CN108111863A - A kind of online real-time three-dimensional model video coding-decoding method - Google Patents
A kind of online real-time three-dimensional model video coding-decoding method Download PDFInfo
- Publication number
- CN108111863A CN108111863A CN201711406406.XA CN201711406406A CN108111863A CN 108111863 A CN108111863 A CN 108111863A CN 201711406406 A CN201711406406 A CN 201711406406A CN 108111863 A CN108111863 A CN 108111863A
- Authority
- CN
- China
- Prior art keywords
- data
- dough sheet
- texture
- video
- index
- 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
Classifications
-
- 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
- H04N19/50—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using predictive coding
- H04N19/597—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using predictive coding specially adapted for multi-view video sequence encoding
-
- 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
- H04N19/44—Decoders specially adapted therefor, e.g. video decoders which are asymmetric with respect to the encoder
-
- 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
- H04N19/70—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals characterised by syntax aspects related to video coding, e.g. related to compression standards
Landscapes
- Engineering & Computer Science (AREA)
- Multimedia (AREA)
- Signal Processing (AREA)
- Compression Or Coding Systems Of Tv Signals (AREA)
Abstract
The present invention provides a kind of online real-time three-dimensional model video coding-decoding methods, it is embedded into customized NAL unit in video flowing after threedimensional model is compressed by Meshsequential algorithms, then the texture maps NAL unit after coding and customized model NAL unit are encapsulated into ts stream packets, and pass through network and be sent to client, client is decoded displaying.For the threedimensional model of certain number of vertex knead dough the piece number, one frame data of this method encoding and decoding can accomplish online real-time exhibition threedimensional model within 30ms, and be shown in the form of video code flow, and ordinary playing device can also play the texture part of the code stream；Compared with FVV, Holopoetation of Microsoft, the present invention can both realize online real-time live broadcast, can also realize coding and decoding video, be more suitable for the stream transmission of network, can reach 360 degree of viewing videos, substantially improve viewing experience.
Description
Technical field
The present invention relates to threedimensional model video coding and decoding technology field, a kind of specifically online real-time three-dimensional model regards
Frequency decoding method.
Background technology
In threedimensional model video field, texture maps encoding and decoding and threedimensional model encoding and decoding two parts are mainly included.At present
It is related to that article, technology or the patent of threedimensional model coding and decoding video are less, and existing article has Microsoft FVV, Holoportation
Deng.
Texture maps are encoded to video flowing by Microsoft FVV according to H.264 method for video coding；For threedimensional model, borrow first
Help Mesh tracking, by the way that threedimensional model is divided into key frame and prediction frame.Key frame includes geological information（Apex coordinate and line
Manage coordinate）And link information, geometry encoding using quantify 16bit methods, link information using delta coding, can mutilation
Save deltic method；And frame is predicted only comprising geological information increment, and utilize Linear motion prediction computational geometry increment, Ran Houjin
Row quantifies and Columbus's compressed encoding.Compressed model data is finally embedded into customized NAL unit, shape in video flowing
Into the mp4 files with grid model.One frame of this method compressed encoding needs the 2.2s times, does not accomplish that online real-time exhibition is three-dimensional
Model.
For texture maps, Holorportation uses LZ4 compression algorithm compressed textures figures；For threedimensional model,
Vertex deduplication operation is first carried out in Holorportation, to reduce vertex coordinates data amount, secondly by texture coordinate data amount
Change to 16bit, then with the compression of LZ4 methods go it is re-quantization after three-dimensional modeling data, finally by TCP by compressed texture
Data and three-dimensional modeling data are directly transferred to client.This method can accomplish the displaying for having delay, but not accomplish that video is compiled
Decoding is unfavorable for threedimensional model in transmission over networks with playing.
