CN111711810B - Stereo video transmission method based on asymmetric code rate distribution - Google Patents

Stereo video transmission method based on asymmetric code rate distribution Download PDF

Info

Publication number
CN111711810B
CN111711810B CN202010617647.4A CN202010617647A CN111711810B CN 111711810 B CN111711810 B CN 111711810B CN 202010617647 A CN202010617647 A CN 202010617647A CN 111711810 B CN111711810 B CN 111711810B
Authority
CN
China
Prior art keywords
video
code rate
quality
model
stereoscopic video
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
Application number
CN202010617647.4A
Other languages
Chinese (zh)
Other versions
CN111711810A (en
Inventor
赵铁松
吴雨旋
翟宇轩
陈炜玲
魏宏安
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Fuzhou University
Original Assignee
Fuzhou University
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by Fuzhou University filed Critical Fuzhou University
Priority to CN202010617647.4A priority Critical patent/CN111711810B/en
Publication of CN111711810A publication Critical patent/CN111711810A/en
Application granted granted Critical
Publication of CN111711810B publication Critical patent/CN111711810B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Classifications

    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
    • H04N13/00—Stereoscopic video systems; Multi-view video systems; Details thereof
    • H04N13/10—Processing, recording or transmission of stereoscopic or multi-view image signals
    • H04N13/194—Transmission of image signals
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L65/00—Network arrangements, protocols or services for supporting real-time applications in data packet communication
    • H04L65/60—Network streaming of media packets
    • H04L65/65—Network streaming protocols, e.g. real-time transport protocol [RTP] or real-time control protocol [RTCP]
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L65/00—Network arrangements, protocols or services for supporting real-time applications in data packet communication
    • H04L65/60—Network streaming of media packets
    • H04L65/70—Media network packetisation
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L65/00—Network arrangements, protocols or services for supporting real-time applications in data packet communication
    • H04L65/80—Responding to QoS
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
    • H04N13/00—Stereoscopic video systems; Multi-view video systems; Details thereof
    • H04N13/10—Processing, recording or transmission of stereoscopic or multi-view image signals
    • H04N13/106—Processing image signals
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
    • H04N13/00—Stereoscopic video systems; Multi-view video systems; Details thereof
    • H04N13/10—Processing, recording or transmission of stereoscopic or multi-view image signals
    • H04N13/106—Processing image signals
    • H04N13/161—Encoding, multiplexing or demultiplexing different image signal components

Landscapes

  • Engineering & Computer Science (AREA)
  • Multimedia (AREA)
  • Signal Processing (AREA)
  • Computer Networks & Wireless Communication (AREA)
  • Testing, Inspecting, Measuring Of Stereoscopic Televisions And Televisions (AREA)

Abstract

The invention relates to a three-dimensional video transmission method based on asymmetric code rate distribution, which comprises the following steps: step S1, establishing a stereo video objective quality model based on the stereo video perception quality under the condition that the left and right viewpoint code rates are asymmetric; step S2, according to the stereoscopic video audience quality model, expressing the process of predicting the network bandwidth and the process of the cache state change by using a mathematical model, and establishing a stereoscopic video code rate adaptive transmission algorithm model based on the HTTP dynamic adaptive streaming media technology; and step S3, in the stereoscopic video transmission system, according to a stereoscopic video code rate adaptive transmission algorithm model based on the HTTP dynamic adaptive streaming media technology, providing an optimal left and right viewpoint code rate combination for each video clip downloading request, and completing the transmission of the stereoscopic video. The invention can effectively predict the quality of the stereo video and realize the self-adaptive transmission of the asymmetric code rate.

