause the minimum-depth saltus step calculated be change in depth very zonule calculate and ought to be uncared-for value.By set depth threshold value, can avoid little cavity being differentiated into macroscopic-void in this case by mistake.
In described step (8), the computing formula of minimum translation size is as follows:
L is minimum translation size; k
2it is the internal reference matrix of the camera of virtual view; [R, T] is the outer ginseng matrix of the camera of true viewpoint; D is the change in depth of this pixel in real camera viewpoint; X is in order to represent the coordinate of pixel at the image coordinate system of true viewpoint; k
1it is the internal reference matrix of the camera of true viewpoint.
The device that the present invention realizes said method is as follows:
This device comprises: degree of depth statistics with histogram module, for generating the degree of depth histogram of original view, adds up the number of pixels corresponding to different gray-scale value; Envelope diagram computing module, carries out envelope detected and neighbor smoothing filtering by the degree of depth histogram obtained, and generation is level and smooth, swarming significantly adds up envelope diagram; Depth layer computing module, detects the swarming of above-mentioned envelope diagram, and casts out by the threshold value preset the swarming that some have jamming pattern, is depth layer, and depth layer is done difference between two by remaining peak-fit processing, obtain minimum-depth interlayer every; Cavity size computing module, according to the position relationship of original image and virtual view, and above-mentioned minimum-depth interlayer every and Epipolar geometry method, calculate the catercorner length in minimum translation size and cavity; Cavity size judge module, utilizes above-mentioned minimum translation size as threshold value, according to the catercorner length in cavity, judges and distinguish in virtual view to empty type.
Method of the present invention, by utilizing the deep statistical histogram picture of depth image in image and sequence of frames of video and its correspondence, carries out type judgement respectively to the size cavity produced in render process, thus takes different to fill up strategy to size cavity.Invention not only simplifies the complicacy of existing Rendering algorithms, accelerate the speed played up, and greatly strengthen the accuracy of empty type judgement, what enhance cavity fills up effect, can obtain good image rendering effect.
Accompanying drawing explanation
In order to be illustrated more clearly in the embodiment of the present invention or technical scheme of the prior art, introduce to the accompanying drawing used required in embodiment or description of the prior art simply below, apparently, accompanying drawing in the following describes is only some embodiments of the present invention, for those of ordinary skill in the art, do not paying under performing creative labour prerequisite, other accompanying drawing can also obtained according to these accompanying drawings.
Fig. 1 is the schematic diagram producing cavity in prior art in rendering image process;
Fig. 2 is the grey level histogram in the embodiment of the present invention 1;
Fig. 3 is the process flow diagram of the empty type judgement method that the embodiment of the present invention 1 provides;
Fig. 4 is the structural representation of the empty kind judging device that the embodiment of the present invention 2 provides.
Embodiment
For making object of the present invention, technical scheme and advantage clearly, are described in further detail the invention process method below in conjunction with accompanying drawing.
Embodiment 1
Present embodiments provide a kind of method playing up the judgement of the empty type occurred in virtual view: the depth map choosing reference picture and frame of video; Histogram statistics calculating is carried out to described depth image, obtains the histogram corresponding with described view; According to described histogram, envelope detected is carried out to it, obtains envelope diagram; Neighbor smoothing filtering is carried out to described envelope diagram, obtains the obvious envelope diagram of level and smooth swarming; Setting swarming threshold value, using being greater than the swarming of threshold value as a depth layer; Between two difference is done to described all depth layer, calculate minimum-depth interlayer every; Set depth threshold value, compares minimum-depth interlayer every the size with depth threshold; If minimum-depth interlayer is every being greater than depth threshold, then get minimum-depth interlayer every as parameter; If minimum-depth interlayer is every being greater than depth threshold, then get the minimum-depth interlayer that is greater than depth threshold every as parameter; Obtain virtual view, according to position relationship and the above-mentioned minimum-depth layer spacing parameter of virtual view view and original view, calculate minimum-depth interlayer and be interposed between the value of the position translation that virtual view causes as translation threshold value; According to above-mentioned threshold value, to the cavity being greater than above-mentioned threshold size, think to block the cavity caused, and for being less than the cavity of above-mentioned threshold size, think the cavity that quantization error causes.
