CN107295264B - One kind being based on homography conversion light-field data compression method - Google Patents
One kind being based on homography conversion light-field data compression method Download PDFInfo
- Publication number
- CN107295264B CN107295264B CN201710648322.0A CN201710648322A CN107295264B CN 107295264 B CN107295264 B CN 107295264B CN 201710648322 A CN201710648322 A CN 201710648322A CN 107295264 B CN107295264 B CN 107295264B
- Authority
- CN
- China
- Prior art keywords
- sub
- image
- aperture image
- field data
- aperture
- 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
Classifications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N23/00—Cameras or camera modules comprising electronic image sensors; Control thereof
- H04N23/80—Camera processing pipelines; Components thereof
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N19/00—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals
- H04N19/42—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals characterised by implementation details or hardware specially adapted for video compression or decompression, e.g. dedicated software implementation
Abstract
The light-field data compression method based on homography conversion that the invention discloses a kind of, comprising: input original light field data collection Φ (S1,S2,…,Si), by sub-aperture image S thereiniIt is divided into center sub-aperture image CiWith corresponding neighborhood sub-aperture image Ai‑j；According to center sub-aperture image CiWith corresponding neighborhood sub-aperture image Ai‑jSeek homography matrix Hi‑j, homography conversion is based on by neighborhood sub-aperture image Ai‑jProject to center sub-aperture image CiPlace imaging plane obtains transformed neighborhood sub-aperture image A 'i‑j, by transformed neighborhood sub-aperture image A 'i‑jWith center sub-aperture image CiSubtract each other to obtain residual plot Ri‑j；Whole center sub-aperture image C is concentrated to original light field dataiWith residual image Ri‑jIt is scanned respectively, generates centered video sequence VCWith residual video sequence VR；To center video sequence VCWith residual video sequence VRIt is compressed respectively.The redundancy in airspace is effectively reduced in the present invention, and then realizes effective promotion of code efficiency.
Description
Technical field
The present invention relates to computer vision and digital image processing field more particularly to a kind of light based on homography conversion
Field data compression method.
Background technique
The full light camera based on microlens structure has caused the extensive concern of academia and industry recently, capture
All-optical data can record the location information and angle information of light simultaneously, in three-dimensional reconstruction, stereoscopic display imaging and virtual
The fields such as reality all illustrate huge application prospect.However, full light image has relative to traditional camera captured image
Ultrahigh resolution, while there is also huge information redundancies for its special lenticule pixel distribution form.The surge of data volume is given
The transimission and storage of light field data brings immense pressure, therefore is badly in need of a kind of effective light-field data compression scheme, this be also by
Plane multimedia march toward the stereopsis epoch be badly in need of and necessary core technology.
For the original light field image of light-field camera capture, can be pressed by rendering the sub-aperture image of generation to it
Contracting, the compression scheme can be roughly divided into three classes: the first kind is generated by sub-aperture image pseudo- by utilizing different scanning forms
Video sequence, and then compression is completed based on video encoder, such scheme is analyzed due to that cannot provide theoretic compression efficiency,
Therefore have certain uncertain；Second class scheme is by sub-aperture image according to different visual angle relationships, from airspace to time domain
Conversion, and then the compression of multi-vision-point encoding completion light field data is utilized, since multi-vision-point encoding needs to configure camera in the program
Parameter, while the viewpoint number of the visual angle number of sub-aperture image and multi-vision-point encoding device limits, so that implementing has certain hardly possible
Degree；Last one kind scheme is then the parallax relationship having using adjacent sub-aperture image, and parallax is introduced in video encoder and is mended
The encoding tool repaid reduces spatial redundancy information, such scheme fails to efficiently use the similar pass of the macro pixel in light field image
System, limits the further promotion of code efficiency.
The disclosure of background above technology contents is only used for auxiliary and understands design and technical solution of the invention, not necessarily
The prior art for belonging to present patent application, no tangible proof show above content present patent application the applying date
In disclosed situation, above-mentioned background technique should not be taken to the novelty and creativeness of evaluation the application.