In addition, 201280033913.2 patents of application number CN disclose " for repetitive structure threedimensional model coding and
Decode the system and method for bit stream ", which is compressed mainly for 3D (three-dimensional) model with repetitive structure, encode and
Decoding, and the threedimensional model for not having repetitive structure can not be compressed, coded and decoded.Application number CN
201010155134.2 patents disclose " three-dimensional model transmitting and interactive drawing method in a kind of Wireless LAN ", this is specially
Profit is one kind three-dimensional model transmitting and interactive drawing method suitable for Wireless LAN, can be realized in mobile client
The interactive real-time drawing function of the arbitrary resolution of implementation model, mainly for wireless local threedimensional model off the net transmission with
It draws, texture can not be compressed and be transmitted.
The content of the invention
In order to solve the problems, such as the online real-time video encoding and decoding of threedimensional model, accomplish to show online in the form of video code flow
Threedimensional model, the present invention provides a kind of online real-time three-dimensional model video coding-decoding methods, are calculated by Meshsequential
Method is embedded into customized NAL in video flowing after threedimensional model is compressed（Network Abstraction Layer, network are taken out
As layer）Then the texture maps NAL unit formed after compressed encoding and self-definition model NAL unit are encapsulated into ts by unit
（Transport Stream, transport stream）In stream packet, and pass through network and be sent to client, client is decoded displaying.
The coding method can accomplish online real-time exhibition threedimensional model, and be shown in the form of video code flow, and ordinary playing device also may be used
To play the texture part of the code stream.
The technical solution that uses to solve above-mentioned technical problem of the present invention for：A kind of online real-time three-dimensional model video compiles solution
Code method, this method includes the server side code stage, network transmission stage and the client decoding stage that carry out successively, specific real
Existing step is as follows：
S1, texture graph code：Video compress is carried out to original texture diagram data using H.264 video encoding standard, is compressed
Texture map data afterwards, and it is encapsulated into video NAL units；
S2, threedimensional model coding：Original three-dimensional model data are compressed using Meshsequential algorithms, are compressed
Compressed three-dimensional modeling data is encapsulated into customized model NAL unit by three-dimensional modeling data afterwards afterwards；The original
Beginning three-dimensional modeling data includes vertex coordinates data and dough sheet index data, using Meshsequential algorithms to initial three-dimensional
Model data be compressed the specific implementation process is as follows：
S21, it is compressed for threedimensional model dough sheet index data：First index datastore of first dough sheet is absolute
Value, the index data of remaining dough sheet is stored as the relative value based on the index data of previous plane piece, until all faces
Piece index data finishes compression and preservation；
S22, it is compressed for threedimensional model vertex coordinates data：First all vertex coordinates datas are quantified, afterwards
Coding is compressed using entropy coding；
S23, before the first frame data are encoded, the camera parameter under each visual angle is transmitted through the network to decoding end；
S3, the step S1 video NAL units formed and the customized model NAL units of step S2 are encapsulated into ts stream packets, shape
Into ts code streams；
Step 2: the network transmission stage：The network transmission stage by the ts code streams after packing, is transmitted through the network to client
End；
Step 3: client decoding stage, specific decoding process is as follows：
T1, the step S3 ts stream packets formed are unpacked, obtains NAL unit；
T2, the NAL unit type obtained according to step T1, isolate texture NAL unit and threedimensional model NAL unit；
X264 Video Decoders and Meshsequential models is respectively adopted in T3, texture NAL unit and threedimensional model NAL unit
Decoder is operated in sub-line journey and unziped it, and respectively obtains the threedimensional model number after the data texturing and decompression after decompression
According to, and the three-dimensional modeling data after the data texturing and decompression after decompression is stored in buffering area；Using Meshsequential moulds
It is that type decoder unzips it the specific implementation process is as follows：
T31, according to coding side compression process, read first index data of first dough sheet, be afterwards converted into relative value
Absolute value realizes the decoding of dough sheet index data；
T32, carry out entropy coding inverse process and quantify inverse process, decompress vertex coordinates data；
T33, indexed for each group of dough sheet, by texture under camera parameter, each visual angle dough sheet starting point and each camera perspective
Figure starting point coordinate calculates the texture coordinate uv of vertex correspondence in every group of dough sheet index data, and texture projection is carried out according to uv；
When T4, area to be buffered are full, according to every mono- frames of 30ms, successively by the data texturing after all decompressions in buffering area according to three-dimensional
On texture coordinate map to threedimensional model in model, a textured threedimensional model video counts frame that can be shown is formed；
T5, by constructing three-dimensional rendering scene, shown successively per frame textured threedimensional model video frame, realize and play band in real time
The threedimensional model video of texture.