The method of above-mentioned acquisition virtual view can be: first demarcated by the inside and outside parameter matrix of the camera to true visual angle, the inside and outside parameter of the virtual camera of virtual view is calculated again according to the position relationship at virtual camera and true visual angle, use the image at true visual angle and the depth image of correspondence thereof, carry out the projection of individual element point thus render the view of virtual view.
As shown in Figure 3, detailed process is as follows:
S1: carry out Histogram statistics to the data that the depth image of original image stores in the matrix form in computer, adds up the pixel number corresponding to all gray-scale values (being in general 0-255 totally 256 levels);
S2: carry out envelope detected (utilizing existing envelope detected technology) according to above-mentioned histogram, obtain deep statistical envelope diagram;
S3: neighbor smoothing filtering (utilizing existing smothing filtering technology) is done to above-mentioned deep statistical envelope diagram, obtains level and smooth deep statistical envelope diagram; (design sketch is see Fig. 2)
S4: according to level and smooth deep statistical envelope diagram, detects all points of peak values, and does difference between two to used point of peak value, finds out minimum positive number difference, utilizes this difference, calculate minimum-depth interlayer every, in conjunction with formula:
thus calculate minimum-depth on virtual view and change the translational movement caused; Wherein L is minimum translation size; k
2be the internal reference matrix of the camera of virtual view, it comprises the internal information such as focal length and principal point that virtual view camera has, characterize the position relationship of image coordinate system and true coordinate system; [R, T] is the outer ginseng matrix of the camera of true viewpoint; D is the change in depth of this pixel in real camera viewpoint.X is in order to represent the coordinate of pixel at the image coordinate system of true viewpoint; k
1be the internal reference matrix of the camera of true viewpoint, it comprises the internal information such as focal length and principal point that true view camera should have, characterize the position relationship of image coordinate system and true coordinate system.
S5: using above-mentioned translational movement as the threshold value judging pore size, detects the cavity point on the virtual view generated, and distinguishes cavity that change in depth and blocking causes and the cavity that quantization error causes.Different empty filling Strategy can be taked to different size cavity in ensuing work, depth jump is blocked to the macroscopic-void caused, can consider to use spatial information (si) to fill up, such as, before and after using, the image of associated frame is filled up to obtain structural information and texture information, for the little cavity that quantization error is brought, can consider to use the classic algorithm of inpainting or Laplce's enthesis (roifill) or the image processing field such as horizontal extrapolation based on depth map.Always obtain good virtual view effect.
The function cvFindContours that the size in above-mentioned cavity can use opencv to carry finds out outline line, re-uses a cvRectangle rectangles encompass and changes cavity, and use the catercorner length of this rectangle, compare with the threshold value set before.The cavity caused is blocked by utilizing depth image and Epipolar geometry to calculate, to different size in render process, the process of different modes is carried out in the cavity that different reason produces, not only simplify the complicacy of existing Rendering algorithms, accelerate the speed played up, but also greatly strengthen the effect of hole-filling.
Example is as follows, such as,
Original image is designated as A, and the depth image of original image is designated as D.Degree of depth Histogram statistics is carried out to D.And be X-axis with gray-scale value, number of pixels is Y-axis.Obtain a statistic histogram C1;
Draw Histogram statistics envelope diagram according to above-mentioned histogram, obtain original Histogram statistics envelope diagram C2;
Neighbor smoothing filtering is carried out to C2, obtains the envelope diagram C3 of the obvious squelch of level and smooth swarming; See Fig. 2.