Summary of the invention
The light-field data compression method based on homography conversion that the main purpose of the present invention is to provide a kind of, is effectively reduced
Redundancy in airspace, and then realize effective promotion of code efficiency.
To achieve the goals above, the invention adopts the following technical scheme:
The light-field data compression method based on homography conversion that the invention discloses a kind of, comprising the following steps:
A1: original light field data collection Φ (S is inputted1,S2,…,Si), by sub-aperture image S thereiniIt is divided into center sub-aperture
Diameter image CiWith corresponding neighborhood sub-aperture image Ai-j；
A2: according to center sub-aperture image CiWith corresponding neighborhood sub-aperture image Ai-jSeek homography matrix Hi-j, base
In homography conversion by neighborhood sub-aperture image Ai-jProject to center sub-aperture image CiPlace imaging plane obtains transformed
Neighborhood sub-aperture image A 'i-j, by transformed neighborhood sub-aperture image A 'i-jWith center sub-aperture image CiSubtract each other to obtain residual error
Scheme Ri-j；
A3: whole center sub-aperture image C is concentrated to light field dataiWith residual image Ri-jIt is scanned, generates respectively
Centered video sequence VCWith residual video sequence VR；
A4: to center video sequence VCWith residual video sequence VRIt is compressed respectively.
Preferably, step A1 specifically: to sub-aperture image S thereiniIt is divided into center according to airspace arrangement positional relationship
Sub-aperture image CiWith corresponding neighborhood sub-aperture image Ai-j。
Preferably, step A1 specifically: the original light field data collection Φ (S of input1,S2,…,Si) it is N × N square figure
As N × N square-shaped image array is divided into (N/3) × (N/ with adjacent 3 × 39 sub-aperture images for one group by array
3) sub-aperture image group is organized, wherein the sub-aperture image S at the center in every group of sub-aperture image groupiDefinition is center sub-aperture figure
As Ci, 8 adjacent sub-aperture image S of surroundingiIt is defined as corresponding neighborhood sub-aperture image Ai-j。
Preferably, step A2 specifically:
A21: it is based on SIFT image matching algorithm and DLT direct linear transformation's algorithm, solves center sub-aperture figure respectively
As CiWith corresponding neighborhood sub-aperture image Ai-jHomography matrix Hi-j:
A22: by neighborhood sub-aperture image Ai-jProject to center sub-aperture image CiPlace imaging plane obtains transformed
Neighborhood sub-aperture image A 'i-j:
[x ', y ', 1]=[x, y, 1] Hi-j
Wherein, (x, y) is neighborhood sub-aperture image Ai-jIn pixel coordinate, (x ', y ') be transformed neighborhood sub-aperture
Image A 'i-jIn pixel coordinate；
A23: by transformed neighborhood sub-aperture image A 'i-jWith corresponding center sub-aperture image CiSubtract each other, obtains residual error
Scheme Ri-j:
IR(u, v)=IA(u,v)-Ic(u,v)
Wherein, (u, v) is the pixel coordinate in image, IR、IA、ICRespectively Ri-j、A′i-jAnd CiIn pixel value.
Preferably, whole center sub-aperture image C is concentrated in step A3 to light field dataiThe method being scanned is Z
Font scanning or S-shaped scanning, wherein zigzag scanning indicates the center sub-aperture image C between adjacent rowsiIt is from beginning to end, S-shaped
Scanning indicates the center sub-aperture image C between adjacent rowsiHead is connected or tail portion is connected.
Preferably, whole residual image R is concentrated in step A3 to light field datai-jThe method being scanned is swept for zigzag
It retouches or S-shaped scans, wherein zigzag scanning indicates the residual image R between adjacent rowsi-jFrom beginning to end, S-shaped scanning indicates phase
The residual image R of neighbour two in the ranksi-jHead is connected or tail portion is connected.