As a preferred embodiment, original texture figure described in step S1 is texture maps under each camera perspective according to visual angle
Order storage is regarded texture maps starting point coordinate under each camera perspective and each into the texture maps of a 2048x2048
Edged surface piece starting point is stored according to dough sheet indexed mode to dough sheet index.
As a preferred embodiment, dough sheet index data described in step S2 is included according to the tactic dough sheet of camera perspective
The starting point coordinate of texture maps and each camera perspective lower panel starting point under index, each camera perspective.
As a preferred embodiment, first index data of the absolute value for first dough sheet is stored as described in step S21
Preserve partial data, the index datastore of remaining dough sheet be each dough sheet after relative value is index value with it is previous
Plane piece index makes the difference, and the compressed data indexed using the difference as the dough sheet and preservation.
As a preferred embodiment, pair all vertex coordinates datas are quantified apex coordinate number described in step S22
According to 16bit is quantized to, the entropy coding uses rANS algorithms.
As a preferred embodiment, the video of the Video coding storehouse of step S1 and step T3 decoding storehouse are
libx264。
As a preferred embodiment, it is by the index information of n-th of dough sheet that relative value described in step T31, which is converted into absolute value,
It is added to obtain this dough sheet index value with the dough sheet index value of previous position, recovers the texture under dough sheet index value, each camera perspective
Figure starting point coordinate and each visual angle lower panel starting point.
Compared with prior art, the invention has the advantages that：
（1）The present invention provides a kind of online real-time three-dimensional model video coding-decoding methods, pass through Meshsequential algorithms
It is embedded into customized NAL unit in video flowing after threedimensional model is compressed, then by the texture maps NAL unit after coding and certainly
The model NAL unit of definition is encapsulated into ts stream packets, and is passed through network and be sent to client, and client is decoded exhibition
Show.The threedimensional model video coding-decoding method is in the few threedimensional model video of processing three-dimensional modeling data amount, one frame of encoding and decoding
Data can reach within 30ms, and in the excessive threedimensional model video of processing three-dimensional modeling data amount, area to be buffered expires, can be real
Existing real-time exhibition threedimensional model, and shown in the form of video code flow, ordinary playing device can also play the texture of the code stream
Part, and third party MPEG edit tools can change video stream data in the case where not losing three-dimensional modeling data；
（2）The present invention provides a kind of online real-time three-dimensional model video coding-decoding method, FVV with Microsoft,
Holopoetation is compared, and the present invention can both realize the online real-time live broadcast of the video that will have been buffered, and can also realize that video is compiled
Decoding, is more suitable for the stream transmission of network, can reach 360 degree of viewing videos, substantially improve viewing experience.And this hair
The bright threedimensional model with repetitive structure that can not only be directed to is compressed, and also may be used for the threedimensional model for not having repetitive structure
To be compressed, code and decode.
Description of the drawings
Fig. 1 is the flow chart of the present invention.
Specific embodiment
Elaborate with reference to specific embodiment to the present invention, the present embodiment based on the technical solution of the present invention,
Give detailed embodiment and specific operating process.