Description

Stereo video transmission method based on asymmetric code rate distribution
Technical Field
The invention relates to the technical field of video transmission, in particular to a three-dimensional video transmission method based on asymmetric code rate distribution.
Background
With the development of science and technology, the demand of users on multimedia information is not limited to the traditional flat video, the appearance of the stereo video provides more real viewing experience for users, media interaction means are enriched, and the method becomes one of the development directions of digital multimedia technology. The stereoscopic video has wide market prospect, the most widely applied stereoscopic video is the film industry at present, and 3D film watching is the mainstream entertainment mode of the public. In addition, the 3D digital television enables users to enjoy vivid three-dimensional viewing experience without going out of home; scenes such as a video conference system, remote medical treatment, remote education, remote industrial control and the like can also adopt a stereo video to enhance interactivity; there is also a potential need in many areas, including virtual reality systems, telerobots, auto-navigation, consumer electronics.
In recent years, the video service market has grown at a high rate. The conventional vod service has been unable to meet the user's requirements, subject to the differences in the end user's equipment performance, network bandwidth, quality of demand, etc. In this case, Dynamic Adaptive Streaming over HTTP (DASH) technology based on HTTP becomes a new choice for Streaming service providers. In such schemes, the scheduling algorithm dynamically selects a suitable code rate in the video transmission process according to the network bandwidth, the buffer status and other factors of the user.
The development of streaming media services simultaneously promotes the rise of online stereoscopic video services, and the internet-based stereoscopic video transmission is an economical transmission mode compatible with the existing plane video distribution. The binocular stereo video generates depth information in the brain through binocular parallax, provides immersive viewing experience for a user, and the stereo display equipment presents slightly different left and right viewpoint videos shot in the same scene to the left and right eyes of the user respectively and generates stereo perception through fusing image information of the left and right viewpoints through the brain. The stereoscopic video has a larger data volume due to the dual-viewpoint characteristic, and the transmission optimization problem is more important.
Disclosure of Invention
In view of this, an object of the present invention is to provide a method for transmitting a stereoscopic video based on asymmetric rate allocation, which can provide videos with different rates according to different network bandwidth conditions, so that the rate change is smoother.
In order to achieve the purpose, the invention adopts the following technical scheme:
a stereoscopic video transmission method based on asymmetric code rate distribution comprises the following steps:
step S1, establishing a stereo video objective quality model based on the stereo video perception quality under the condition that the left and right viewpoint code rates are asymmetric;
step S2, according to the stereoscopic video audience quality model, expressing the process of predicting the network bandwidth and the process of the cache state change by using a mathematical model, and establishing a stereoscopic video code rate adaptive transmission algorithm model based on the HTTP dynamic adaptive streaming media technology;
and step S3, in the stereoscopic video transmission system, according to a stereoscopic video code rate adaptive transmission algorithm model based on the HTTP dynamic adaptive streaming media technology, providing an optimal left and right viewpoint code rate combination for each video clip downloading request, and completing the transmission of the stereoscopic video.
Further, the step S1 is specifically:
step S11, constructing a stereo video subjective database;
step S12, obtaining subjective experiment data of the stereo video based on a single-stimulus subjective test method; (ii) a
Step S13, preprocessing the data of the subjective experiment to obtain the subjective quality scores Q of the left and right viewpoints of the reliable test sequence1And Q2And the average quality score Q of the stereoscopic video;
step S14, constructing a linear model of the visual stimulation intensity of the left and right viewpoints according to the subjective experimental data, and calculating to obtain the visual stimulation intensity of the left and right viewpoints;
step S15, constructing a weight function of the left and right viewpoints according to the visual stimulation intensity of the left and right viewpoints to obtain the weights of the left and right viewpoints;
step S16, establishing a model of the quality of the stereo video based on the obtained weights of the left and right viewpoints;
and step S17, establishing a linear model by using the video space information SI and the time information TI to predict the parameters to be predicted. Solving parameters to be predicted based on least square fitting;
and step S18, obtaining a final stereoscopic video objective quality model according to the linear model of the prediction parameters and the model of the stereoscopic video quality.