Preferably, whole residual image R is concentrated in step A3 to light field datai-jThe method being scanned is to turn-take to sweep
It retouches, scanning of turn-taking is indicated to 8 residual image R in sub- subaperture image groupi-jIt turn-takes after scanning, with other group of adjacent sub-aperture
Residual image R in diameter image groupi-jIt joins end to end.
Preferably, step A4 further includes to whole homography matrix Hi-jCoding compression.
Preferably, step A4 further includes to whole homography matrix Hi-jCarry out block code compression.
Preferably, step A4 specifically: to center video sequence VCWith residual video sequence VRIt is utilized respectively video encoder
Carry out compressed encoding.
Compared with prior art, the beneficial effects of the present invention are the light field datas of the invention based on homography conversion
Original light field data collection is divided first as center sub-aperture image and corresponding neighborhood sub-aperture image, is based on by compression method
Homography conversion completes the mapping of neighborhood sub-aperture image and center sub-aperture image, so by residual image to acquisition with
Center sub-aperture image is scanned the pseudo- video sequence of generation respectively, realizes that light field data is compiled to video sequence and then after compressing
Effective promotion of code efficiency；The light-field data compression method, takes full advantage of the design feature of light field data, reduces light as best one can
The redundancy of field data, to effectively improve the coding compression efficiency of light field image.
In further embodiment, sub-aperture image is divided according to the airspace of sub-aperture image arrangement positional relationship, so
The visual angle effect of neighborhood sub-aperture image and center sub-aperture image is completed by homography conversion afterwards, and then scans building center
Multi-view video sequence and residual video sequence, to make full use of the geometry of imaged viewing angle between light field data sub-aperture image to close
System and have the characteristics that similar light intensity is arranged each other, due to the similitude between image, residual image relative smooth, so that
Sufficiently reduce code rate expense in cataloged procedure, the redundancy of light field data is further effectively reduced, to further improve
The coding compression efficiency of light field image.
Detailed description of the invention
Fig. 1 is the flow diagram of the light-field data compression method based on homography conversion of the preferred embodiment of the present invention；
Fig. 2 a is the schematic diagram of the light field data collection of the preferred embodiment of the present invention；
Fig. 2 b is the enlarged diagram in Fig. 2 a at T；
Fig. 3 a is that the zigzag for center sub-aperture image in the preferred embodiment of the present invention scans schematic diagram；
Fig. 3 b is that the S-shaped for center sub-aperture image in the preferred embodiment of the present invention scans schematic diagram；
Fig. 4 a is that the zigzag for residual image in the preferred embodiment of the present invention scans schematic diagram；
Fig. 4 b is that the S-shaped for residual image in the preferred embodiment of the present invention scans schematic diagram；
Fig. 4 c is to scan schematic diagram for turn-taking for residual image in the preferred embodiment of the present invention.
Specific embodiment
Below against attached drawing and in conjunction with preferred embodiment, the invention will be further described.
As shown in Figure 1, the preferred embodiment of the present invention discloses a kind of light-field data compression method based on homography conversion,
The following steps are included:
A1: original light field data collection Φ (S is inputted1,S2,…,Si), to sub-aperture image S thereiniIt arranges according to airspace
Positional relationship is divided into center sub-aperture image CiWith corresponding neighborhood sub-aperture image Ai-j；
Wherein, the original light field data collection Φ (S of input1,S2,…,Si) by sub-aperture image SiComposition.Sub-aperture image Si
A series of original full light image generated after pre-processings as obtained by light-field camera (such as LytroIllum) acquisition is through wash with watercolours
It contaminates obtained by algorithm, and is N × N square-shaped image array according to the arrangement of its imaged viewing angle relationship；It is generated with LytroIllum
For light field data, Φ (S1,S2,…,Si) it is 15 × 15 totally 225 sub-aperture images, as shown in Figure 2 a.For the image battle array
Former N × N square-shaped image array is divided into (N/3) × (N/3) with adjacent 3 × 39 sub-aperture images for one group by column
Group, as shown in Figure 2 b, the sub-aperture image S at every group of centeriDefinition is center sub-aperture image Ci, 8 adjacent sub-aperture figures of surrounding
As being defined as neighborhood sub-aperture image Ai-j, the invalid sub-aperture images of light field some for quadrangle around are ignored and are skipped (such as
Shown in light gray color lump in Fig. 2 a at four angles).