A kind of online real-time three-dimensional model video coding-decoding method, this method include the server side code rank carried out successively
Section, network transmission stage and client decoding stage, specific implementation step are as follows：
Step 1: the server side code stage, the server side code stage includes texture graph code and threedimensional model encodes, specifically
Realization process is as follows：
S1, texture graph code：Video compress is carried out to original texture diagram data using H.264 video encoding standard, is compressed
Texture map data afterwards, original texture diagram data described herein are that the texture maps under each camera perspective are stored according to visual angle order
It is risen into the texture maps of a 2048x2048, and by texture maps starting point coordinate under each camera perspective and each visual angle dough sheet
Initial point is stored according to dough sheet indexed mode to dough sheet index, and is encapsulated into video NAL units；H.264 it is the VCEG of ITU-T
（Video Coding Experts Group）With the MPEG of ISO/IEC（Motion Picture Expert Group）Joint video team（JVT：joint
video team）One digital video coding standard of exploitation, can also be video compression format；X264 is one and is based on
H.264 the coding and decoding video project freely increased income；Libx264 is the function library of x264 coding generations.
S2, threedimensional model coding：Original three-dimensional model data are compressed using Meshsequential algorithms, are obtained
Compressed three-dimensional modeling data, as preference, herein Meshsequential algorithms by the parallel storehouse OpenMP of CPU to it
Parallelization acceleration is carried out, compressed three-dimensional modeling data is encapsulated into customized model NAL unit afterwards；It is described original
Three-dimensional modeling data includes vertex coordinates data and dough sheet index data, and the vertex coordinates data refers to each in threedimensional model
The coordinate value on vertex, for representing the location information of each vertex in three dimensions；The dough sheet index number residence refers to three-dimensional
Which vertex each dough sheet has form in model, for representing the link information of dough sheet in threedimensional model；As preference, institute
It states dough sheet index data and includes the starting point according to texture maps under the tactic dough sheet index of camera perspective, each camera perspective
Coordinate and each camera perspective lower panel starting point press original three-dimensional model data using Meshsequential algorithms
Contracting the specific implementation process is as follows：
S21, it is compressed for threedimensional model dough sheet index data：First index datastore of first dough sheet is absolute
Value, the index data of remaining dough sheet are stored as the relative value based on previous position index data, until all dough sheets index
Data finish compression and preservation；Dough sheet of the present invention is triangle surface, and each dough sheet includes three index values, respectively
For representing the index value on three vertex, first index data of first dough sheet refers to first vertex of first dough sheet
Index value；As preference, the absolute value that is stored as preserves partial data for first index data of first dough sheet,
The index datastore of remaining dough sheet is that the index value of each dough sheet after relative value is is done with previous plane piece index
Difference, and the compressed data indexed using the difference as the dough sheet and preservation；
S22, it is compressed for threedimensional model vertex coordinates data：First all vertex coordinates datas are quantified, afterwards
Coding is compressed using entropy coding；As preference, all vertex coordinates datas are quantified apex coordinate
Data quantization to 16bit, the entropy coding uses rANS algorithms；
S23, before the first frame data are encoded, the camera parameter under each visual angle is transmitted through the network to decoding end；
S3, the step S1 video NAL units formed and the customized model NAL units of step S2 are encapsulated into ts stream packets, shape
Into ts code streams；
Step 2: the network transmission stage：The network transmission stage by the ts code streams after packing, is transmitted through the network to client
End；
Step 3: client decoding stage, specific decoding process is as follows：
T1, the step S3 ts stream packets formed are unpacked, obtains NAL unit；
T2, the NAL unit type obtained according to step T1, isolate texture NAL unit and threedimensional model NAL unit；
X264 Video Decoders and Meshsequential models is respectively adopted in T3, texture NAL unit and threedimensional model NAL unit
Decoder is operated in sub-line journey and unziped it, and respectively obtains the threedimensional model number after the data texturing and decompression after decompression
According to, and the three-dimensional modeling data after the data texturing and decompression after decompression is stored in buffering area；Compared with sub-line journey, main thread is used
In entire decoding process, and main thread includes sub-line journey.The tool unziped it using Meshsequential model decoders
Body realizes that process is as follows：
T31, according to coding side compression process, read first index data of first dough sheet, be afterwards converted into relative value
Absolute value realizes the decoding of dough sheet index data；As preference, it is by n-th that relative value described in step T32, which is converted into absolute value,
The index information of dough sheet is added to obtain this dough sheet index value with the dough sheet index value of previous position, recovers dough sheet index value, Mei Gexiang
Texture maps starting point coordinate and each visual angle lower panel starting point under machine visual angle；
T32, carry out entropy coding inverse process and quantify inverse process, decompress vertex coordinates data；
T33, indexed for each group of dough sheet, by texture under camera parameter, each visual angle dough sheet starting point and each camera perspective
Figure starting point coordinate calculates the texture coordinate of vertex correspondence in every group of dough sheet index data, as preference, every to calculate herein
The texture coordinate of three vertex correspondences in group dough sheet index data, texture projection is carried out according to uv；
When T4, area to be buffered are full, according to every mono- frames of 30ms, successively by the data texturing after all decompressions in buffering area according to three-dimensional
On texture coordinate map to threedimensional model in model, a textured threedimensional model video frame that can be shown is formed；
T5, by constructing three-dimensional rendering scene, shown successively per frame textured threedimensional model video frame, realize and play band in real time
The threedimensional model video of texture.
As preference, the Video coding storehouse of step S1 and the video decoding storehouse of step T3 are libx264.
Test data
（1）It is 150k for the continuous number of vertex of 200 frames, the texture of the threedimensional model that dough sheet number is 270k and 2048x2048
Figure utilizes a kind of online real-time three-dimensional model video coding-decoding method of the present invention, the average volume of one frame model video data of coding
The code time is 180ms, and the average decoding time of one frame data of decoding is 300ms, i.e., excessive in processing three-dimensional modeling data amount
During threedimensional model video, by client buffer, area to be buffered is full, it can be achieved that real-time exhibition threedimensional model；
（2）For the texture maps of 200 threedimensional models that frame number of vertex is 50k, dough sheet number is 100k and 2048x2048, utilize
A kind of online real-time three-dimensional model video coding-decoding method of the present invention, the average coding time of one frame model video data of coding are
80ms, decoding end decode the average decoding time of a frame data as 120ms, i.e., in the excessive three-dimensional of processing three-dimensional modeling data amount
During model video, by client buffer, area to be buffered is full, it can be achieved that real-time exhibition threedimensional model；
（3）For the texture maps of 200 threedimensional models that frame number of vertex is 11k, dough sheet number is 22k and 2048x2048, utilize
A kind of online real-time three-dimensional model video coding-decoding method of the present invention, the average coding time of one frame model video data of coding are
28ms, decoding end decode the average decoding time of a frame data as 26ms, and realization plays textured threedimensional model video in real time.
Threedimensional model video coding-decoding method i.e. of the present invention is in the few threedimensional model video of processing three-dimensional modeling data amount, encoding and decoding one
Frame data can reach within 30ms.
Furthermore, it is necessary to explanation, the few threedimensional model video of three-dimensional modeling data amount of the present invention refer to number of vertex
Within 20k, dough sheet number is within 30k, at this point, one frame data of encoding and decoding can reach within 30ms；The threedimensional model number
Refer to that number of vertex is more than 20k according to excessive threedimensional model video is measured, dough sheet number is more than 30k, at this time, it may be necessary to delay by client
Punching, area to be buffered are full, it can be achieved that real-time exhibition threedimensional model.
The above described is only a preferred embodiment of the present invention, not make limitation in any form to the present invention, though
So the present invention is described as above with preferred embodiment, however is not limited to the present invention, any to be familiar with this professional technology people
Member, without departing from the scope of the present invention, when a little change or modification made using technology contents described above
It is the equivalent embodiment of equivalent variations, as long as being without departing from technical solution of the present invention content, technical spirit according to the invention
To any simple modification, equivalent change and modification that above example is made, in the range of still falling within technical solution of the present invention.