Further, the linear model of the visual stimulation intensity of the left and right viewpoints is specifically:
Figure BDA0002561926420000031
wherein s ismaxFor maximum resolution of video, s1And s2Resolution of left and right viewpoints of original video, s1/smaxAnd s2/smaxThe blurring distortion intensity of left and right viewpoints determined by the resolution ratio; q. q.sminTo minimize the quantization parameter, q1And q is2Quantized values of left and right viewpoints, q, respectively, of the original videomin/q1And q ismin/q2Is the block effect distortion strength of left and right view points determined by quantization parameters, a and b are parameters to be predicted, c is constant, g1And g2The stimulus intensity of the left and right viewpoints is indicated.
Further, the weighting functions of the left and right viewpoints are as follows:
Figure BDA0002561926420000041
wherein, w1And w2The weights of the left and right views, respectively.
Further, the model of the stereoscopic video quality specifically includes:
Q＝w1·Q1+w2·Q2
wherein Q is1And Q2The subjective quality scores of the left viewpoint and the right viewpoint are respectively, and Q is the average quality score of the stereo video.
Further, the step S17 is specifically:
calculating the SI used to express the amount of detail in the image spatial domain:
SI＝maxtime{stdspace[Sobel(Fn)]},
wherein, FnFor video frame information, Sobel is the edge detection operator, stdspaceTaking the maximum value as SI for calculating the standard deviation of the filtered image;
calculating the TI used for expressing the change situation of the video content:
TI＝maxtime{stdspace[Fn(i,j)-Fn-1(i,j)]}
wherein, Fn(i, j) is the pixel at the ith and jth lines of the nth frame in time, the standard deviation of the pixel difference values of the successive frames is calculated, and the maximum value is taken as TI
Establishing the following linear model prediction parameters a and b according to the video spatial information SI and the time information TI:
Figure BDA0002561926420000051
wherein, as、at、ac、bs、bt、bcAre all coefficients.
Further, the step S2 is specifically:
step S21, using the throughput of downloading the previous video segment as the predicted bandwidth
Figure BDA0002561926420000052
Predicting the bandwidth x required to be divided into left and right view channels1[n]And x2[n]：
Figure BDA0002561926420000053
Step S22, monitoring the current buffer length of the client player as B [ n ], wherein n represents the nth video segment of the starting request, and establishing a user experience quality model as follows:
Figure BDA0002561926420000054
wherein, w1、w2Is a weight, alpha, beta are parameters relating to the video content, r1[n]、r2[n]Are the left and right view code rates, u, k, b0For coefficients, t is the video segment length, E is for B [ n ]]Function of (c):
Figure BDA0002561926420000055
wherein p is a coefficient, BrefThe target cache length is larger than 0 and smaller than the maximum cache length.
And step S23, finding the optimal left and right viewpoint code rate combination by adopting an optimized exhaustion method according to the user experience quality model.
Further, the step S2 is specifically: the code rate of the default left viewpoint is always higher than that of the default right viewpoint; when the network bandwidth condition improves, the current predicted bandwidth value is larger than the predicted value of the previous video segment downloading, namely when the network bandwidth condition improves
Figure BDA0002561926420000061
Then, the following solving process is carried out:
Figure BDA0002561926420000062
wherein
Figure BDA0002561926420000063
And
Figure BDA0002561926420000064
to make it possible to
Figure BDA0002561926420000065
Maximum current left and right view code rate, r1[n]And r2[n]For left and right view rates, R, of the current videoLIs the maximum selectable view point code rate;
when the network condition is deteriorated, the current predicted bandwidth value is smaller than the previous onePrediction value when downloading video segments, i.e. when
Figure BDA0002561926420000066
Then, the following solving process:
Figure BDA0002561926420000067
wherein
Figure BDA0002561926420000068
And
Figure BDA0002561926420000069
to make it possible to
Figure BDA00025619264200000610
Maximum current left and right view code rate, r1[n-1]And r2[n-1]Left and right view rates, R, for a previous video1Minimum view rate, R, selectable for left view2A minimum view bitrate that is selectable for the right view.
Further, the step S3 specifically includes:
step S31, establishing a stereo video transmission system, including a server and a client, where the server is a streaming media server supporting DASH and storing coded video segments and MPD files, the client is a streaming media player with code rate adaptive algorithm, and the server and the client form a local area network through a router;
step S32, Wondershaper software is installed on the server, the throughput limit of the Linux server is dynamically changed over time through the software, and the change of network bandwidth in the local area network is simulated;
step S33, modifying the MPD file of the server, and increasing the information of SI, TI, alpha and beta of the video;