A2: according to center sub-aperture image CiWith corresponding neighborhood sub-aperture image Ai-jSeek homography matrix Hi-j, base
In homography conversion by neighborhood sub-aperture image Ai-jProject to center sub-aperture image CiPlace imaging plane obtains transformed
Neighborhood sub-aperture image A 'i-j, by transformed neighborhood sub-aperture image A 'i-jWith center sub-aperture image CiSubtract each other to obtain residual error
Scheme Ri-j；
For each ready-portioned 3 × 3 sub-aperture image group, directly linearly become based on SIFT image matching algorithm and DLT
Scaling method solves 8 neighborhood sub-aperture image A around respectivelyi-jWith center sub-aperture image CiHomography matrix Hi-j；Into
And utilize homography conversion by neighborhood sub-aperture image Ai-jProject to center sub-aperture image CiPlace imaging plane is converted
Neighborhood sub-aperture image A ' afterwardsi-j:
[x ', y ', 1]=[x, y, 1] Hi-j
Wherein, (x, y) is former neighborhood sub-aperture image Ai-jIn pixel coordinate, (x ', y ') be transformed neighborhood sub-aperture
Diameter image A 'i-jIn pixel coordinate.
Resulting neighborhood sub-aperture image A ' will be convertedi-jWith corresponding center sub-aperture image CiSubtract each other, obtains residual plot
Ri-j:
IR(u, v)=IA(u,v)-Ic(u,v)
Wherein, (u, v) is the pixel coordinate in image, IR、IA、ICRespectively Ri-j、A′i-jAnd CiIn pixel value.
A3: whole center sub-aperture image C is concentrated to original light field dataiWith residual image Ri-jAccording to respective sky
Domain scanning mode generates centered video sequence V respectivelyCWith residual video sequence VR；
Wherein, for center sub-aperture image Ci, (S- can be scanned using zigzag scanning (Zigzag scanning) or S-shaped
Shape scanning), wherein zigzag scans center sub-aperture image C of the form between adjacent rowsiIt is from beginning to end, such as Fig. 3 a institute
Show；And center sub-aperture image C of the S-shaped scanning form between adjacent rowsiHead is connected or tail portion is connected, such as Fig. 3 b institute
Show.
For residual image Ri-j, (Zigzag scanning), S-shaped scanning (S-Shape scanning) can be scanned with zigzag or are turned
Circle scanning (Circle scanning), wherein zigzag scanning and S-shaped scan form and center bore image CiSimilar namely zigzag is swept
Retouch residual image R of the form between adjacent rowsi-jIt is from beginning to end, as shown in fig. 4 a；And S-shaped scans form between adjacent rows
Residual image Ri-jHead is connected or tail portion is connected, as shown in Figure 4 b；And scanning form of turn-taking is the sub-aperture figure to 3 × 3
As 8 residual image R in groupi-jIt turn-takes after scanning, joins end to end with the residual image in other adjacent groups, such as Fig. 4 c institute
Show.
A4: to center video sequence VCWith residual video sequence VRIt is utilized respectively video encoder to be compressed, while to complete
The homography matrix H in portioni-jCoding compression.
Wherein, to center video sequence VCWith residual video sequence VRIt is utilized respectively video encoder and carries out compressed encoding, together
When to whole homography matrix Hi-jCarry out block code compression.And the process of decoding and rebuilding is then the inverse process of the above process,
Details are not described herein.