Claims (7)
1. a kind of online real-time three-dimensional model video coding-decoding method, this method includes server side code stage, network transmission rank
Section and client decoding stage, it is characterised in that：It is as follows to implement step：
Step 1: the server side code stage, the server side code stage includes texture graph code and threedimensional model encodes, specifically
Realization process is as follows：
S1, texture graph code：Video compress is carried out to original texture diagram data using H.264 video encoding standard, is compressed
Texture map data afterwards, and it is encapsulated into video NAL units；
S2, threedimensional model coding：Original three-dimensional model data are compressed using Meshsequential algorithms, are compressed
Compressed three-dimensional modeling data is encapsulated into customized model NAL unit by three-dimensional modeling data afterwards afterwards；The original
Beginning three-dimensional modeling data includes vertex coordinates data and dough sheet index data, using Meshsequential algorithms to initial three-dimensional
Model data be compressed the specific implementation process is as follows：
S21, it is compressed for threedimensional model dough sheet index data：First index datastore of first dough sheet is absolute
Value, the index data of remaining dough sheet is stored as the relative value based on the index data of previous plane piece, until all faces
Piece index data finishes compression and preservation；
S22, it is compressed for threedimensional model vertex coordinates data：First all vertex coordinates datas are quantified, afterwards
Coding is compressed using entropy coding；
S23, before the first frame data are encoded, the camera parameter under each visual angle is transmitted through the network to decoding end；
S3, the step S1 video NAL units formed and the customized model NAL units of step S2 are encapsulated into ts stream packets, shape
Into ts code streams；
Step 2: the network transmission stage：The network transmission stage by the ts code streams after packing, is transmitted through the network to client
End；
Step 3: client decoding stage, specific decoding process is as follows：
T1, the step S3 ts stream packets formed are unpacked, obtains NAL unit；
T2, the NAL unit type obtained according to step T1, isolate texture NAL unit and threedimensional model NAL unit；
X264 Video Decoders and Meshsequential models is respectively adopted in T3, texture NAL unit and threedimensional model NAL unit
Decoder is operated in sub-line journey and unziped it, and respectively obtains the threedimensional model number after the data texturing and decompression after decompression
According to, and the three-dimensional modeling data after the data texturing and decompression after decompression is stored in buffering area；Using Meshsequential moulds
It is that type decoder unzips it the specific implementation process is as follows：
T31, according to coding side compression process, read first dough sheet index data, relative value be converted into absolute value afterwards, it is real
Existing dough sheet index data decoding；
T32, carry out entropy coding inverse process and quantify inverse process, decompress vertex coordinates data；
T33, indexed for each group of dough sheet, by texture under camera parameter, each visual angle dough sheet starting point and each camera perspective
Figure starting point coordinate calculates the texture coordinate of vertex correspondence in every group of dough sheet index data, and texture throwing is carried out according to texture coordinate
Shadow；
When T4, area to be buffered are full, according to every mono- frames of 30ms, successively by the data texturing after all decompressions in buffering area according to three-dimensional
On texture coordinate map to threedimensional model in model, a textured threedimensional model video frame is formed；
T5, by constructing three-dimensional rendering scene, shown successively per frame textured threedimensional model video frame, realize and play band in real time
The threedimensional model video of texture.
2. a kind of online real-time three-dimensional model video coding-decoding method according to claim 1, it is characterised in that：Step S1
The original texture diagram data is that the texture maps under each camera perspective are stored according to visual angle order to the line of a 2048x2048
It manages in figure, and by texture maps starting point coordinate under each camera perspective and each visual angle dough sheet starting point according to dough sheet indexed mode
After storage is indexed to dough sheet.
3. a kind of online real-time three-dimensional model video coding-decoding method according to claim 1, it is characterised in that：Step S2
The dough sheet index data includes the starting according to texture maps under the tactic dough sheet index of camera perspective, each camera perspective
Point coordinates and each camera perspective lower panel starting point.
4. a kind of online real-time three-dimensional model video coding-decoding method according to claim 1, it is characterised in that：Step
Absolute value is stored as described in S21 and preserves partial data, the index number of remaining dough sheet for first index data of first dough sheet
According to each dough sheet being stored as after relative value is index value with previous plane piece index make the difference, and using the difference as
The compressed data of dough sheet index and preservation.
5. a kind of online real-time three-dimensional model video coding-decoding method according to claim 1, it is characterised in that：Step
Pair all vertex coordinates datas described in S22 are quantified vertex coordinates data being quantized to 16bit, and the entropy coding is adopted
With rANS algorithms.
6. a kind of online real-time three-dimensional model video coding-decoding method according to claim 1, it is characterised in that：Step S1
The video of Video coding storehouse and step T3 decoding storehouse are libx264.
7. a kind of online real-time three-dimensional model video coding-decoding method according to claim 1, it is characterised in that：Step
Relative value described in T31 is converted into absolute value as the index information of n-th of dough sheet to be added to obtain with the dough sheet index value of previous position
This dough sheet index value recovers dough sheet index value, the texture maps starting point coordinate under each camera perspective and each visual angle lower panel
Starting point.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201711406406.XA CN108111863A (en) | 2017-12-22 | 2017-12-22 | A kind of online real-time three-dimensional model video coding-decoding method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201711406406.XA CN108111863A (en) | 2017-12-22 | 2017-12-22 | A kind of online real-time three-dimensional model video coding-decoding method |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN108111863A true CN108111863A (en) | 2018-06-01 |
Family
ID=62212386
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201711406406.XA Pending CN108111863A (en) | 2017-12-22 | 2017-12-22 | A kind of online real-time three-dimensional model video coding-decoding method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN108111863A (en) |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109255429A (en) * | 2018-07-27 | 2019-01-22 | 中国人民解放军国防科技大学 | Parameter decompression method for sparse neural network model |
| CN109408425A (en) * | 2018-12-27 | 2019-03-01 | 中科院计算技术研究所南京移动通信与计算创新研究院 | A kind of high speed magnanimity Flash memory module based on VPX |
| CN110971906A (en) * | 2018-09-29 | 2020-04-07 | 上海交通大学 | Hierarchical point cloud code stream packaging method and system |
| CN112929705A (en) * | 2021-02-02 | 2021-06-08 | 腾讯科技（深圳）有限公司 | Texture compression and decompression method and device, computer equipment and storage medium |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1946180A (en) * | 2006-10-27 | 2007-04-11 | 北京航空航天大学 | Three dimension mode compression coding/decoding method based on octree |
| CN101610411A (en) * | 2009-07-16 | 2009-12-23 | 中国科学技术大学 | A kind of method and system of video sequence mixed encoding and decoding |
| CN101859443A (en) * | 2010-06-22 | 2010-10-13 | 清华大学 | Self-adaptive triangular multi-resolution grid vertex position predicting method and device |
| CN105046725A (en) * | 2015-07-10 | 2015-11-11 | 清华大学 | Head shoulder image reconstruction method based on model and object in low bit rate video call |
| CN105139449A (en) * | 2015-08-24 | 2015-12-09 | 上海卫高网络科技有限公司 | Three-dimensional model compression method based on three-dimensional mesh subdivision and coding |
| US20170301110A1 (en) * | 2015-08-26 | 2017-10-19 | Electronic Arts Inc. | Producing three-dimensional representation based on images of an object |
-
2017
- 2017-12-22 CN CN201711406406.XA patent/CN108111863A/en active Pending
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1946180A (en) * | 2006-10-27 | 2007-04-11 | 北京航空航天大学 | Three dimension mode compression coding/decoding method based on octree |
| CN101610411A (en) * | 2009-07-16 | 2009-12-23 | 中国科学技术大学 | A kind of method and system of video sequence mixed encoding and decoding |
| CN101859443A (en) * | 2010-06-22 | 2010-10-13 | 清华大学 | Self-adaptive triangular multi-resolution grid vertex position predicting method and device |
| CN105046725A (en) * | 2015-07-10 | 2015-11-11 | 清华大学 | Head shoulder image reconstruction method based on model and object in low bit rate video call |
| CN105139449A (en) * | 2015-08-24 | 2015-12-09 | 上海卫高网络科技有限公司 | Three-dimensional model compression method based on three-dimensional mesh subdivision and coding |
| US20170301110A1 (en) * | 2015-08-26 | 2017-10-19 | Electronic Arts Inc. | Producing three-dimensional representation based on images of an object |
Cited By (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109255429A (en) * | 2018-07-27 | 2019-01-22 | 中国人民解放军国防科技大学 | Parameter decompression method for sparse neural network model |
| CN109255429B (en) * | 2018-07-27 | 2020-11-20 | 中国人民解放军国防科技大学 | Parameter decompression method for sparse neural network model |
| CN110971906A (en) * | 2018-09-29 | 2020-04-07 | 上海交通大学 | Hierarchical point cloud code stream packaging method and system |
| CN109408425A (en) * | 2018-12-27 | 2019-03-01 | 中科院计算技术研究所南京移动通信与计算创新研究院 | A kind of high speed magnanimity Flash memory module based on VPX |
| CN112929705A (en) * | 2021-02-02 | 2021-06-08 | 腾讯科技（深圳）有限公司 | Texture compression and decompression method and device, computer equipment and storage medium |
| CN112929705B (en) * | 2021-02-02 | 2022-04-22 | 腾讯科技（深圳）有限公司 | Texture compression and decompression method and device, computer equipment and storage medium |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108111863A (en) | A kind of online real-time three-dimensional model video coding-decoding method | |
| Lengyel | Compression of time-dependent geometry | |
| Briceño Pulido | Geometry videos: a new representation for 3D animations | |
| US6947045B1 (en) | Coding of animated 3-D wireframe models for internet streaming applications: methods, systems and program products | |
| CN1906945B (en) | Method and apparatus for scalable video encoding and decoding | |
| CN104885471B (en) | Coding/uncoded data content-adaptive entropy coding for next-generation video | |
| CN107113432A (en) | Rate Control for parallel video coding | |
| CN104469378A (en) | Coding position data for the last non-zero transform coefficient in a coefficient group | |
| Noimark et al. | Streaming scenes to MPEG-4 video-enabled devices | |
| CN102239693A (en) | Moving picture decoding method and moving picture encoding method | |
| TWI511528B (en) | Encoder, decoder and method | |
| JP2016226001A (en) | Decoder and decoding method | |
| CN102246526A (en) | Moving picture decoding method and device, and moving picture encoding method and device | |
| CN107231557A (en) | Reset for the intelligence in the recurrence block subregion of the senior infra-frame prediction in Video coding | |
| IL143438A (en) | Method and system for predictive encoding of arrays of data | |
| Yang et al. | Real-time 3d video compression for tele-immersive environments | |
| CN107409212A (en) | The gradual renewal using conversion coefficient for coding and decoding | |
| Stengel et al. | A distributed, decoupled system for losslessly streaming dynamic light probes to thin clients | |
| KR102238091B1 (en) | System and method for 3d model compression and decompression | |
| Chew et al. | A fuzzy clustering algorithm for virtual character animation representation | |
| CN102065293B (en) | Image compression method based on space domain predictive coding | |
| CN105519108B (en) | The weight predicting method and device of quantization matrix coding | |
| US20230290009A1 (en) | Remeshing for efficient compression | |
| RU2016101046A (en) | Encoder, decoder and method of operation using interpolation | |
| Stefanoski et al. | Spatially and temporally scalable compression of animated 3D meshes with MPEG-4/FAMC |
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: 20180601 |