and step S33, the client downloads the MPD file on the server according to the IP address in the local area network, analyzes the MPD file to acquire all related information of the selectable video segments on the server, then selects a proper code rate according to the predicted bandwidth and the cache state by the adaptive algorithm and sends a request to the server, and finally decodes and renders the downloaded video segments into stereo video to finish the transmission of the stereo video.
Compared with the prior art, the invention has the following beneficial effects:
1. the invention can effectively predict the quality of the stereo video, and the QoE model designed on the basis can guide the allocation of left and right viewpoint code rates to realize the self-adaptive transmission of asymmetric code rates, thereby effectively improving the transmission quality;
2. the invention can better adapt to the fluctuation of bandwidth, so that the code rate change is smoother.
Drawings
FIG. 1 is a flow chart of the operation of an embodiment of the present invention;
FIG. 2 is a block diagram of an objective quality model for stereoscopic video according to an embodiment of the present invention;
FIG. 3 is a video bitrate adaptation process in an embodiment of the invention;
fig. 4 is a DASH-based stereoscopic video transmission system according to an embodiment of the present invention.
Detailed Description
The invention is further explained below with reference to the drawings and the embodiments.
Referring to fig. 1, the present invention provides a method for transmitting a stereoscopic video based on asymmetric code rate allocation, comprising the following steps:
step S1, establishing a stereo video objective quality model based on the stereo video perception quality under the condition that the left and right viewpoint code rates are asymmetric;
step S2, according to the stereoscopic video audience quality model, expressing the process of predicting the network bandwidth and the process of the cache state change by using a mathematical model, and establishing a stereoscopic video code rate adaptive transmission algorithm model based on the HTTP dynamic adaptive streaming media technology;
and step S3, in the stereoscopic video transmission system, according to a stereoscopic video code rate adaptive transmission algorithm model based on the HTTP dynamic adaptive streaming media technology, providing an optimal left and right viewpoint code rate combination for each video clip downloading request, and completing the transmission of the stereoscopic video.
Referring to fig. 2, the stereoscopic video quality model takes the planar video quality, resolution, and quantization parameters of the left and right viewpoints as input and the stereoscopic video quality model as output in this example, thereby constructing the stereoscopic video quality model.
In this embodiment, the step S1 specifically includes:
step S11, constructing a stereo video subjective database;
step S12, obtaining the subjective quality score of the three-dimensional video based on the single-stimulus subjective testing method;
step S13, preprocessing the data of the subjective experiment to obtain the subjective quality scores Q of the left and right viewpoints of the reliable test sequence1And Q2And the average quality score Q of the stereoscopic video;
step S14, constructing a linear model of the visual stimulation intensity of the left and right viewpoints according to the subjective experimental data, and calculating to obtain the visual stimulation intensity of the left and right viewpoints;
the linear model of the visual stimulation intensity of the left and right viewpoints is specifically as follows:
Figure BDA0002561926420000091
wherein s ismaxFor maximum resolution of video, s1And s2Resolution of left and right viewpoints of original video, s1/smaxAnd s2/smaxThe blurring distortion intensity of left and right viewpoints determined by the resolution ratio; q. q.sminTo minimize the quantization parameter, q1And q is2Quantized values of left and right viewpoints, q, respectively, of the original videomin/q1And q ismin/q2Is the block effect distortion strength of left and right view points determined by quantization parameters, a and b are parameters to be predicted, c is constant, g1And g2The stimulus intensity of the left and right viewpoints is indicated.
Step S15, constructing a weight function of the left and right viewpoints according to the visual stimulation intensity of the left and right viewpoints to obtain the weights of the left and right viewpoints; the weighting functions of the left and right viewpoints are as follows:
Figure BDA0002561926420000092
wherein, w1And w2The weights of the left and right views, respectively.
Step S16, establishing a model of the quality of the stereo video based on the obtained weights of the left and right viewpoints; the model of the stereo video quality specifically comprises:
Q＝w1·Q1+w2·Q2
wherein Q is1And Q2The subjective quality scores of the left viewpoint and the right viewpoint are respectively, and Q is the average quality score of the stereo video.
And step S17, establishing a linear model by using the video space information SI and the time information TI to predict the parameters to be predicted. And solving the parameters to be predicted based on least square fitting to obtain a linear model of the predicted parameters.
Calculating the SI used to express the amount of detail in the image spatial domain:
SI＝maxtime{stdspace[Sobel(Fn)]},
wherein, FnFor video frame information, Sobel is the edge detection operator, stdspaceTaking the maximum value as SI for calculating the standard deviation of the filtered image;