The preferred embodiment of the present invention discloses a kind of light-field data compression method based on homography conversion, first according to
The positional relationship of sub-aperture image space arrangement is concentrated to carry out center sub-aperture image and field according to original light field data is inputted
The division of subaperture image carries out homography conversion using the similitude of sub-aperture image and neighborhood sub-aperture image, completes neighborhood
Mapping of the sub-aperture image to center sub-aperture image；It is obtained using transformed neighborhood sub-aperture image and center sub-aperture image
To residual image, respectively by scanning generation center sub-aperture path image sequence and neighborhood residual error sub-aperture path image sequence, due to figure
Similitude as between, residual image relative smooth, so that sufficiently reducing code rate expense in an encoding process；The wherein light field
Encoding scheme takes full advantage of in light field data visual angle relationship and adjacent sub-aperture in the imaging process of adjacent sub-aperture image
Similar light intensity arrangement relationship, smooth residual image avoid a large amount of code rate expense, can select different sweep between image
Retouch mode and provide multiple choices for the promotion of code efficiency, the light-field data compression method make light field image research field and
The marketization of light-field camera is benefited extensively.
The above content is a further detailed description of the present invention in conjunction with specific preferred embodiments, and it cannot be said that
Specific implementation of the invention is only limited to these instructions.For those skilled in the art to which the present invention belongs, it is not taking off
Under the premise of from present inventive concept, several equivalent substitute or obvious modifications can also be made, and performance or use is identical, all answered
When being considered as belonging to protection scope of the present invention.
Claims (9)
1. a kind of light-field data compression method based on homography conversion, which comprises the following steps:
A1: original light field data collection Φ (S is inputted1,S2,…,Si), by sub-aperture image S thereiniIt is divided into center sub-aperture image
CiWith corresponding neighborhood sub-aperture image Ai-j；
A2: according to center sub-aperture image CiWith corresponding neighborhood sub-aperture image Ai-jSeek homography matrix Hi-j, based on singly answering
Property transformation by neighborhood sub-aperture image Ai-jProject to center sub-aperture image CiPlace imaging plane obtains transformed neighborhood
Subaperture image A 'i-j, by transformed neighborhood sub-aperture image A 'i-jWith center sub-aperture image CiSubtract each other to obtain residual plot Ri-j；
A3: whole center sub-aperture image C is concentrated to light field dataiWith residual image Ri-jIt is scanned respectively, generates center
Video sequence VCWith residual video sequence VR；
A4: to center video sequence VCWith residual video sequence VRIt is compressed respectively；
Wherein, step A2 specifically:
A21: it is based on SIFT image matching algorithm and DLT direct linear transformation's algorithm, solves center sub-aperture image C respectivelyiWith
Corresponding neighborhood sub-aperture image Ai-jHomography matrix Hi-j:
A22: by neighborhood sub-aperture image Ai-jProject to center sub-aperture image CiPlace imaging plane obtains transformed neighborhood
Sub-aperture image A 'i-j:
[x ', y ', 1]=[x, y, 1] Hi-j
Wherein, (x, y) is neighborhood sub-aperture image Ai-jIn pixel coordinate, (x ', y ') be transformed neighborhood sub-aperture image
A′i-jIn pixel coordinate；
A23: by transformed neighborhood sub-aperture image A 'i-jWith corresponding center sub-aperture image CiSubtract each other, obtains residual plot
Ri-j:
IR(u, v)=IA(u,v)-Ic(u,v)
Wherein, (u, v) is the pixel coordinate in image, IR、IA、ICRespectively Ri-j、A′i-jAnd CiIn pixel value.