Set a swarming threshold value P, detect each swarming being greater than threshold value P in above-mentioned envelope diagram C3, the value of taking out X-axis is designated as h1, h2 ... hn;
Between two difference is done to above-mentioned point of peak value, B=argmin [hx-hy (x, y=1,2 ... n)]; In order to the robustness of system, preset a threshold value Q, make B be greater than Q;
According to the inside and outside parameter matrix of the position relationship combining camera of original viewpoint and virtual view, according to formula:
carry out computing, the displacement L that the minimum-depth saltus step B asked produces on virtual view;
According to above-mentioned displacement L, determine empty size parameter, thus distinguish, what depth jump caused block cavity and the cavity that quantization error causes, take different fill methods after being convenient to.
Embodiment 2
See Fig. 4, embodiments providing the device that a kind of empty type judges, for judging figure cavity type, comprising:
Degree of depth statistics with histogram module 301, for the degree of depth histogram of synthetic image and frame of video, in order to add up the number of pixels that different depth has, and represent degree of depth gray-scale value with X-axis, Y-axis represents number of pixels.
Wherein degree of depth histogram has multiple implementation algorithm, as matlab and opencv has corresponding hist function, its specific implementation thinking, that the size of all pixel values of whole picture is classified, then by the number of pixels mapping that each value has, what have more number of pixels has higher height in Histogram statistics.Be not specifically limited in the present embodiment.
Degree of depth envelope diagram generation module 302, this module obtains the data of degree of depth statistics with histogram CMOS macro cell, carries out envelope detected, then carry out neighbor smoothing filtering for the degree of depth histogram obtained according to described degree of depth statistics with histogram module, level and smooth in order to generate, swarming significantly adds up envelope diagram.
Wherein the method for envelope detected and filtering is multiple, does not do concrete restriction in the present embodiment.Generate design sketch and see Fig. 2
Depth layer computing module 303, this module obtains the data of degree of depth envelope statistical graph module, for detecting the swarming of above-mentioned envelope diagram, and cast out by the threshold value preset the swarming that some have jamming pattern, be depth layer by remaining peak-fit processing, and do difference between two, obtain minimum-depth saltus step.
Cavity size computing module 304, this module obtains the data of depth layer computing module, according to the position relationship of original view and virtual view, and above-mentioned minimum-depth saltus step and Epipolar geometry method, calculate the locus translation threshold value that minimum-depth saltus step brings.
Cavity type judging module 305, this module obtains the data of empty size computing module, utilizes above-mentioned locus translation threshold calculations to go out minimum empty size, judges and distinguish in virtual view to empty type.
By utilizing depth jump to judge the type in cavity, in render process, the size hole area produced is divided into and then and blocks cavity and quantization error cavity, again it is processed respectively, not only simplify the complicacy of existing Rendering algorithms, more accelerate rendering speed, and greatly strengthen cavity fill up effect.
Whole parts of the technique scheme that the embodiment of the present invention provides can have been come by the hardware that programmed instruction is relevant, described program can be stored in the access media that can read, this storage medium comprises: ROM, RAM, magnetic disc or CD medium various can be program code stored medium.
The foregoing is only preferred embodiment of the present invention, not in order to limit the present invention, within the spirit and principles in the present invention all, any amendment made, equivalent replacement, improves, all should be included among protection scope of the present invention.
Claims (6)
1., based on a shadowing method for depth layer, it is characterized in that, comprise the steps:
(1) depth map of reference-view is chosen;
(2) number of pixels corresponding to all gray-scale values of described depth map is added up, draw degree of depth histogram;
(3) envelope detected is carried out to described degree of depth histogram, draw out initial envelope diagram;
(4) neighbor smoothing filtering is carried out to the initial envelope diagram of described degree of depth histogram, then draw filtered envelope diagram;
(5) peakvalue's checking is carried out to described filtered envelope diagram, find out a point peak position;
(6) to described filtered envelope diagram, according to described point of peak position, mark each depth layer, between two difference done to described depth layer, find out minimum-depth interlayer every;
(7) obtain virtual view, determine the position of virtual view to be rendered;
(8) distance of described virtual view relative to described reference-view is calculated, according to the distance calculated, utilize described minimum-depth interlayer every calculating minimum translation size, namely minimum-depth interlayer is interposed between the value of the position translation that virtual view causes;
(9) carry out cavity to the virtual view obtained to detect, find out all cavities, calculate the catercorner length in cavity;
(10) utilize described minimum translation size as threshold value, and compare with the catercorner length in above-mentioned cavity, judge the type in cavity, thus different hole-filling methods is adopted respectively to different image cavity types.
2. a kind of shadowing method based on depth layer according to claim 1, is characterized in that, in described step (6), setting swarming threshold value, using being greater than point peak position of swarming threshold value as a depth layer.
3. a kind of shadowing method based on depth layer according to claim 1 and 2, is characterized in that, in described step (6), finds out minimum-depth interlayer every rear set depth threshold value, compares minimum-depth interlayer every the size with depth threshold:
If minimum-depth interlayer is every being greater than depth threshold, then get minimum-depth interlayer every as parameter;
If minimum-depth interlayer is every being less than depth threshold, then get the minimum-depth interlayer that is greater than depth threshold every as parameter.
4. a kind of shadowing method based on depth layer according to claim 1, is characterized in that, in described step (8), the computing formula of minimum translation size is as follows:
In formula, L is minimum translation size; k
2it is the internal reference matrix of the camera of virtual view; [R, T] is the outer ginseng matrix of the camera of true viewpoint; D is the change in depth of this pixel in real camera viewpoint; X is in order to represent the coordinate of pixel at the image coordinate system of true viewpoint; k
1it is the internal reference matrix of the camera of true viewpoint.
5. a kind of shadowing method based on depth layer according to claim 1, it is characterized in that, the type in cavity is judged: the catercorner length for cavity is greater than the cavity of threshold value in described step (10), block the cavity caused, and the catercorner length in cavity is less than to the cavity of threshold value, be the cavity that quantization error causes.
6. realize as claimed in claim 1 based on a device for the shadowing method of depth layer, it is characterized in that, this device comprises:
Degree of depth statistics with histogram module, for generating the degree of depth histogram of original view, adds up the number of pixels corresponding to different gray-scale value;
Envelope diagram computing module, carries out envelope detected and neighbor smoothing filtering by the degree of depth histogram obtained, and generates level and smooth, the obvious envelope diagram of swarming;
Depth layer computing module, detects the swarming of above-mentioned envelope diagram, and casts out by the threshold value preset the swarming that some have jamming pattern, is depth layer, and depth layer is done difference between two by remaining peak-fit processing, obtain minimum-depth interlayer every;
Cavity size computing module, according to the position relationship of original view and virtual view, and above-mentioned minimum-depth interlayer every and Epipolar geometry method, calculate the catercorner length in minimum translation size and cavity;
Cavity size judge module, utilizes above-mentioned minimum translation size as threshold value, according to the catercorner length in cavity, judges and distinguish in virtual view to empty type.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201310294784.9A CN103337081B (en) | 2013-07-12 | 2013-07-12 | A kind of shadowing method based on depth layer and device |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201310294784.9A CN103337081B (en) | 2013-07-12 | 2013-07-12 | A kind of shadowing method based on depth layer and device |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN103337081A CN103337081A (en) | 2013-10-02 |
| CN103337081B true CN103337081B (en) | 2016-03-16 |
Family
ID=49245229
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201310294784.9A Active CN103337081B (en) | 2013-07-12 | 2013-07-12 | A kind of shadowing method based on depth layer and device |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN103337081B (en) |
Families Citing this family (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103581648B (en) * | 2013-10-18 | 2015-08-26 | 清华大学深圳研究生院 | Draw the hole-filling method in new viewpoint |
| CN103731657B (en) * | 2014-01-26 | 2016-03-16 | 冠捷显示科技（厦门）有限公司 | A kind of to the filling of the cavity containing the empty image processing method after DIBR algorithm process |
| CN104657961B (en) * | 2014-12-17 | 2017-07-04 | 长安大学 | One kind is based on the histogrammic faulting of slab ends three-dimensional filtering method and system of bimodal road surface depth |
| CN104657993B (en) * | 2015-02-12 | 2018-04-17 | 北京格灵深瞳信息技术有限公司 | A kind of camera lens occlusion detection method and device |
| CN106060512B (en) * | 2016-06-28 | 2017-08-01 | 华中科技大学 | It is a kind of to choose and fill up rationally mapping point methods in virtual view synthesis |
| CN109697444B (en) * | 2017-10-20 | 2021-04-13 | ***通信有限公司研究院 | Object identification method and device based on depth image, equipment and storage medium |
| CN109635723B (en) * | 2018-12-11 | 2021-02-09 | 讯飞智元信息科技有限公司 | Shielding detection method and device |
| CN112991193B (en) * | 2020-11-16 | 2022-09-23 | 武汉科技大学 | Depth image restoration method, device and computer-readable storage medium |
Family Cites Families (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102325259A (en) * | 2011-09-09 | 2012-01-18 | 青岛海信数字多媒体技术国家重点实验室有限公司 | Method and device for synthesizing virtual viewpoints in multi-viewpoint video |
| CN103024421B (en) * | 2013-01-18 | 2015-03-04 | 山东大学 | Method for synthesizing virtual viewpoints in free viewpoint television |
-
2013
- 2013-07-12 CN CN201310294784.9A patent/CN103337081B/en active Active
Also Published As
| Publication number | Publication date |
|---|---|
| CN103337081A (en) | 2013-10-02 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN103337081B (en) | A kind of shadowing method based on depth layer and device | |
| TWI489418B (en) | Parallax Estimation Depth Generation | |
| CN104835164B (en) | A kind of processing method and processing device of binocular camera depth image | |
| AU2016355215B2 (en) | Methods and systems for large-scale determination of RGBD camera poses | |
| KR100793076B1 (en) | Edge-adaptive stereo/multi-view image matching apparatus and its method | |
| US9747694B2 (en) | Method and apparatus for optimizing depth information | |
| CN101873509B (en) | Method for eliminating background and edge shake of depth map sequence | |
| CN101610423B (en) | Method and device for rendering image | |
| CN109636732A (en) | A kind of empty restorative procedure and image processing apparatus of depth image | |
| US20200349757A1 (en) | Object capture coverage evaluation | |
| CN111047709B (en) | Binocular vision naked eye 3D image generation method | |
| TWI549477B (en) | Method and apparatus for generating depth information | |
| CN110544300B (en) | Method for automatically generating three-dimensional model based on two-dimensional hand-drawn image characteristics | |
| CN103384343B (en) | A kind of method and device thereof filling up image cavity | |
| EP2887314A1 (en) | Video frame conversion method and apparatus | |
| CN102436671A (en) | Virtual viewpoint drawing method based on depth value non-linear transformation | |
| CN104270624B (en) | A kind of subregional 3D video mapping method | |
| CN105007475B (en) | Produce the method and apparatus of depth information | |
| CN104778673B (en) | A kind of improved gauss hybrid models depth image enhancement method | |
| CN107155100A (en) | A kind of solid matching method and device based on image | |
| CN110149508A (en) | A kind of array of figure generation and complementing method based on one-dimensional integrated imaging system | |
| CN113450274B (en) | Self-adaptive viewpoint fusion method and system based on deep learning | |
| CN110602476A (en) | Hole filling method of Gaussian mixture model based on depth information assistance | |
| EP4149110A1 (en) | Virtual viewpoint synthesis method, electronic device and computer readable medium | |
| CN105025285B (en) | Optimize the method and apparatus of depth information |
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 |