calculating the TI used for expressing the change situation of the video content:
TI＝maxtime{stdspace[Fn(i,j)-Fn-1(i,j)]}
wherein, Fn(i, j) is the pixel at the ith and jth lines of the nth frame in time, the standard deviation of the pixel difference values of the successive frames is calculated, and the maximum value is taken as TI
Establishing the following linear model prediction parameters a and b according to the video spatial information SI and the time information TI:
Figure BDA0002561926420000101
wherein, as、at、ac、bs、bt、bcAre all coefficients.
And step S18, obtaining a final stereoscopic video objective quality model according to the linear model of the prediction parameters and the model of the stereoscopic video quality.
Referring to fig. 3, a stereoscopic video user experience quality model is constructed according to a stereoscopic video quality model and a buffer status model to obtain an optimal bitrate, thereby constructing a stereoscopic video bitrate adaptive scheme
In an embodiment, the step S2 specifically includes:
step S21, using the throughput of downloading the previous video segment as the predicted bandwidth
Figure BDA0002561926420000102
Predicting the bandwidth x required to be divided into left and right view channels1[n]And x2[n]：
Figure BDA0002561926420000103
Step S22, monitoring the current buffer length of the client player and recording as B [ n ], wherein n represents that the nth video segment is requested, and establishing a user quality of experience (QoE) model as follows:
Figure BDA0002561926420000111
wherein, w1、w2Is a weight, alpha, beta are parameters relating to the video content, r1[n]、r2[n]Are the left and right view code rates, u, k, b0For coefficients, t is the video segment length, E is for B [ n ]]Function of (c):
Figure BDA0002561926420000112
wherein p is a coefficient, BrefTo the eyesMarking the cache length, wherein the marking cache length is larger than 0 and smaller than the maximum value of the cache length.
And step S23, finding the optimal left and right viewpoint code rate combination by adopting an optimized exhaustion method according to the user experience quality model.
In this embodiment, preferably, the step S23 specifically includes: the code rate of the default left viewpoint is always higher than that of the default right viewpoint; when the network bandwidth condition improves, the current predicted bandwidth value is larger than the predicted value of the previous video segment downloading, namely when the network bandwidth condition improves
Figure BDA0002561926420000113
Then, the following solving process is carried out:
Figure BDA0002561926420000114
wherein
Figure BDA0002561926420000115
And
Figure BDA0002561926420000116
to make it possible to
Figure BDA0002561926420000117
Maximum current left and right view code rate, r1[n]And r2[n]For left and right view rates, R, of the current videoLIs the maximum selectable view point code rate;
when the network condition deteriorates, the current predicted bandwidth value will be less than the predicted value of the previous video segment download, i.e. when the network condition deteriorates
Figure BDA0002561926420000118
Then, the following solving process:
Figure BDA0002561926420000121
wherein
Figure BDA0002561926420000122
And
Figure BDA0002561926420000123
to make it possible to
Figure BDA0002561926420000124
Maximum current left and right view code rate, r1[n-1]And r2[n-1]Left and right view rates, R, for a previous video1Minimum view rate, R, selectable for left view2A minimum view bitrate that is selectable for the right view.
Referring to fig. 4, in this embodiment, the stereoscopic video transmission system is composed of a server and a client, where the server is a streaming server supporting DASH and stores encoded video segments and MPD files, the client is a streaming player with a bitrate adaptive algorithm, and the server and the client form a local area network through a router.
In an embodiment, the step S3 specifically includes:
step S31, establishing a stereo video transmission system, including a server and a client, where the server is a streaming media server supporting DASH and storing coded video segments and MPD files, the client is a streaming media player with code rate adaptive algorithm, and the server and the client form a local area network through a router;
step S32, Wondershaper software is installed on the server, the throughput limit of the Linux server is dynamically changed over time through the software, and the change of network bandwidth in the local area network is simulated;
step S33, modifying the MPD file of the server, and increasing the information of SI, TI, alpha and beta of the video;
and step S33, the client downloads the MPD file on the server according to the IP address in the local area network, analyzes the MPD file to acquire all related information of the selectable video segments on the server, then selects a proper code rate according to the predicted bandwidth and the cache state by the adaptive algorithm and sends a request to the server, and finally decodes and renders the downloaded video segments into stereo video to finish the transmission of the stereo video.
The above description is only a preferred embodiment of the present invention, and all equivalent changes and modifications made in accordance with the claims of the present invention should be covered by the present invention.

Claims (8)

1. A three-dimensional video transmission method based on asymmetric code rate distribution is characterized by comprising the following steps:
step S1, establishing a stereo video objective quality model based on the stereo video perception quality under the condition that the left and right viewpoint code rates are asymmetric;
step S2, according to the stereoscopic video audience quality model, expressing the process of predicting the network bandwidth and the process of the cache state change by using a mathematical model, and establishing a stereoscopic video code rate adaptive transmission algorithm model based on the HTTP dynamic adaptive streaming media technology;
step S3, in the stereo video transmission system, according to the stereo video code rate adaptive transmission algorithm model based on the HTTP dynamic adaptive streaming media technology, providing the optimal left and right viewpoint code rate combination for each video clip downloading request, and completing the transmission of the stereo video;
the step S1 specifically includes:
step S11, constructing a stereo video subjective database;
step S12, obtaining subjective experiment data of the stereo video based on a single-stimulus subjective test method;
step S13, preprocessing the data of the subjective experiment to obtain the subjective quality scores Q of the left and right viewpoints of the reliable test sequence1And Q2And the average quality score Q of the stereoscopic video;
step S14, constructing a linear model of the visual stimulation intensity of the left and right viewpoints according to the subjective experimental data, and calculating to obtain the visual stimulation intensity of the left and right viewpoints;
step S15, constructing a weight function of the left and right viewpoints according to the visual stimulation intensity of the left and right viewpoints to obtain the weights of the left and right viewpoints;
step S16, establishing a model of the quality of the stereo video based on the obtained weights of the left and right viewpoints;
step S17, establishing a linear model by utilizing the video space information SI and the time information TI to predict the parameters to be predicted, and solving the parameters to be predicted based on least square fitting;
and step S18, obtaining a final stereoscopic video objective quality model according to the linear model of the prediction parameters and the model of the stereoscopic video quality.
2. The method for transmitting stereoscopic video based on asymmetric rate allocation according to claim 1, wherein the linear models of the visual stimulus intensities of the left and right viewpoints are specifically:
Figure FDA0003038577530000021
wherein s ismaxFor maximum resolution of video, s1And s2Resolution of left and right viewpoints of original video, s1/smaxAnd s2/smaxThe blurring distortion intensity of left and right viewpoints determined by the resolution ratio; q. q.sminTo minimize the quantization parameter, q1And q is2Quantized values of left and right viewpoints, q, respectively, of the original videomin/q1And q ismin/q2Is the block effect distortion strength of left and right view points determined by quantization parameters, a and b are parameters to be predicted, c is constant, g1And g2The stimulus intensity of the left and right viewpoints is indicated.
3. The method for stereoscopic video transmission based on asymmetric rate allocation according to claim 2,
the weighting functions of the left and right viewpoints are as follows:
Figure FDA0003038577530000022
wherein, w1And w2The weights of the left and right views, respectively.
4. The method for transmitting stereoscopic video based on asymmetric rate allocation according to claim 3, wherein the model of stereoscopic video quality is specifically:
Q＝w1·Q1+w2·Q2
wherein Q is1And Q2The subjective quality scores of the left viewpoint and the right viewpoint are respectively, and Q is the average quality score of the stereo video.
5. The method for transmitting stereoscopic video based on asymmetric rate allocation according to claim 2, wherein the step S15 specifically comprises:
calculating the SI used to express the amount of detail in the image spatial domain:
SI＝maxtime{stdspace[Sobel(Fn)]},
wherein, FnFor video frame information, Sobel is the edge detection operator, stdspaceTaking the maximum value as SI for calculating the standard deviation of the filtered image;
calculating the TI used for expressing the change situation of the video content:
TI＝maxtime{stdspace[Fn(i,j)-Fn-1(i,j)]}
wherein, Fn(i, j) calculating the standard deviation of the pixel difference values of the continuous frames for the pixels at the ith row and the jth row of the nth frame in time, and taking the maximum value of the standard deviation as TI;
the parameters a and b are predicted by establishing the following linear model according to the video space information SI and the time information TI:
Figure FDA0003038577530000031
wherein, as、at、ac、bs、bt、bcAre all coefficients.
6. The method for transmitting stereoscopic video based on asymmetric rate allocation according to claim 1, wherein the step S2 specifically comprises:
step S21, using the throughput of downloading the previous video segment as the predicted bandwidth
Figure FDA0003038577530000045
Predicting the bandwidth x required to be divided into left and right view channels1[n]And x2[n]：
Figure FDA0003038577530000041
Step S22, monitoring the current buffer length of the client player as B [ n ], wherein n represents the serial number of the nth video segment which is requested, and establishing a user experience quality model as follows:
Figure FDA0003038577530000042
wherein, w1、w2Is a weight, alpha, beta are parameters relating to the video content, r1[n]、r2[n]The code rates of the left and right viewpoints, u, k, and b0For coefficients, t is the video segment length, E is for B [ n ]]Function of (c):
Figure FDA0003038577530000043
wherein p is a coefficient, BrefThe target cache length is greater than 0 and smaller than the maximum value of the cache length;
and step S23, finding the optimal left and right viewpoint code rate combination by adopting an optimized exhaustion method according to the user experience quality model.
7. The method for transmitting stereoscopic video based on asymmetric rate allocation according to claim 6, wherein the step S2 specifically comprises: the code rate of the default left viewpoint is always higher than that of the default right viewpoint; bandwidth of networkWhen the situation improves, the current predicted bandwidth value is larger than the predicted bandwidth value of the previous video segment downloading, namely, when the situation improves
Figure FDA0003038577530000044
Then, the following solving process is carried out:
Figure FDA0003038577530000051
wherein r is1 *[n]And
Figure FDA0003038577530000052
to make it possible to
Figure FDA0003038577530000053
Maximum current left and right view code rate, r1[n]And r2[n]For left and right view rates, R, of the current videoLIs the maximum selectable view point code rate;
when the network condition deteriorates, the current predicted bandwidth value will be less than the predicted value of the previous video segment download, i.e. when the network condition deteriorates
Figure FDA0003038577530000054
Then, the following solving process:
Figure FDA0003038577530000055
wherein r is1 *[n]And
Figure FDA0003038577530000056
to make it possible to
Figure FDA0003038577530000057
Maximum current left and right view code rate, r1[n-1]And r2[n-1]Left and right view rates, R, for a previous video1Is a left viewpointSelected minimum view rate, R2A minimum view bitrate that is selectable for the right view.
8. The method for transmitting stereoscopic video based on asymmetric rate allocation according to claim 1, wherein the step S3 specifically comprises:
step S31: establishing a three-dimensional video transmission system, which comprises a server and a client, wherein the server is a streaming media server supporting DASH and stores coded video segments and MPD files, the client is a streaming media player with a code rate adaptive algorithm, and the server and the client form a local area network through a router;
step S32: wondershaper software is installed in a server, the throughput limit of a Linux server is dynamically changed over time through the software, and the change of network bandwidth in a local area network is simulated;
step S33: modifying an MPD file of a server, and increasing SI, TI, alpha and beta information of a video, wherein alpha and beta are parameters related to video content;
step S33: the client downloads an MPD file on the server according to an IP address in the local area network, analyzes the MPD file to acquire all related information of the selectable video segments on the server, then selects a proper code rate according to the predicted bandwidth and the cache state by using a self-adaptive algorithm, sends a request to the server, and finally decodes and renders the downloaded video segments into stereo video to finish the transmission of the stereo video.
CN202010617647.4A 2020-06-30 2020-06-30 Stereo video transmission method based on asymmetric code rate distribution Active CN111711810B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202010617647.4A CN111711810B (en) 2020-06-30 2020-06-30 Stereo video transmission method based on asymmetric code rate distribution

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202010617647.4A CN111711810B (en) 2020-06-30 2020-06-30 Stereo video transmission method based on asymmetric code rate distribution

Publications (2)

Publication Number Publication Date
CN111711810A CN111711810A (en) 2020-09-25
CN111711810B true CN111711810B (en) 2021-06-22

Family

ID=72543962

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202010617647.4A Active CN111711810B (en) 2020-06-30 2020-06-30 Stereo video transmission method based on asymmetric code rate distribution

Country Status (1)

Country Link
CN (1) CN111711810B (en)

Families Citing this family (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN113905221B (en) * 2021-09-30 2024-01-16 福州大学 Stereoscopic panoramic video asymmetric transport stream self-adaption method and system

Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN103391436A (en) * 2013-07-12 2013-11-13 宁波大学 Stereo video bitrate control method based on binocular vision characteristic
CN106454317A (en) * 2016-11-15 2017-02-22 天津大学 Three-dimensional video quality adaptation algorithm based on fuzzy control
CN106888374A (en) * 2015-12-16 2017-06-23 联芯科技有限公司 A kind of 3 d video encoding method, device and video processing equipment
CN109379632A (en) * 2018-10-25 2019-02-22 中国地质大学（武汉） A kind of progressive switching method of code rate and system of dynamic self-adapting HTTP stream

Patent Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN103391436A (en) * 2013-07-12 2013-11-13 宁波大学 Stereo video bitrate control method based on binocular vision characteristic
CN106888374A (en) * 2015-12-16 2017-06-23 联芯科技有限公司 A kind of 3 d video encoding method, device and video processing equipment
CN106454317A (en) * 2016-11-15 2017-02-22 天津大学 Three-dimensional video quality adaptation algorithm based on fuzzy control
CN109379632A (en) * 2018-10-25 2019-02-22 中国地质大学（武汉） A kind of progressive switching method of code rate and system of dynamic self-adapting HTTP stream

Non-Patent Citations (2)

* Cited by examiner, † Cited by third party
Title
《基于HTTP的动态自适应流媒体传输算法》;侯永宏等;《吉林大学学报（工学版）》;20180326;第1-3节 *
《基于HTTP自适应流媒体传输的 3D 视频质量评价》;翟宇轩等;《北京航空航天大学学报》;20190822;第1-2节 *

Also Published As

Publication number Publication date
CN111711810A (en) 2020-09-25

Similar Documents

Publication Publication Date Title
US10817988B2 (en) Method and apparatus for streaming data
US10817985B2 (en) Apparatuses and methods for performing artificial intelligence encoding and artificial intelligence decoding on image
US20200219292A1 (en) Methods and apparatuses for performing artificial intelligence encoding and artificial intelligence decoding on image
US10904639B1 (en) Server-side fragment insertion and delivery
CN110324621B (en) Video encoding method, video encoding device, electronic equipment and storage medium
Yuan et al. Spatial and temporal consistency-aware dynamic adaptive streaming for 360-degree videos
EP2364190B1 (en) Centralized streaming game server
Duanmu et al. Quality-of-experience of adaptive video streaming: Exploring the space of adaptations
CN113905221B (en) Stereoscopic panoramic video asymmetric transport stream self-adaption method and system
CN112584119B (en) Self-adaptive panoramic video transmission method and system based on reinforcement learning
CN109792545A (en) The transmission of spectators' importance adaptive bitrate
CN106303562B (en) Multi-view point video adaptive transmitted control algorithm based on PI control
US20210264567A1 (en) Apparatus and method for performing artificial intelligence encoding and artificial intelligence decoding on image by using pre-processing
CN109792547A (en) The transmission of viewer's attention degree adaptive bitrate
Devlic et al. QoE-aware optimization for video delivery and storage
Liu et al. QoE-oriented 3D video transcoding for mobile streaming
Yao et al. Video streaming adaptation strategy for multiview navigation over DASH
CN111711810B (en) Stereo video transmission method based on asymmetric code rate distribution
JP5507770B2 (en) Video encoding method, apparatus, program, and recording medium
Janowski et al. Content driven QoE assessment for video frame rate and frame resolution reduction
CN117440209B (en) Implementation method and system based on singing scene
Murroni et al. Guest editorial special issue on quality of experience for advanced broadcast services
CN107124603A (en) A kind of multi-view point video adaptive transmission method based on caching
Karn et al. User-perceived quality aware adaptive streaming of 3D multi-view video plus depth over the internet
Tanjung et al. Qoe optimization in dash-based multiview video streaming

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