2. light-field data compression method according to claim 1, which is characterized in that step A1 specifically: to son therein
Subaperture image SiIt is divided into center sub-aperture image C according to airspace arrangement positional relationshipiWith corresponding neighborhood sub-aperture image Ai-j。
3. light-field data compression method according to claim 1, which is characterized in that step A1 specifically: input it is original
Light field data collection Φ (S1,S2,…,Si) it is N × N square-shaped image array, with adjacent 3 × 39 sub-aperture images for one
N × N square-shaped image array is divided into (N/3) × (N/3) group sub-aperture image group, wherein every group of sub-aperture image group by group
In center sub-aperture image SiDefinition is center sub-aperture image Ci, 8 adjacent sub-aperture image S of surroundingiIt is defined as phase
The neighborhood sub-aperture image A answeredi-j。
4. light-field data compression method according to claim 1, which is characterized in that concentrated in step A3 to light field data complete
The center sub-aperture image C in portioniThe method being scanned is that zigzag scans or S-shaped scanning, wherein zigzag scanning indicate phase
The center sub-aperture image C of neighbour two in the ranksiFrom beginning to end, S-shaped scanning indicates the center sub-aperture image C between adjacent rowsiHead
It is connected or tail portion is connected.
5. light-field data compression method according to claim 1, which is characterized in that concentrated in step A3 to light field data complete
The residual image R in portioni-jThe method being scanned is that zigzag scans or S-shaped scanning, wherein zigzag scanning indicate adjacent rows
Between residual image Ri-jFrom beginning to end, S-shaped scanning indicates the residual image R between adjacent rowsi-jHead is connected or tail portion phase
Even.
6. light-field data compression method according to claim 3, which is characterized in that concentrated in step A3 to light field data complete
The residual image R in portioni-jThe method being scanned is scanning of turn-taking, and scanning of turn-taking is indicated to 8 residual errors in sub- subaperture image group
Image Ri-jIt turn-takes after scanning, with the residual image R in other group of adjacent sub-aperture image groupi-jIt joins end to end.
7. light-field data compression method according to claim 1, which is characterized in that step A4 further includes answering whole lists
Property matrix Hi-jCoding compression.
8. light-field data compression method according to claim 1, which is characterized in that step A4 further includes answering whole lists
Property matrix Hi-jCarry out block code compression.
9. light-field data compression method according to claim 1, which is characterized in that step A4 specifically: to centered video
Sequence VCWith residual video sequence VRIt is utilized respectively video encoder and carries out compressed encoding.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710648322.0A CN107295264B (en) | 2017-08-01 | 2017-08-01 | One kind being based on homography conversion light-field data compression method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710648322.0A CN107295264B (en) | 2017-08-01 | 2017-08-01 | One kind being based on homography conversion light-field data compression method |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN107295264A CN107295264A (en) | 2017-10-24 |
| CN107295264B true CN107295264B (en) | 2019-09-06 |
Family
ID=60105057
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201710648322.0A Active CN107295264B (en) | 2017-08-01 | 2017-08-01 | One kind being based on homography conversion light-field data compression method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN107295264B (en) |
Families Citing this family (12)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107770537B (en) * | 2017-11-02 | 2020-03-31 | 中国科学技术大学 | Light field image compression method based on linear reconstruction |
| CN108184064B (en) * | 2018-01-04 | 2020-06-26 | 中国科学技术大学 | Visual angle graph array dividing method |
| CN108921781B (en) * | 2018-05-07 | 2020-10-02 | 清华大学深圳研究生院 | Depth-based optical field splicing method |
| CN109099857B (en) * | 2018-08-24 | 2020-03-17 | 中国工程物理研究院机械制造工艺研究所 | Subaperture splicing method based on SURF feature matching |
| CN109489559B (en) * | 2018-10-08 | 2020-07-03 | 北京理工大学 | Point light source space positioning method based on time-frequency analysis and light field imaging technology |
| CN109064410B (en) * | 2018-10-24 | 2023-03-14 | 清华大学深圳研究生院 | Super-pixel-based light field image splicing method |
| CN109271803B (en) * | 2018-11-08 | 2021-09-28 | 首都师范大学 | Method for encrypting, reading and writing information by using nanotechnology |
| CN111757125B (en) * | 2019-03-29 | 2024-02-27 | 曜科智能科技（上海）有限公司 | Multi-view video compression method based on light field, device, equipment and medium thereof |
| CN110191359A (en) * | 2019-05-16 | 2019-08-30 | 华侨大学 | A kind of light field image compression method chosen based on crucial sub-aperture image |
| CN110392266B (en) * | 2019-07-25 | 2021-07-16 | 清华大学深圳研究生院 | Light field video coding method based on pseudo video sequence, terminal equipment and storage medium |
| CN111147848B (en) * | 2019-12-30 | 2021-10-01 | 清华大学深圳国际研究生院 | Light field video coding method based on content self-adaptation |
| CN111182312B (en) * | 2020-01-03 | 2021-10-29 | 杭州电子科技大学 | Hierarchical residual error light field video coding method |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106375766A (en) * | 2016-09-08 | 2017-02-01 | 电子科技大学 | Light field image compression method |
| CN106791869A (en) * | 2016-12-21 | 2017-05-31 | 中国科学技术大学 | Quick motion search method based on light field sub-aperture image relative position relation |
| CN106973293A (en) * | 2017-04-21 | 2017-07-21 | 中国科学技术大学 | The light field image coding method predicted based on parallax |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US7991235B2 (en) * | 2007-02-27 | 2011-08-02 | Xerox Corporation | Light compression for color images using error diffusion |
-
2017
- 2017-08-01 CN CN201710648322.0A patent/CN107295264B/en active Active
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106375766A (en) * | 2016-09-08 | 2017-02-01 | 电子科技大学 | Light field image compression method |
| CN106791869A (en) * | 2016-12-21 | 2017-05-31 | 中国科学技术大学 | Quick motion search method based on light field sub-aperture image relative position relation |
| CN106973293A (en) * | 2017-04-21 | 2017-07-21 | 中国科学技术大学 | The light field image coding method predicted based on parallax |
Also Published As
| Publication number | Publication date |
|---|---|
| CN107295264A (en) | 2017-10-24 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN107295264B (en) | One kind being based on homography conversion light-field data compression method | |
| US10484664B2 (en) | Mapping of spherical image data into rectangular faces for transport and decoding across networks | |
| Cho et al. | Modeling the calibration pipeline of the lytro camera for high quality light-field image reconstruction | |
| US20190141323A1 (en) | Video image encoding method and apparatus, and video image decoding method and apparatus | |
| Zhu et al. | Efficient single image super-resolution via hybrid residual feature learning with compact back-projection network | |
| KR101723738B1 (en) | Apparatus and method for resolution enhancement based on dictionary learning | |
| WO2022016350A1 (en) | Light field image processing method, light field image encoder and decoder, and storage medium | |
| CN106254719B (en) | A kind of light field image compression method based on linear transformation and image interpolation | |
| CN103024421A (en) | Method for synthesizing virtual viewpoints in free viewpoint television | |
| CN114897680B (en) | Angle super-resolution method for fusing light field sub-aperture image and macro-pixel image | |
| CN106056656A (en) | Three-dimensional display data acquisition method | |
| CN112017228A (en) | Method for three-dimensional reconstruction of object and related equipment | |
| Sauer et al. | Improved motion compensation for 360° video projected to polytopes | |
| Majidi et al. | A deep model for super-resolution enhancement from a single image | |
| CN111369443B (en) | Zero-order learning super-resolution method of light field cross-scale | |
| Jin et al. | Image reshaping for efficient compression of plenoptic content | |
| CN112785502B (en) | Light field image super-resolution method of hybrid camera based on texture migration | |
| CN106961605A (en) | A kind of light field image compression method matched based on grand pixel boundary | |
| Shima et al. | Comparison of image alignment on hexagonal and square lattices | |
| Niu et al. | Lightweight and accurate single image super-resolution with channel segregation network | |
| CN110223357B (en) | Multi-pixel imaging method based on compressed sensing | |
| Zhong et al. | Implicit epipolar geometric function based light field continuous angular representation | |
| CN116309067B (en) | Light field image space super-resolution method | |
| WO2023174336A1 (en) | Encoding method and apparatus, decoding method and apparatus, and device | |
| US20230171390A1 (en) | Image conversion apparatus and method |
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 | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |