WO2014114109A1 - 一种视频流质量监测方法及装置 - Google Patents

一种视频流质量监测方法及装置 Download PDF

Info

Publication number
WO2014114109A1
WO2014114109A1 PCT/CN2013/084795 CN2013084795W WO2014114109A1 WO 2014114109 A1 WO2014114109 A1 WO 2014114109A1 CN 2013084795 W CN2013084795 W CN 2013084795W WO 2014114109 A1 WO2014114109 A1 WO 2014114109A1
Authority
WO
WIPO (PCT)
Prior art keywords
frame
video frame
score
current video
quality
Prior art date
Application number
PCT/CN2013/084795
Other languages
English (en)
French (fr)
Inventor
吴宝春
魏芳
王雨
Original Assignee
中兴通讯股份有限公司
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 中兴通讯股份有限公司 filed Critical 中兴通讯股份有限公司
Publication of WO2014114109A1 publication Critical patent/WO2014114109A1/zh

Links

Classifications

    • HELECTRICITY
    • H04ELECTRIC COMMUNICATION TECHNIQUE
    • H04NPICTORIAL COMMUNICATION, e.g. TELEVISION
    • H04N17/00Diagnosis, testing or measuring for television systems or their details
    • H04N17/004Diagnosis, testing or measuring for television systems or their details for digital television systems
    • HELECTRICITY
    • H04ELECTRIC COMMUNICATION TECHNIQUE
    • H04LTRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L1/00Arrangements for detecting or preventing errors in the information received
    • H04L1/20Arrangements for detecting or preventing errors in the information received using signal quality detector
    • H04L1/201Frame classification, e.g. bad, good or erased
    • HELECTRICITY
    • H04ELECTRIC COMMUNICATION TECHNIQUE
    • H04LTRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L1/00Arrangements for detecting or preventing errors in the information received
    • H04L1/20Arrangements for detecting or preventing errors in the information received using signal quality detector
    • H04L1/205Arrangements for detecting or preventing errors in the information received using signal quality detector jitter monitoring
    • HELECTRICITY
    • H04ELECTRIC COMMUNICATION TECHNIQUE
    • H04NPICTORIAL COMMUNICATION, e.g. TELEVISION
    • H04N19/00Methods or arrangements for coding, decoding, compressing or decompressing digital video signals
    • H04N19/85Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using pre-processing or post-processing specially adapted for video compression
    • HELECTRICITY
    • H04ELECTRIC COMMUNICATION TECHNIQUE
    • H04NPICTORIAL COMMUNICATION, e.g. TELEVISION
    • H04N21/00Selective content distribution, e.g. interactive television or video on demand [VOD]
    • H04N21/60Network structure or processes for video distribution between server and client or between remote clients; Control signalling between clients, server and network components; Transmission of management data between server and client, e.g. sending from server to client commands for recording incoming content stream; Communication details between server and client 
    • H04N21/63Control signaling related to video distribution between client, server and network components; Network processes for video distribution between server and clients or between remote clients, e.g. transmitting basic layer and enhancement layers over different transmission paths, setting up a peer-to-peer communication via Internet between remote STB's; Communication protocols; Addressing
    • H04N21/647Control signaling between network components and server or clients; Network processes for video distribution between server and clients, e.g. controlling the quality of the video stream, by dropping packets, protecting content from unauthorised alteration within the network, monitoring of network load, bridging between two different networks, e.g. between IP and wireless
    • H04N21/64723Monitoring of network processes or resources, e.g. monitoring of network load

Definitions

  • the present invention relates to the field of mobile communications, and in particular, to a video stream quality monitoring method and apparatus. Background technique
  • Online monitoring of video stream service quality can be implemented on the service side, the network side, and the terminal side.
  • the service quality monitoring on the service side mainly monitors the quality provided by the service side and cannot reflect the quality of the user experience on the terminal side.
  • the terminal-side service quality monitoring needs to be deployed on the user terminal. Generally, the user needs to obtain the user's consent to deploy. The deployment is difficult and can only reflect the experience of the individual user. It is difficult to reflect the experience of the network community users.
  • Some terminal-side service quality monitoring methods also affect the user's use of the service, which affects the user's service experience.
  • the network-side service quality monitoring can be deployed on each network node without user cooperation or user impact.
  • Multiple service sessions passing through the network node can be monitored at the same time to reflect the quality of the network responsible for the network node. If it is deployed on a network node that is close to the user terminal, it cannot directly reflect the user experience as the terminal-side method, but it can also reflect the business running status.
  • the network side service data throughput is large, especially for video services. Therefore, the data processing speed is very high, and the corresponding quality monitoring method cannot be too complicated.
  • the original data of the frequency service can only be evaluated by analyzing the video service stream. This is called the no-reference method in the media quality evaluation.
  • the no-reference method has always been a difficult point in the quality evaluation field, or the accuracy is not high, or it is too complicated.
  • the monitoring basis of the network-side service quality monitoring needs to be the data packet that has arrived at the current time, and the quality status needs to be dynamically updated according to the currently received data packet, which requires the monitoring method to process the real-time service flow data in real time.
  • the existing non-reference method applicable to the network side generally considers the impact of source video coding and network performance on video quality, but differs in specific coverage factors.
  • ITU-T G.1070 proposes a more complete non-reference frame for the quality evaluation of video telephony services, but when evaluating video quality, source video coding only considers the encoder type, bit rate and frame rate.
  • the network performance only considers the packet loss factor;
  • the source video coding of the Moving Picture Quality Metric (MPQM) model only gives a fixed score according to the type of the encoder, and the network performance only considers the packet loss. , the monitoring accuracy of the video stream quality is low.
  • MPQM Moving Picture Quality Metric
  • Other models have similar problems.
  • the video stream quality monitoring method on the network side of the related art has a low monitoring accuracy and a complicated implementation.
  • the embodiment of the invention provides a video stream quality monitoring method and device for solving the above technical problem.
  • An embodiment of the present invention provides a video stream quality monitoring method, where the method includes: acquiring video stream information on a network side; determining, according to the video stream information, a quality score of a reference frame, and a coding quality score of a current video frame. According to the quality score of the reference frame, the current Determining a source coding quality score according to the coding quality score of the video frame and the coding type of the current video frame; determining a network impairment score according to the video stream information; wherein the network impairment is caused by packet loss and delay jitter Video stream quality impairment; determining a quality monitoring score for the video stream based on the source coding quality score and the network impairment score.
  • the obtaining the video stream information on the network side comprises: decapsulating and demultiplexing the video service network data on the network side to obtain a video stream; parsing the video stream, and positioning the current video frame and the previous video frame in the video. a starting point and an ending point in the stream; respectively mapping the starting point and the ending point of the current video frame, and the starting point and ending point of the previous video frame, respectively, to the video service network data packet, respectively Determining the arrival time of all network packets of the current video frame and the previous video frame; determining the current video frame arrival time and the previous video frame arrival time according to the arrival time respectively; according to the end point and the end of the previous video frame Determining an end point of the current video frame, determining a number of bits of the current video frame; determining a real-time packet loss rate, a real-time bit rate, and a real-time frame rate of the current video frame; parsing the encoded bit stream of the current video frame, determining The encoding type, size information, and
  • determining the quality score of the reference frame according to the video stream information includes: determining, according to the encoded bitstream of the current video frame, whether the current video frame is a reference frame; determining that the current video frame is a reference frame, Performing a block effect evaluation operation and a fuzzy evaluation operation on the reference frame; performing weighted averaging on the result of the block effect evaluation operation and the result of the fuzzy evaluation operation to obtain a quality score of the reference frame.
  • determining, according to the video stream information, the coding quality score of the current video frame comprises: obtaining, according to the number of bits and size information of the current video frame, a number of bits per pixel of the current video frame; The number of bits and the quantization parameter are determined by a determining operation corresponding to the encoding type of the current video frame, and the time and space complexity are obtained; and the current video frame is obtained according to the time and space complexity and the quantization parameter.
  • the coding quality score is: obtaining, according to the number of bits and size information of the current video frame, a number of bits per pixel of the current video frame; The number of bits and the quantization parameter are determined by a determining operation corresponding to the encoding type of the current video frame, and the time and space complexity are obtained; and the current video frame is obtained according to the time and space complexity and the quantization parameter.
  • the coding quality of the current video frame And determining the source coding quality score according to the coding type of the current video frame, if the coding type of the current video frame is an intra prediction I frame, and the current video frame is a reference frame, the source coding The quality score is equal to the quality score of the reference frame; if the encoding type of the current video frame is an intra prediction I frame, the current video frame is not a reference frame, and the quality score of the reference frame exists, according to the a quality score of the reference frame, an encoding quality score of the current video frame, and a coding quality difference between the current video frame and the reference frame, determining the source coding quality score; if the coding type of the current video frame is Intra prediction I frame, the current video frame is not a reference frame, and there is no quality score of the reference frame, the source coding quality score is equal to an encoding quality score of the current video frame; if the current video frame The coding type is forward prediction
  • determining the network impairment score according to the video stream information includes: setting a first time window length; wherein, the end point of the first time window is the current video frame, and the beginning of the first time window a point is a video frame whose time interval from the current video frame is not greater than the length of the first time window; determining a real-time average packet loss rate according to a real-time packet loss rate of each video frame in the first time window; Determining a real-time average bit rate of the real-time bit rate of each video frame in the first time window; determining a real-time average frame rate according to a real-time frame rate of each video frame in the first time window; Determining a packet loss robustness factor according to the bit rate and the real-time average frame rate; determining a packet loss score according to the packet loss robustness factor and the real-time average packet loss rate; according to the current video frame arrival time The real-time average frame rate, and the real-time average frame rate of the arrival time of the previous video frame, determining a
  • determining the quality monitoring score of the video stream according to the source coding quality score and the network impairment score comprises: determining, according to the source coding quality score, the packet loss impairment score, Determining a dynamic score of the current video frame; setting a second time window length; wherein the end point of the second time window is the current video frame, and the beginning of the second time window Pointing a video frame whose time interval from the current video frame is not greater than the length of the second time window; averaging the dynamic scores of the respective video frames in the second time window to obtain the quality of the video stream Monitor the score.
  • an embodiment of the present invention further provides a video stream quality monitoring apparatus, where the apparatus includes: an information acquiring module configured to acquire video stream information on a network side; and a source coding quality scoring module And configured to determine, according to the video stream information, a quality score of the reference frame, and a coding quality score of the current video frame; determining the source coding according to the quality score of the reference frame, the coding quality score of the current video frame, and the coding type of the current video frame.
  • Network impairment scoring module configured to determine network impairment score according to video stream information; wherein, network impairment is video stream quality damage caused by packet loss and delay jitter; quality monitoring scoring module, configured to be based on source coding quality Scoring and network impairment scores, which determine the quality monitoring score of the video stream.
  • the information acquiring module includes: a video stream obtaining sub-unit configured to de-encapsulate and de-multiplex the video service network data on the network side to obtain a video stream; and a time determining sub-unit configured to parse the video stream, Locating a start point and an end point of the current video frame and a previous video frame in the video stream; respectively, starting and ending points of the current video frame, and starting and ending points of the previous video frame, respectively Mapping to the video service network data packet, respectively, obtaining an arrival time of all network packets of the current video frame and the previous video frame; determining, according to the arrival time, a current video frame arrival time and a previous video frame arrival time; a number determining subunit, configured to determine a number of bits of the current video frame according to an end point of the previous video frame and an end point of the current video frame; and an encoding information determining subunit configured to determine the current video Real-time packet loss rate, real-time bit rate, and real-time frame rate of the frame; par
  • the source coding quality scoring module includes a quality scoring unit of the reference frame, configured to determine a quality score of the reference frame according to the video stream information, where the quality scoring unit of the reference frame includes: a reference frame determiner a unit, configured to determine, according to the encoded bitstream of the current video frame, whether the current video frame is a reference frame; the evaluation subunit, configured to perform a blockiness evaluation operation and a fuzzy evaluation on the reference frame if it is a reference frame
  • the quality scoring sub-unit of the reference frame is configured to perform weighted averaging on the result of the block-effect evaluation operation and the result of the fuzzy evaluation operation to obtain a quality score of the reference frame.
  • the source coding quality scoring module includes an encoding quality scoring unit configured to determine an encoding quality score of a current video frame according to the video stream information, where the encoding quality scoring unit includes: a per-pixel number determining determinant a unit, configured to obtain, according to the number of bits and size information of the current video frame, a number of bits per pixel of the current video frame; a time and space complexity determining subunit configured to use the number of bits per pixel and the Quantizing a parameter, taking a determining operation corresponding to the encoding type of the current video frame, to obtain time and space complexity; and a coding quality scoring subunit configured to obtain the according to the time and space complexity and the quantization parameter The encoding quality score of the current video frame.
  • the encoding quality scoring unit includes: a per-pixel number determining determinant a unit, configured to obtain, according to the number of bits and size information of the current video frame, a number of bits per pixel of the current video frame; a time and space complexity determining
  • the source coding quality scoring module includes a source coding quality scoring unit configured to determine a source according to a quality score of the reference frame, an encoding quality score of the current video frame, and an encoding type of the current video frame.
  • the coding quality scoring unit includes: a first scoring sub-unit, configured to: when the coding type of the current video frame is an intra prediction I frame, and the current video frame is a reference frame The source code quality score is equal to the quality score of the reference frame; the second score subunit is configured to: the coding type of the current video frame is an intra prediction I frame, and the current video frame is not a reference frame, And if there is a quality score of the reference frame, determining a quality according to a quality score of the reference frame, an encoding quality score of the current video frame, and a coding quality difference between the current video frame and a reference frame.
  • a source coding quality score configured to be in the coding class of the current video frame
  • the type is an intra prediction I frame, where the current video frame is not a reference frame, and if there is no quality score of the reference frame, the source coding quality score is equal to the coding quality score of the current video frame
  • a scoring sub-unit configured to: obtain a quality monitoring score of a previous video frame if the encoding type of the current video frame is a forward prediction P frame; and monitor a score according to a quality of the previous video frame, the current The coding quality score of the video frame, and the difference in coding quality between the current video frame and the previous video frame, the source coding quality score is determined.
  • the network impairment scoring module includes: a first setting unit configured to set a first time window length; wherein, an end point of the first time window is the current video frame, the first time The starting point of the window is a video frame whose time interval from the current video frame is not greater than the length of the first time window; the averaging unit is configured to be based on a real-time packet loss rate of each video frame in the first time window, Determining a real-time average packet loss rate; determining a real-time average bit rate according to a real-time bit rate of each video frame in the first time window; determining a real-time average according to a real-time frame rate of each video frame in the first time window a frame loss; a packet loss scoring unit configured to determine a packet loss robustness factor according to the real-time average bit rate and the real-time average frame rate; a delay jitter scoring unit configured to arrive according to the current video frame The real-time average frame rate of the moment, and the real-time average frame rate of the arrival time
  • the quality monitoring and scoring module includes: a dynamic scoring unit, configured to determine a dynamic score of a current video frame according to the source coding quality score, the packet loss impairment score, and the delay jitter score; a determining unit, configured to set a second time window length; wherein, an ending point of the second time window is the current video frame, and a starting point of the second time window is a time interval from the current video frame a video frame not greater than the length of the second time window; a quality monitoring scoring unit configured to move to each video frame in the second time window The state scores are averaged to obtain a quality monitoring score for the video stream.
  • a dynamic scoring unit configured to determine a dynamic score of a current video frame according to the source coding quality score, the packet loss impairment score, and the delay jitter score
  • a determining unit configured to set a second time window length; wherein, an ending point of the second time window is the current video frame, and a starting point of the second time window is a time interval from the current video frame a video frame
  • the source coding quality score and the network impairment score are obtained according to the video stream information, where the reference frame is used in the process of determining the source coding quality score, in the network damage.
  • the video stream quality damage caused by packet loss and delay jitter is considered.
  • the quality monitoring score of the video stream is determined according to the source coding quality score and the network impairment score, and the network side of the related technology is solved.
  • the video stream quality monitoring method has low monitoring accuracy and complicated implementation, and more accurately reflects the dynamic change of video quality, improves the accuracy of video stream quality monitoring, and has low complexity, which can be conveniently applied.
  • FIG. 1 is a flowchart of a video stream quality monitoring method according to an embodiment of the present invention
  • FIG. 2 is a flowchart of acquiring video stream information on the network side according to an embodiment of the present invention
  • FIG. 3 is a flow chart of quality scoring of a reference frame according to an embodiment of the present invention.
  • FIG. 5 is a flowchart of source coding quality scoring according to an embodiment of the present invention.
  • FIG. 6 is a flowchart of a network impairment score according to an embodiment of the present invention.
  • FIG. 7 is a schematic structural diagram of a video stream quality monitoring apparatus 1 according to an embodiment of the present invention
  • FIG. 8 is a schematic structural diagram of a video stream quality monitoring apparatus 2 according to an embodiment of the present invention.
  • FIG. 1 is a flowchart of a video stream quality monitoring method according to an embodiment of the present invention. As shown in FIG. 1, the method includes:
  • Step S102 Obtain video stream information on the network side.
  • the video stream information provides a calculation basis for subsequent determination of the source coding quality score and the network impairment score.
  • the video stream information includes: an encoded bit stream of a current video frame, a bit number, a size information, a coding type, a quantization parameter, a real-time packet loss rate, a real-time bit rate, a real-time frame rate, and a current video frame of a current video frame.
  • step S104 determining, according to the video stream information, a quality score of the reference frame, and an encoding quality score of the current video frame; according to the quality score of the reference frame, the current video frame a coding quality score, and an encoding type of the current video frame, determining a source coding quality score;
  • Step S106 determining a network impairment score according to the video stream information; wherein the network impairment is caused by packet loss and delay jitter Caused video stream quality damage;
  • Step S108 Determine a quality monitoring score of the video stream according to the source coding quality score and the network impairment score.
  • the source coding quality score and the network impairment score are obtained according to the video stream information, where the reference frame is used in determining the source coding quality score, and the network impairment score is used. In the process of determining, the video stream quality damage caused by packet loss and delay jitter is considered. Finally, the quality monitoring score of the video stream is determined according to the source coding quality score and the network impairment score, and the network side video in the related technology is solved.
  • the flow quality monitoring method has low monitoring accuracy and complicated implementation, and accurately reflects the dynamic change of video quality, improves the accuracy of video stream quality monitoring, and has low complexity, which can be conveniently applied to the network side. Video quality is monitored in real time.
  • the embodiment provides a preferred implementation manner, that is, acquiring the video stream information on the network side includes: Decapsulating and demultiplexing video service network data on the network side to obtain a video stream; parsing the video stream, and locating a start point and an end point of the current video frame and a previous video frame in the video stream; The start point and the end point of the frame, and the start point and the end point of the previous video frame are respectively inversely mapped to the video service network data packet, respectively, to obtain all networks of the current video frame and the previous video frame respectively.
  • the arrival time of the packet determining the current video frame arrival time and the previous video frame arrival time respectively; determining the number of bits of the current video frame according to the end point of the previous video frame and the end point of the current video frame; a real-time packet loss rate, a real-time bit rate, and a real-time frame rate of the current video frame; parsing an encoded bit stream of the current video frame, determining an encoding type, a size information, and a quantization parameter of the current video frame.
  • the calculation basis is provided for the subsequent determination of the source coding quality score and the network impairment score, and the monitoring accuracy and accuracy of the video stream quality are mentioned.
  • step S202-step S210 is a flowchart of acquiring video stream information on the network side according to an embodiment of the present invention. As shown in FIG. 2, the process includes the following steps (step S202-step S210):
  • step S202 the video service network data is decapsulated and demultiplexed to obtain a video stream.
  • the video service network data is decapsulated and demultiplexed to obtain a video stream, for example, in a real-time transport protocol (RTP).
  • RTP real-time transport protocol
  • UDP User Datagram Protocol
  • IP Inter-network protocol
  • step S202 the video service network data is decapsulated and demultiplexed to obtain a video stream.
  • RTP real-time transport protocol
  • IP Inter-network protocol
  • IP Internet Protocol
  • Step S204 parsing the video stream, and searching for a start code, thereby positioning a start point and an end point of the coded bit stream of the current video frame in the video stream;
  • Step S206 Reverse mapping the start point and the end point of the current video frame to the video service network data packet, to obtain a sequence number and an arrival time of all network data packets used for encapsulating the current video frame;
  • Step S208 the arrival time of the video frame is obtained from the arrival time of all network packets of the video frame. Time, earliest arrival time and latest arrival time;
  • the earliest arrival time of the video frame is the arrival time of the first network data packet of the video frame
  • the arrival time and the latest arrival time are the arrival times of the last network data packet of the video frame.
  • Step S210 The interval between the end points of the two frames is the number of bits included in the current video frame, and the encoded bit stream of the current video frame is continued to be parsed, and other encoded information is extracted.
  • the other encoding information includes: an encoding type of the frame (I frame or P frame); a size (width and height) of the frame; a quantization parameter of the frame.
  • the frame header of each frame can be parsed in the H.261, H.263 video stream, and the sequence parameter set (SPS, Sequence Parameter Set) and image parameter set (PPS, Picture) can be parsed in the H.264 video stream. Parameter Set ).
  • the frame coding type of the current video frame is an I frame and the interval between the previous reference frame and the previous reference frame is greater than G GoPs, continue to determine whether it is a non-dropped I frame, and if yes, use it as a new reference frame, and The encoded bit stream of the frame is saved. If no, the frame cannot be used as a new reference frame.
  • the method for determining whether a frame has a packet loss is: the sequence number of the last network packet corresponding to the frame minus the number of the first network packet is equal to the number of network packets corresponding to the frame, and the frame is considered to be no packet loss. of.
  • the value of G can be determined according to the nature of the service, the accuracy requirements of the evaluation, and the real-time requirements of the evaluation.
  • the implementation of the method for acquiring video stream information in this embodiment is not limited to the process, as long as the video stream information can be accurately obtained.
  • the current video frame can be located by cooperating with the encapsulation protocol in the system. For example, if the video frame is directly encapsulated by using the RTP protocol, and the fragmentation unit (FU) method is adopted for the RTP encapsulation, the current frame can be located by the flag bit in the FU.
  • the FU fragmentation unit
  • the embodiment provides a preferred embodiment, that is, whether the current video frame is a reference frame according to the encoded bit stream of the current video frame; if it is a reference frame, performing a block effect evaluation operation and a fuzzy evaluation on the reference frame And performing weighted averaging on the result of the block effect evaluation operation and the result of the fuzzy evaluation operation to obtain a quality score of the reference frame.
  • the reference frame is introduced for quality scoring, which improves the monitoring accuracy of the video stream quality.
  • FIG. 3 is a flowchart of the quality scoring of the reference frame according to the embodiment of the present invention. As shown in FIG. 3, the process includes the following steps (step S302-step S310):
  • Step S302 determining whether the current video frame is a reference frame according to the encoded bit stream in the video stream information; specifically, determining whether the encoded bit stream of the current video frame exists, and if present, the current video frame is a reference frame; if not, The current video frame is a non-reference frame;
  • Step S304 performing video decoding on the current video frame.
  • Step S306 performing blockiness evaluation on the reconstructed image of the current video frame
  • the specific method can be implemented by: Using a simple block effect level statistical method to determine.
  • the luminance component of the reconstructed image is segmented into 8 x 8 blocks that do not overlap each other.
  • 8 X 8 luma block that is, there are 8 rows and 8 ⁇ ij
  • the block is considered to be "block effect is serious"; except for the edge, each 8
  • the x 8 luma block has four 8 X 8 luma blocks adjacent thereto, and if the difference between more than 5 of the 8 adjacent gray values of 3 or 4 adjacent edges is greater than the threshold, It is considered that the "block effect is severe" of the block; if two or four of the four adjacent edges occur adjacent to each other, the difference between the eight gray values is more than five, and the difference is considered to be the threshold value.
  • the block effect is generally ". All 8 x 8 blocks are classified according to the method, and the proportions of various blockiness degree blocks in the current video frame are respectively counted
  • Step S308 Perform fuzzy evaluation on the reconstructed image of the current frame.
  • the specific can be determined by a simple average edge width method: Using Sobel The vertical direction operator obtains the edge map. For an edge point in the edge map, find the extreme point near the corresponding point of the image (for example, the same line). The distance between the extreme points is the width of the edge. The edge width is obtained for all edge points, and the ratio of the sum of all edge widths to the number of edge points is the fuzzy evaluation of the image.
  • step S310 the weighted average of the results of the blockiness evaluation and the fuzzy evaluation is the quality score Qo of the reference frame.
  • the manner of the block effect evaluation and the fuzzy evaluation is not limited to the manner listed in the embodiment, as long as the quality score of the reference frame can be accurately obtained.
  • a specific implementation method of selecting a quality score of a reference frame may be selected according to service quality, evaluation accuracy requirements, and evaluation real-time requirements.
  • blockiness can be evaluated using block boundary difference evaluation based on singular value decomposition, or based on partial block difference and HVS block effect evaluation. Fuzzy can be evaluated by using edge contour gradient statistics and local kurtosis statistics.
  • the first lossless I frame in one or more group of pictures may be used as a reference frame, and the coding of the reference frame may be extracted from the video stream.
  • the bit stream after the reference frame is completely decoded, performs spatial domain no reference monitoring. This monitoring result is used as a scoring benchmark for other subsequent frames in the one or more GoPs.
  • the interval at which the reference frame appears can be selected based on the nature of the service, the accuracy requirements of the monitoring, and the real-time requirements for monitoring.
  • this embodiment For determining the coding quality score of the current video frame according to the video stream information in the step S104, this embodiment provides a preferred implementation manner, that is, according to the bit number and size information of the current video frame, each current video frame is obtained. a number of pixel bits; determining, according to the number of bits per pixel and the quantization parameter, a determination operation corresponding to a coding type of the current video frame, to obtain time and space complexity; and obtaining, according to time and space complexity and the quantization parameter, The encoding quality score of the current video frame.
  • the monitoring accuracy of the video stream quality is improved.
  • the larger the quantization parameter the larger the quantization step size and the lower the video quality.
  • the coding quality is also related to the complexity of time and space. Different time and space complexity will affect the slope of the linear model.
  • the temporal and spatial complexity can be characterized by the nature of the residual frame, where the quantization parameters of the frame and the number of bits per pixel after encoding are used. Therefore, the encoding quality of the current video frame can be determined according to the frame encoding type, the quantization parameter, the number of bits per pixel, and the time and space complexity.
  • the encoding quality of the current video frame is further measured in terms of the temporal and spatial complexity of the video frame, in addition to using conventional quantization parameters (QP, Quantization Parameter).
  • the temporal and spatial complexity of a video frame has the following characteristics: The temporal and spatial complexity of a video frame is simply and efficiently characterized using the quantization parameter QP of the video frame and the number of coded bits per pixel of the frame, and depending on the different coding types employed by the frame. Take different determination methods. Since real-time video services require high delay, generally only intra-prediction I frames and forward-predicted P-frames are used, instead of bi-predictive B-frames with delay introduced, so only two coding types are considered here, I Frame or P frame.
  • step S402-step S406 is a flowchart of coding quality scoring according to an embodiment of the present invention. As shown in FIG. 4, the flow includes the following steps (step S402-step S406):
  • Step S402 determining, according to the bit number B and the size (width W and height H) of the current video frame in the video stream information, the number of bits per pixel R of the current video frame;
  • the determining process of the number of bits per pixel R can be specifically implemented by the following formula: ⁇ ⁇
  • Step S404 determining a time and space complexity of the current frame according to the coding type of the current video frame and the quantization parameter Q P in the video stream information.
  • the time and space complexity is determined as follows: 2 ⁇ + ⁇ . If the current frame is a P frame, the time and space complexity is determined as follows: ⁇ 2 (a 2 x R + b 2 Qp), where the bi, ⁇ 2 and b 2 in the formula are model values of constant values Number.
  • Step S406 determining, according to the quantization parameter Q P , and the time and space complexity, the coding quality Q c of the current video frame;
  • ⁇ 3 , ⁇ 4 , and b 4 in the formula are model parameters of a constant value.
  • the source coding quality score can be determined.
  • the error propagation impairment assessment is used to measure the impact of error propagation on the current video frame and to determine the source coding quality score Q E . If the current frame is an I frame, the quality of the frame is completely determined by itself and is not affected by error propagation. If the current frame is a P frame, the quality of the frame is not completely determined by itself, but also affected by the quality of its reference frame. If the quality of its reference frame is degraded, this degradation will propagate to the frame. Real-time video services require strict delay, so multi-reference frame technology is generally not used. Therefore, only the influence of the previous frame on the error propagation of the current frame is considered here.
  • the embodiment For the determining operation of the source coding quality score in the step S104, the embodiment provides a preferred embodiment, that is, according to the quality score of the reference frame, the coding quality score of the current video frame, and the coding type of the current video frame. Determining the source coding quality score includes:
  • the source coding quality score is equal to the quality score of the reference frame
  • the current video frame is not a reference frame, and there is a quality score of the reference frame, according to the quality score of the reference frame, the encoding quality score of the current video frame, and the current video frame and a difference in coding quality between reference frames to determine a source coding quality score;
  • the source coding quality score is equal to the coding quality score of the current video frame; If the encoding type of the current video frame is a forward predicted P frame, the quality monitoring score of the previous video frame is obtained; the quality monitoring score according to the previous video frame, the encoding quality score of the current video frame, and the current video frame and the previous video frame The difference in coding quality between video frames determines the source coding quality score.
  • FIG. 5 is a flowchart of source coding quality scoring according to an embodiment of the present invention. As shown in FIG. 5, the process includes the following steps (step S502-step S506):
  • Step S502 If the current video frame is an I frame and is a reference frame (determining whether the encoded bit stream of the current video frame exists, if the current video frame is the reference frame if present), the source coding quality score
  • Step S504 if the current video frame is an I frame, and is not a reference frame, and the quality of the reference frame is scored by Q. Already exist, the source coding quality score Q E is scored on the quality of the reference frame. Based on, and considering the difference in coding quality between the current video frame and the reference frame, as follows: Q E . among them,
  • the determination of Qc,o is the same as the determination of Qc in step S406 in the embodiment.
  • Step S506 If the current video frame is a P frame, consider the error propagation of the current video frame by the previous video frame, that is, the final quality dynamic score Q prc of the video frame above the source coding quality score Q E is referenced, and consider the current video frame.
  • the difference in coding quality from the previous video frame as follows:
  • the network damage score is determined according to the video stream information.
  • the network impairment also considers the influence of the delay jitter factor on the video quality.
  • the effect of delay on video quality is not considered, because too much delay will result in packet loss. Think about losing the package.
  • the overall latency of the video traffic flow is initially buffered and can be mitigated by some additional means such as advertising.
  • the real-time average packet loss rate can be used instead of the commonly used real-time packet loss rate to make the dynamic scoring result more stable.
  • the real-time packet loss rate is the packet loss rate for a period of time before the current video frame arrives.
  • the real-time average packet loss rate is the average of the packet loss rate for a period of time before the current video frame arrival time.
  • the robustness factor of packet loss can be determined to improve the accuracy of video stream quality monitoring.
  • the tolerance for packet loss is different. The higher the real-time average bit rate or real-time frame rate, the lower the tolerance for packet loss.
  • the real-time bit rate is the video stream rate through the network node for a period of time before the current video frame arrives.
  • the real-time average bit rate is the video stream rate through the network node for a period of time before the current video frame arrives. Since the embodiment is applied to the network side, the frame rate refers to the number of video frames passing through the network node in a certain period of time.
  • the real-time frame rate is the number of video frames that pass through the network node for a period of time before the current video frame arrives.
  • the real-time average frame rate is the average number of video frames passing through the network node for a period of time before the current video frame arrives.
  • the jitter of the real-time average frame rate can be characterized.
  • the change in the real-time average frame rate is characterized by the difference in the real-time average frame rate of the video frame arrival time. Since the delay jitter of a video frame relative to the previous video frame can be alleviated by the terminal buffer on the terminal side, for the terminal, the real-time average frame rate change may cause the video to play fast and slow. , affecting the subjective experience of end users.
  • this embodiment provides a preferred implementation manner, that is, determining the network impairment score according to the video stream information includes: setting a first time window length; wherein, the end point of the first time window is a current video frame The start point of the first time window is a video frame whose time interval from the current video frame is not greater than the length of the first time window; according to each of the first time windows Real-time packet loss rate of the video frame, determining the real-time average packet loss rate; determining the real-time average bit rate according to the real-time bit rate of each video frame in the first time window; according to the real-time frame rate of each video frame in the first time window Determine the real-time average frame rate; determine the packet loss robustness factor according to the real-time average bit rate and the real-time average frame rate; determine the packet loss score according to the packet loss robustness factor and the real-time average packet loss rate; according to the current video frame The real-time average frame rate at the arrival time, and the real-time average frame rate of the arrival time of the
  • FIG. 6 is a flowchart of network impairment scoring according to an embodiment of the present invention. As shown in FIG. 6, the process includes the following steps (step S602 to step S620):
  • Step S602 setting a time window used for determining a real-time packet loss rate, a real-time average packet loss rate, a real-time bit rate, a real-time average bit rate, a real-time frame rate, and a real-time average frame rate.
  • the end of the time window 1 time window can be set as the current video frame, and the start point of the time window is the first video frame whose interval from the arrival time of the current video frame is not greater than T.
  • Step S604 determining a real-time packet loss rate PLR r of the current video frame arrival time
  • the method for determining the PLR r is: subtracting the difference between the sequence number of the last network packet corresponding to the current video frame minus the sequence number of the first network packet corresponding to the first video frame in the time window, and subtracting the time window The number of all network packets corresponding to all video frames, and the obtained difference is divided by the sequence number of the last network packet corresponding to the current video frame and the sequence number of the first network packet corresponding to the first video frame in the time window. The difference.
  • Step S606 determining a real-time average packet loss rate PLR a of the current video frame arrival time
  • the method of determining PLR a is: determining an average of real-time packet loss rates at the arrival times of all video frames in the time window. Using the real-time average packet loss rate instead of the real-time packet loss rate can make the evaluation result smoother.
  • Step S608 determining a real-time bit rate BR r of the current video frame arrival time;
  • the method of determining BR r is: dividing the number of bits of all video frames in the time window by the difference between the latest arrival time of the current video frame and the earliest arrival time of the first video frame in the time window.
  • Step S610 determining a real-time average bit rate BR a of the current video frame arrival time
  • the determined method is: determining the average of the real-time bit rates of all video frames arriving in the time window. Using the real-time average bit rate instead of the real-time bit rate makes the evaluation results smoother.
  • Step S612 determining a real-time frame rate FR r of the current video frame arrival time
  • the determined FR r method is: the number of all video frames in the time window, divided by the difference between the latest arrival time of the current video frame and the earliest arrival time of the first video frame in the time window.
  • Step S614 determining a real-time average frame rate FR a of the current video frame arrival time
  • the method of determining FR a is: determining an average of the real-time frame rates of all video frame arrival times in the time window. Using the real-time average frame rate instead of the real-time frame rate makes the results smoother.
  • Step S616 determining a packet loss robustness factor D plv of the current video frame arrival time
  • a 5 and b 5 in the formula are model parameters of a constant value.
  • Step S620 determining a delay jitter score DF of the current video frame arrival time.
  • the method of determining the DF is: the average frame rate jitter is the real-time average frame rate of the current video frame arrival time minus the real-time average frame rate of the previous frame arrival time.
  • time window smoothing techniques can be used. Set the time window length Tq, the end point of the time window is the current video frame, and the starting point of the time window is the first video frame with the current video frame time interval not greater than Tq.
  • the embodiment Based on the implementation process of the quality monitoring score of the video stream, the embodiment provides a preferred implementation manner, that is, the source coding quality score and the network impairment score, and determining the quality monitoring score of the video stream includes: according to the source coding quality score, Packet loss score, delay jitter score, determining the dynamic score of the current video frame; setting the second time window length; wherein, the end point of the second time window is the current video frame, and the start point of the second time window is current and current The video frame has a time interval not greater than the video frame of the second time window length; the dynamic scores of the respective video frames in the second time window are averaged to obtain a quality monitoring score of the video stream.
  • the accuracy and stability of the video stream quality monitoring result is improved.
  • FIG. 7 is a schematic structural diagram of a video stream quality monitoring apparatus 1 according to an embodiment of the present invention. As shown in FIG. 7, the apparatus includes: an information acquiring module 10, a source encoding quality scoring module 20, a network impairment scoring module 30, and a quality monitoring scoring module. 40. The structure is described in detail below.
  • the information acquiring module 10 is configured to acquire video stream information on the network side.
  • the source coding quality scoring module 20 is configured to determine, according to the video stream information, a quality score of the reference frame, and an encoding quality score of the current video frame; a quality score according to the reference frame, a coding quality score of the current video frame, and a current video frame.
  • Encoding type determining a source coding quality score
  • the network impairment scoring module 30 is configured to determine a network impairment score according to the video stream information; wherein, the network impairment is a video stream quality impairment caused by packet loss and delay jitter
  • the quality monitoring scoring module 40 is configured to determine a quality monitoring score of the video stream based on the source coding quality score and the network impairment score.
  • the source coding quality scoring module 20 and the network impairment scoring module 30 determine the source coding quality score and the network impairment score according to the video stream information, where
  • the reference frame is used in the process of determining the source coding quality score, and the video stream quality damage caused by packet loss and delay jitter is considered in the process of determining the network impairment score
  • the quality monitoring scoring module 40 is based on the source coding quality score and the location.
  • the network damage score is used to determine the quality monitoring score of the video stream, which solves the problem that the monitoring quality of the video stream quality monitoring method on the network side in the related art is not high and the implementation is complicated, and the dynamic change of the video quality is accurately reflected.
  • the accuracy of video stream quality monitoring is not high enough, and it can be easily applied to the network side video quality real-time monitoring environment.
  • the device may further include a network data acquiring module, where the network data acquiring module is configured to intercept the video stream from the network.
  • the present embodiment provides a preferred implementation manner, that is, the information acquiring module 10 includes: a video stream acquiring subunit, and the network data acquiring module is connected to the video service network on the network side.
  • the data is decapsulated and demultiplexed to obtain a video stream;
  • the time determining subunit is configured to parse the video stream, and locate a starting point and an ending point of the current video frame and the previous video frame in the video stream; and starting a current video frame And the end point, and the start point and the end point of the previous video frame, respectively, are inversely mapped to the video service network data packet, respectively, and the arrival times of all network packets of the current video frame and the previous video frame are respectively obtained; Determining the current video frame arrival time and the previous video frame arrival time respectively; the bit number determining subunit is configured to determine the number of bits of the current video frame according to the starting point of the previous video frame and the starting point of the current video frame; Determining a subunit configured to determine a real-time packet loss rate, a real-time bit rate, and Fps; parsing a current video frame encoded bit stream, determine encoding type, size information of the current video frame and a quantization parameter.
  • the information acquiring module 10 may further have the following features: the extracted video encoding information includes: an encoded bit stream of a reference frame; an encoding type of the frame (I frame or P frame); a frame size (width and height) ; quantization parameter Q P of the frame; number of bits per frame; arrival time of the frame.
  • the extracted network packet information includes a sequence number and an arrival time of all network packets used to encapsulate one video frame.
  • the source coding quality scoring module 20 For the source coding quality scoring module 20 to determine the quality score of the reference frame according to the video stream information, this embodiment provides a preferred implementation manner, that is, the source coding quality scoring module 20 includes a quality scoring unit of the reference frame, configured to be based on the video. The stream information is used to determine a quality score of the reference frame.
  • the quality scoring unit of the reference frame includes: a reference frame determining subunit configured to determine whether the current video frame is a reference frame according to the encoded bit stream of the current video frame; If it is a reference frame, performing a block effect evaluation operation and a fuzzy evaluation operation on the reference frame; the quality score subunit of the reference frame is configured to perform weighted averaging on the result of the block effect evaluation operation and the result of the fuzzy evaluation operation to obtain a reference frame Quality rating.
  • the reference frame is introduced for quality scoring, which improves the monitoring accuracy of the video stream quality. The quality scoring process of the reference frame is described in detail above, and will not be described here.
  • the quality scoring unit of the reference frame may further have the following characteristics: for the two most obvious spatial damages in the real-time video service: blockiness and blurring, respectively, using a less complex spatial domain quality evaluation method for evaluation, After the synthesis, the benchmark frame quality evaluation result is obtained. While ensuring the effect, the amount of calculation does not increase much.
  • the coding quality score of the current video frame is determined according to the video stream information.
  • the source coding quality scoring module 20 includes an encoding quality scoring unit, configured to be configured according to The video stream information is used to determine a coding quality score of the current video frame, where the coding quality scoring unit includes: a sub-unit for determining the number of bits per pixel, The method is configured to obtain, according to the number of bits and size information of the current video frame, the number of bits per pixel of the current video frame; the time and space complexity determining subunit, configured to adopt the encoding of the current video frame according to the number of bits per pixel and the quantization parameter.
  • the type determines the operation to obtain the time and space complexity; the coding quality score subunit is configured to obtain the coding quality score of the current video frame according to the time and space complexity and the quantization parameter. With this preferred embodiment, the monitoring accuracy of the video stream quality is improved.
  • the coding quality scoring process is described in detail above and will not be described here.
  • the present embodiment provides a preferred embodiment, that is, the source coding quality scoring module 20 includes a source coding quality scoring unit configured to be based on the quality of the reference frame. The score, the coding quality score of the current video frame, and the coding type of the current video frame, determine the source coding quality score; wherein, the source coding quality scoring unit includes:
  • a first scoring subunit configured to: when the encoding type of the current video frame is an intra prediction I frame, and the current video frame is a reference frame, the source encoding quality score is equal to the quality score of the reference frame; the second scoring subunit, When the encoding type of the current video frame is an intra prediction I frame, the current video frame is not a reference frame, and there is a quality score of the reference frame, according to the quality score of the reference frame, the encoding quality score of the current video frame, and The difference in coding quality between the current video frame and the reference frame, determining the source coding quality score;
  • a third scoring sub-unit configured to: when the encoding type of the current video frame is an intra prediction I frame, the current video frame is not a reference frame, and the quality score of the reference frame does not exist, the source coding quality score is equal to the current video frame. Coding quality score;
  • a fourth scoring sub-unit configured to: obtain a quality monitoring score of a previous video frame if the encoding type of the current video frame is a forward prediction P frame; and monitor a score of the current video frame according to a quality of the previous video frame The quality score, and the difference in coding quality between the current video frame and the previous video frame, determines the source coding quality score.
  • the source coding quality scoring process is described in detail above and will not be described here.
  • the source coding quality scoring unit may further have the following features: According to the frame coding type, it is determined whether an error propagation mechanism is introduced, and only the P frame is affected by the error propagation. After the quality of a frame is damaged, other frames in the same GoP after the frame will be affected. Although the I frame does not consider error propagation, in order to improve the evaluation accuracy, the source coding quality is evaluated based on the reference frame.
  • the network impairment scoring module 30 of the embodiment obtains the network impairment score according to the video stream information. In addition to considering the packet loss factor, the network impairment also considers the influence of the delay jitter factor on the video quality.
  • the present embodiment provides a preferred embodiment, that is, the network impairment scoring module 30 includes: a first setting unit configured to set a first time window length; wherein, an end point of the first time window is a current video frame, The starting point of a time window is a video frame whose time interval from the current video frame is not greater than the length of the first time window; the averaging unit is configured to determine the real-time average lost according to the real-time packet loss rate of each video frame in the first time window.
  • Packet rate determining a real-time average bit rate according to a real-time bit rate of each video frame in the first time window; determining a real-time average frame rate according to a real-time frame rate of each video frame in the first time window; And configured to determine a packet loss robustness factor according to a real-time average bit rate and a real-time average frame rate; determine a packet loss score according to a packet loss robustness factor and a real-time average packet loss rate; and a delay jitter scoring unit configured to The real-time average frame rate based on the current video frame arrival time, and the real-time average of the previous video frame arrival time Frame rate, determining the delay jitter score; wherein, the network impairment score includes the packet loss score and the delay jitter score.
  • the network damage scoring process is described in detail above and will not be described here.
  • the network impairment scoring module 30 may further have the following characteristics: the real-time average packet loss rate, the real-time average bit rate, and the real-time average frame rate used in determining are real-time packet loss rates within a period of time before the current video frame arrival time, The average of the real-time bit rate and real-time frame rate. This will make the evaluation results more stable.
  • the real-time packet loss rate, real-time bit rate, and real-time frame rate are also determined within a period of time before the current video frame arrival time.
  • the network impairment scoring module 30 can also have the following features: To determine the real-time packet loss rate, the real-time bit rate, and the real-time frame rate within a period of time before the current video frame arrival time, the time window concept can be used, and a storage time can be used. A "first in, first out" queue of all video frame information within the window.
  • the time window length is the length of time to determine the real-time packet loss rate, real-time bit rate, and real-time frame rate.
  • the current video frame information is added to the queue, and all the information of the video frames on the queue head that are different from the current video frame arrival time by more than the time window length are deleted.
  • a queue can also be established for the real-time packet loss rate and the real-time frame rate, configured to determine the real-time average packet loss rate and the real-time average frame rate.
  • the quality monitoring scoring module 40 includes: a dynamic scoring unit configured to score according to the source coding quality and packet loss. Scoring, delay jitter score, determining the dynamic score of the current video frame;
  • a second setting unit configured to set a second time window length; wherein, an ending point of the second time window is a current video frame, and a starting point of the second time window is a time interval from the current video frame not greater than a second time a video frame of a window length; a quality monitoring scoring unit configured to average the dynamic scores of the respective video frames in the second time window to obtain a quality monitoring score of the video stream.
  • the quality monitoring scoring module 40 is configured to combine the results of the source coding quality evaluation and the network impairment evaluation to determine the dynamic scoring result of the current video frame, and may use the time window to smooth the dynamic scoring of the current video frame, and the output is more stable. Dynamic rating results.
  • FIG. 8 is a schematic structural diagram of a video stream quality monitoring apparatus 2 according to an embodiment of the present invention.
  • the video stream quality monitoring apparatus includes a network data acquiring module 81, an information acquiring module 82, and a source encoding quality scoring module 83.
  • the network damage scoring module 84 and the quality monitoring scoring module 85, the network data obtaining module 81 is responsible for acquiring network data on the network side, and separating the video service network data to be evaluated according to the protocol and the port number.
  • the network data is extracted from the evaluation related information, including the video encoding information and the network packet information; the source encoding quality scoring module 83 and the network impairment scoring module 84 perform the itemized evaluation according to the related information, respectively.
  • the source coding quality scoring module 83 continues to be subdivided into three units, a quality score unit 831 of the reference frame, an encoding quality scoring unit 832, and a source coding quality scoring unit 833; the network impairment scoring module 84 performs packet loss assessment and delay jitter impairment. Evaluation; Quality Monitoring Scoring Module 85 combines the results of source coding quality assessment and network impairment assessment to give a final dynamic score.
  • the embodiment of the present invention acquires a real-time video stream of a specified service and a session on the network side, parses the real-time video stream, and scores the quality of the source code.
  • the video quality damage caused by the error propagation factor is evaluated, and the video quality damage caused by the network performance is evaluated.
  • the dynamic score of the real-time video stream quality is obtained based on the evaluation result.
  • the service quality dynamic score of the frame arrival time can be given for the video stream acquired by the network side.
  • the quality of service dynamic score of a video frame arrival time not only considers the current frame, but also considers other frames in the previous period of the current frame, so that the score is both dynamic and stable.
  • the network data acquisition module, the information acquisition module, the source coding quality scoring module, the network impairment scoring module, and the quality monitoring scoring module in the video stream quality monitoring apparatus proposed in the embodiment of the present invention, and each subunit in each module may
  • the base station controller may be implemented by a processor in a base station controller; the base station controller may be a radio network controller (RNC) or a base station controller (BSC); in practical applications, the processing
  • the device can be a central processing unit (CPU), a microprocessor (MPU, a Micro Processor Unit), a digital signal processor (DSP), or a Field Programmable Gate Array (FPGA). Wait.
  • the video stream quality monitoring apparatus proposed in the embodiment of the present invention can also be implemented by a specific logic circuit.
  • Embodiments of the present invention simultaneously consider the effects of source video coding and network performance on video quality.
  • the reference frame is innovatively used, and the quantization quality, time and space complexity, frame coding type, error propagation and other factors are considered in measuring the coding quality.
  • network performance is considered.
  • the quality damage caused by delay jitter is considered.
  • the determining process of the embodiment of the present invention is simple, and can also monitor the dynamic quality of real-time services, and can be well applied to online monitoring of real-time video service quality on the network side.
  • the embodiment of the present invention After acquiring the video stream information on the network side, the embodiment of the present invention obtains a source coding quality score and a network impairment score according to the video stream information, where the reference frame is used in determining the source coding quality score, and the network impairment score is used.
  • the video stream quality damage caused by the packet loss and the delay jitter is considered in the process of determining, and finally, the quality monitoring score of the video stream is determined according to the source coding quality score and the network impairment score;
  • the solution can solve the problem that the monitoring quality of the video stream quality monitoring method on the network side in the related technology is not high and the implementation is complicated, and the dynamic change of the video quality is accurately reflected, and the accuracy of the video stream quality monitoring is improved, and the complexity is improved. Not high, can be easily applied to the network side video quality real-time monitoring environment.

Landscapes

  • Engineering & Computer Science (AREA)
  • Signal Processing (AREA)
  • Computer Networks & Wireless Communication (AREA)
  • Multimedia (AREA)
  • Quality & Reliability (AREA)
  • Computer Security & Cryptography (AREA)
  • Health & Medical Sciences (AREA)
  • Biomedical Technology (AREA)
  • General Health & Medical Sciences (AREA)
  • Compression Or Coding Systems Of Tv Signals (AREA)

Abstract

本发明公开了一种视频流质量监测方法及装置。其中,该方法包括:获取网络侧的视频流信息;根据视频流信息,确定基准帧的质量评分,以及当前视频帧的编码质量评分;根据基准帧的质量评分、当前视频帧的编码质量评分、以及当前视频帧的编码类型,确定源编码质量评分;根据视频流信息,确定网络损伤评分;其中,网络损伤是由丟包以及时延抖动引起的视频流质量损伤;根据源编码质量评分和网络损伤评分,确定视频流的质量监测评分。通过本发明,解决了相关技术中网络侧视频流质量监测方法监测精度不高、实现较复杂的问题,准确反映出视频质量的动态变化,提高了视频流质量监测的准确度,复杂度不高,能很方便地应用于网络侧视频质量实时监测环境中。

Description

一种视频流质量监测方法及装置 技术领域
本发明涉及移动通讯领域, 特别是涉及一种视频流质量监测方法及装 置。 背景技术
近年来随着网络带宽的增加, 实时视频业务如流媒体业务、 视频电话 业务、 视频会议业务等发展速度非常迅速。 对这些视频业务的质量进行准 确评估, 既可以让用户了解到视频业务的服务水平, 又可以为服务提供商 提供服务水平提升的标尺, 对业务的良好运营有着重要作用。 尤其在业务 实际运行中, 实现业务的在线监测, 可以动态掌握业务运行的质量状态, 及时发现业务运行过程中的质量问题, 具有很强的实用价值。
视频流业务质量的在线监测可在服务侧、 网络侧和终端侧实现。 服务 侧业务质量监测主要监测服务侧提供的质量, 无法体现终端侧的用户体验 质量。 终端侧业务质量监测需要部署在用户终端上, 一般需征得用户同意 才能部署, 部署存在一定难度, 并且只能反映单个用户对业务的体验, 很 难反映网络群体用户的体验。 有些终端侧业务质量监测方法还会影响用户 对业务的使用, 从而影响用户的业务体验。 而网络侧业务质量监测可部署 在各网络节点, 无需用户配合, 也不会影响用户, 可对网络节点上通过的 多个业务会话同时进行监测, 从而反映该网络节点负责的网络的质量状况。 如果部署在离用户终端很近的网络节点上, 虽无法像终端侧方法那样直接 反映用户体验, 但也能很好反映业务运行状况。
但是, 在网络侧进行视频流业务质量在线监测存在一些挑战。 首先, 网络侧业务数据通过量较大, 视频业务尤其如此, 因此对数据处理速度的 要求很高, 相应的质量监测方法就不能太复杂; 其次, 在网络侧不存在视 频业务的原始数据, 只能通过解析视频业务流来进行质量评测, 这在媒体 质量评测中称为无参考方法, 无参考方法在质量评测领域一直是难点所在, 要么精度不高、 要么过于复杂; 再次, 网络侧业务质量监测的监测依据需 是当前时刻已到达的数据包, 且需要根据当前收到的数据包进行质量状态 的动态更新, 这就要求监测方法能够实时处理实时业务流数据。
现有的可应用于网络侧的无参考方法, 一般会考虑源视频编码和网络 性能两方面对视频质量的影响,但在具体涵盖因素方面各有不同。如 ITU-T G.1070针对视频电话业务的质量评测提出了一套较完善的无参考框架, 但 在针对视频质量进行评测时, 源视频编码只考虑了编码器类型、 比特率和 帧率三个因素, 网络性能只考虑了丟包因素; 移动图像质量度量(MPQM, Moving Picture Quality Metric )模型的源视频编码只是根据编码器类型的不 同给出了固定分数, 网络性能也只考虑了丟包, 导致视频流质量的监测精 度较低。 另外一些模型, 考虑的因素很多, 但其具体实现却很模糊。 如 T-V-model, 考虑了视频编码、 分辨率变化、 传输过程、 及显示过程带来的 损伤, 但是这些损伤与最终评分之间的关系过于简单, 且并未阐明损伤的 具体表示方法和影响因素。 其他一些模型, 也存在类似的问题。
针对相关技术中的网络侧的视频流质量监测方法监测精度不高、 实现 较复杂的问题, 目前尚未提出有效的解决方案。 发明内容
针对相关技术中网络侧的视频流质量监测方法监测精度不高、 实现较 复杂的问题, 本发明实施例提供了一种视频流质量监测方法及装置, 用以 解决上述技术问题。
本发明实施例提供了一种视频流质量监测方法, 其中, 该方法包括: 获取网络侧的视频流信息; 根据所述视频流信息, 确定基准帧的质量评分, 以及当前视频帧的编码质量评分; 根据所述基准帧的质量评分、 所述当前 视频帧的编码质量评分、 以及所述当前视频帧的编码类型, 确定源编码质 量评分; 根据所述视频流信息, 确定网络损伤评分; 其中, 所述网络损伤 是由丟包以及时延抖动引起的视频流质量损伤; 根据所述源编码质量评分 和所述网络损伤评分, 确定视频流的质量监测评分。
优选地, 获取网络侧的视频流信息包括: 对网络侧的视频业务网络数 据进行解封装和解复用, 得到视频流; 解析所述视频流, 定位当前视频帧 和上一视频帧在所述视频流中的起始点和结束点; 将当前视频帧的起始点 和结束点, 以及所述上一视频帧的起始点和结束点, 分别反向映射到所述 视频业务网络数据包, 分别得到所述当前视频帧和上一视频帧的所有网络 包的到达时刻; 才艮据该到达时刻分别确定当前视频帧到达时刻和上一视频 帧到达时刻; 根据所述上一视频帧的结束点以及所述当前视频帧的结束点, 确定所述当前视频帧的比特数; 确定所述当前视频帧的实时丟包率、 实时 比特率以及实时帧率; 解析所述当前视频帧的编码比特流, 确定所述当前 视频帧的编码类型、 尺寸信息以及量化参数。
优选地, 根据所述视频流信息, 确定基准帧的质量评分包括: 根据所 述当前视频帧的编码比特流判断所述当前视频帧是否为基准帧; 确定所述 当前视频帧为基准帧时, 对所述基准帧执行块效应评价操作以及模糊评价 操作; 对所述块效应评价操作的结果以及所述模糊评价操作的结果进行加 权平均, 得到所述基准帧的质量评分。
优选地, 根据所述视频流信息, 确定当前视频帧的编码质量评分包括: 根据所述当前视频帧的比特数和尺寸信息, 得到所述当前视频帧的每像素 比特数; 根据所述每像素比特数以及所述量化参数, 采取与所述当前视频 帧的编码类型相应的确定操作, 得到时间和空间复杂度; 根据所述时间和 空间复杂度以及所述量化参数, 得到所述当前视频帧的编码质量评分。
优选地, 根据所述基准帧的质量评分、 所述当前视频帧的编码质量评 分、 以及所述当前视频帧的编码类型, 确定源编码质量评分包括: 如果所 述当前视频帧的编码类型为帧内预测 I帧,且所述当前视频帧是基准帧, 则 所述源编码质量评分等于所述基准帧的质量评分; 如果所述当前视频帧的 编码类型为帧内预测 I帧, 所述当前视频帧不是基准帧,且存在所述基准帧 的质量评分, 则根据所述基准帧的质量评分、 所述当前视频帧的编码质量 评分, 以及所述当前视频帧与基准帧之间的编码质量差异, 确定所述源编 码质量评分; 如果所述当前视频帧的编码类型为帧内预测 I帧,所述当前视 频帧不是基准帧, 且不存在所述基准帧的质量评分, 则所述源编码质量评 分等于所述当前视频帧的编码质量评分; 如果所述当前视频帧的编码类型 为前向预测 P帧, 则获取上一视频帧的质量监测评分; 根据所述上一视频 帧的质量监测评分、 所述当前视频帧的编码质量评分、 以及所述当前视频 帧与所述上一视频帧之间的编码质量差异, 确定所述源编码质量评分。
优选地, 根据所述视频流信息, 确定网络损伤评分包括: 设定第一时 间窗口长度; 其中, 所述第一时间窗口的结束点为所述当前视频帧, 所述 第一时间窗口的开始点为与所述当前视频帧的时间间隔不大于所述第一时 间窗口长度的视频帧; 根据所述第一时间窗口内的各个视频帧的实时丟包 率, 确定实时平均丟包率; 根据所述第一时间窗口内的各个视频帧的实时 比特率, 确定实时平均比特率; 根据所述第一时间窗口内的各个视频帧的 实时帧率, 确定实时平均帧率; 根据所述实时平均比特率和所述实时平均 帧率, 确定丟包鲁棒性因子; 根据所述丟包鲁棒性因子以及所述实时平均 丟包率, 确定丟包损伤评分; 根据所述当前视频帧到达时刻的所述实时平 均帧率, 以及所述上一视频帧到达时刻的实时平均帧率, 确定时延抖动评 分; 其中, 所述网络损伤评分包括所述丟包损伤评分和所述时延抖动评分。
优选地, 根据所述源编码质量评分和所述网络损伤评分, 确定所述视 频流的质量监测评分包括: 根据所述源编码质量评分、 所述丟包损伤评分、 所述时延抖动评分, 确定得到当前视频帧的动态评分; 设定第二时间窗口 长度; 其中, 所述第二时间窗口的结束点为所述当前视频帧, 所述第二时 间窗口的开始点为与所述当前视频帧的时间间隔不大于所述第二时间窗口 长度的视频帧; 对所述第二时间窗口内的各个视频帧的动态评分取平均值, 得到所述视频流的质量监测评分。
根据本发明实施例的另一方面, 本发明实施例还提供了一种视频流质 量监测装置, 其中, 该装置包括: 信息获取模块, 配置为获取网络侧的视 频流信息; 源编码质量评分模块, 配置为根据视频流信息, 确定基准帧的 质量评分, 以及当前视频帧的编码质量评分; 根据基准帧的质量评分、 当 前视频帧的编码质量评分、 以及当前视频帧的编码类型, 确定源编码质量 评分; 网络损伤评分模块, 配置为根据视频流信息, 确定网络损伤评分; 其中, 网络损伤是由丟包以及时延抖动引起的视频流质量损伤; 质量监测 评分模块, 配置为根据源编码质量评分和网络损伤评分, 确定视频流的质 量监测评分。
优选地, 所述信息获取模块包括: 视频流获取子单元, 配置为对网络 侧的视频业务网络数据进行解封装和解复用, 得到视频流; 时刻确定子单 元, 配置为解析所述视频流, 定位当前视频帧和上一视频帧在所述视频流 中的起始点和结束点; 将当前视频帧的起始点和结束点, 以及所述上一视 频帧的起始点和结束点, 分别反向映射到所述视频业务网络数据包, 分别 得到所述当前视频帧和上一视频帧的所有网络包的到达时刻; 根据该到达 时刻分别确定当前视频帧到达时刻和上一视频帧到达时刻; 比特数确定子 单元, 配置为根据所述上一视频帧的结束点以及所述当前视频帧的结束点, 确定所述当前视频帧的比特数; 编码信息确定子单元, 配置为确定所述当 前视频帧的实时丟包率、 实时比特率以及实时帧率; 解析所述当前视频帧 的编码比特流, 确定所述当前视频帧的编码类型、 尺寸信息以及量化参数。 优选地, 所述源编码质量评分模块包括基准帧的质量评分单元, 配置 为根据所述视频流信息, 确定基准帧的质量评分; 其中, 所述基准帧的质 量评分单元包括: 基准帧判断子单元, 配置为根据所述当前视频帧的编码 比特流判断所述当前视频帧是否为基准帧; 评价子单元, 配置为如果是基 准帧, 则对所述基准帧执行块效应评价操作以及模糊评价操作; 基准帧的 质量评分子单元, 配置为对所述块效应评价操作的结果以及所述模糊评价 操作的结果进行加权平均, 得到所述基准帧的质量评分。
优选地, 所述源编码质量评分模块包括编码质量评分单元, 配置为根 据所述视频流信息, 确定当前视频帧的编码质量评分; 其中, 所述编码质 量评分单元包括: 每像素比特数确定子单元, 配置为根据所述当前视频帧 的比特数和尺寸信息, 得到所述当前视频帧的每像素比特数; 时间和空间 复杂度确定子单元, 配置为根据所述每像素比特数以及所述量化参数, 采 取与所述当前视频帧的编码类型相应的确定操作, 得到时间和空间复杂度; 编码质量评分子单元, 配置为根据所述时间和空间复杂度以及所述量化参 数, 得到所述当前视频帧的编码质量评分。
优选地, 所述源编码质量评分模块包括源编码质量评分单元, 配置为 根据所述基准帧的质量评分、 所述当前视频帧的编码质量评分、 以及所述 当前视频帧的编码类型, 确定源编码质量评分; 其中, 所述源编码质量评 分单元包括: 第一评分子单元, 配置为在所述当前视频帧的编码类型为帧 内预测 I帧,且所述当前视频帧是基准帧的情况下, 所述源编码质量评分等 于所述基准帧的质量评分; 第二评分子单元, 配置为在所述当前视频帧的 编码类型为帧内预测 I帧, 所述当前视频帧不是基准帧,且存在所述基准帧 的质量评分的情况下, 根据所述基准帧的质量评分、 所述当前视频帧的编 码质量评分, 以及所述当前视频帧与基准帧之间的编码质量差异, 确定所 述源编码质量评分; 第三评分子单元, 配置为在所述当前视频帧的编码类 型为帧内预测 I帧, 所述当前视频帧不是基准帧,且不存在所述基准帧的质 量评分的情况下, 所述源编码质量评分等于所述当前视频帧的编码质量评 分; 第四评分子单元, 配置为在所述当前视频帧的编码类型为前向预测 P 帧的情况下, 获取上一视频帧的质量监测评分; 根据所述上一视频帧的质 量监测评分、 所述当前视频帧的编码质量评分、 以及所述当前视频帧与所 述上一视频帧之间的编码质量差异, 确定所述源编码质量评分。
优选地, 所述网络损伤评分模块包括: 第一设定单元, 配置为设定第 一时间窗口长度; 其中, 所述第一时间窗口的结束点为所述当前视频帧, 所述第一时间窗口的开始点为与所述当前视频帧的时间间隔不大于所述第 一时间窗口长度的视频帧; 平均单元, 配置为根据所述第一时间窗口内的 各个视频帧的实时丟包率, 确定实时平均丟包率; 根据所述第一时间窗口 内的各个视频帧的实时比特率, 确定实时平均比特率; 根据所述第一时间 窗口内的各个视频帧的实时帧率, 确定实时平均帧率; 丟包损伤评分单元, 配置为根据所述实时平均比特率和所述实时平均帧率, 确定丟包鲁棒性因 分; 时延抖动评分单元, 配置为根据所述当前视频帧到达时刻的所述实时 平均帧率, 以及所述上一视频帧到达时刻的实时平均帧率, 确定时延抖动 评分; 其中, 所述网络损伤评分包括所述丟包损伤评分和所述时延抖动评 分。
优选地, 所述质量监测评分模块包括: 动态评分单元, 配置为根据所 述源编码质量评分、 所述丟包损伤评分、 所述时延抖动评分, 确定当前视 频帧的动态评分; 第二设定单元, 配置为设定第二时间窗口长度; 其中, 所述第二时间窗口的结束点为所述当前视频帧, 所述第二时间窗口的开始 点为与所述当前视频帧的时间间隔不大于所述第二时间窗口长度的视频 帧; 质量监测评分单元, 配置为对所述第二时间窗口内的各个视频帧的动 态评分取平均值, 得到所述视频流的质量监测评分。
通过本发明实施例, 在获取到网络侧的视频流信息后, 根据该视频流 信息获取源编码质量评分和网络损伤评分, 其中, 在源编码质量评分的确 定过程中使用基准帧, 在网络损伤评分的确定过程中考虑了丟包以及时延 抖动引起的视频流质量损伤, 最后根据所述源编码质量评分和所述网络损 伤评分, 确定视频流的质量监测评分, 解决了相关技术中网络侧的视频流 质量监测方法监测精度不高、 实现较复杂的问题, 较为准确地反映出视频 质量的动态变化, 提高了视频流质量监测的准确度, 且复杂度不高, 能够 很方便地应用于网络侧视频质量实时监测环境中。 附图说明
图 1是本发明实施例的视频流质量监测方法的流程图;
图 2是本发明实施例的网络侧的视频流信息的获取流程图;
图 3是本发明实施例的基准帧的质量评分流程图;
图 4是本发明实施例的编码质量评分流程图;
图 5是本发明实施例的源编码质量评分流程图;
图 6是本发明实施例的网络损伤评分流程图;
图 7是本发明实施例的视频流质量监测装置一的结构示意图; 图 8是本发明实施例的视频流质量监测装置二的结构示意图。 具体实施方式
为了解决相关技术中网络侧的视频流质量监测方法监测精度不高、 实 现较复杂的问题, 本发明实施例提供了一种视频流质量监测方法及装置, 以下结合附图以及实施例, 对本发明的技术方案进行进一步详细说明。 应 当理解, 此处所描述的具体实施例仅用以解释本发明的技术方案, 并不限 定本发明的保护范围。 本发明实施例提供了一种视频流质量监测方法, 该方法可以在网络侧 实现, 图 1是本发明实施例的视频流质量监测方法的流程图, 如图 1所示, 该方法包括:
步骤 S102, 获取网络侧的视频流信息;
这里, 所述视频流信息是为后续确定源编码质量评分和网络损伤评分 提供计算基础。 具体地: 所述视频流信息包括: 当前视频帧的编码比特流, 当前视频帧的比特数、 尺寸信息、 编码类型、 量化参数、 实时丟包率、 实 时比特率、 实时帧率以及当前视频帧到达时刻、 上一视频帧到达时刻; 步骤 S104, 根据所述视频流信息, 确定基准帧的质量评分, 以及当前 视频帧的编码质量评分; 根据所述基准帧的质量评分、 所述当前视频帧的 编码质量评分、 以及所述当前视频帧的编码类型, 确定源编码质量评分; 步骤 S106, 根据所述视频流信息, 确定网络损伤评分; 其中, 所述网 络损伤是由丟包以及时延抖动引起的视频流质量损伤;
步骤 S108, 根据所述源编码质量评分和所述网络损伤评分, 确定视频 流的质量监测评分。
通过上述方法, 在获取到网络侧的视频流信息后, 根据该视频流信息 获取源编码质量评分和网络损伤评分, 其中, 在源编码质量评分的确定过 程中使用基准帧, 在网络损伤评分的确定过程中考虑了丟包以及时延抖动 引起的视频流质量损伤, 最后根据所述源编码质量评分和所述网络损伤评 分, 确定视频流的质量监测评分, 解决了相关技术中网络侧的视频流质量 监测方法监测精度不高、 实现较复杂的问题, 较为准确地反映出视频质量 的动态变化, 提高了视频流质量监测的准确度, 且复杂度不高, 能够很方 便地应用于网络侧视频质量实时监测环境中。
对于上述步骤 S102中的视频流信息的获取操作, 本实施例提供了一种 优选实施方式, 即获取网络侧的视频流信息包括: 对网络侧的视频业务网络数据进行解封装和解复用, 得到视频流; 解 析所述视频流, 定位当前视频帧和上一视频帧在所述视频流中的起始点和 结束点; 将当前视频帧的起始点和结束点, 以及所述上一视频帧的起始点 和结束点, 分别反向映射到所述视频业务网络数据包, 分别得到所述当前 视频帧和上一视频帧的所有网络包的到达时刻; 居该到达时刻分别确定 当前视频帧到达时刻和上一视频帧到达时刻; 艮据上一视频帧的结束点以 及当前视频帧的结束点, 确定当前视频帧的比特数; 确定所述当前视频帧 的实时丟包率、 实时比特率以及实时帧率; 解析所述当前视频帧的编码比 特流, 确定所述当前视频帧的编码类型、 尺寸信息以及量化参数。
通过该优选实施方式, 为后续确定源编码质量评分和网络损伤评分提 供了计算基础, 提到了视频流质量的监测精度和准确度。
图 2是本发明实施例的网络侧的视频流信息的获取流程图, 如图 2所 示, 该流程包括以下步骤(步骤 S202-步骤 S210 ):
步骤 S202, 对视频业务网络数据进行解封装和解复用, 得到视频流; 这里, 所述对视频业务网络数据进行解封装和解复用, 得到视频流, 例如可以是,在实时传送协议 ( RTP ) /用户数据包协议( UDP, User Datagram Protocol ) /网络之间互连的协议(IP, Internet Protocol )协议框架下, 可通 过以太网帧解封装、 IP包解封装、 UDP包解封装、 RTP包解封装得到视频 业务流, 继而通过解复用 (如 H.223解复用)得到视频流。
步骤 S204, 解析视频流, 查找起始码, 从而定位当前视频帧的编码比 特流在视频流中的起始点和结束点;
步骤 S206, 将当前视频帧的起始点和结束点反向映射到视频业务网络 数据包, 可得到用于封装当前视频帧的所有网络数据包的序号以及到达时 刻;
步骤 S208, 由视频帧的所有网络包的到达时间, 可得到视频帧的到达 时刻、 最早到达时刻和最晚到达时刻;
这里, 所述视频帧的最早到达时刻为视频帧的第一个网络数据包的到 达时刻, 到达时刻和最晚到达时刻都为视频帧的最后一个网络数据包的到 达时刻。
步骤 S210, 两帧之间结束点的间隔即为当前视频帧所包含的比特数, 继续解析当前视频帧的编码比特流, 提取其他编码信息。
这里, 所述其他编码信息包括: 帧的编码类型 (I帧或 P帧); 帧的尺 寸 (宽度和高度); 帧的量化参数。
为得到所述信息只需解析少量信息即可, 运算量很小。 举例来说, 在 H.261、 H.263视频流中可解析每帧的帧头, 在 H.264视频流中可解析序列 参数集 ( SPS, Sequence Parameter Set )和图像参数集 ( PPS, Picture Parameter Set )。
其中,如果当前视频帧的帧编码类型为 I帧,且与上一基准帧出现间隔 大于 G个 GoP, 则继续判断其是否为无丟包 I帧, 如是则将其作为新的基 准帧, 并保存该帧的编码比特流, 如否, 则该帧不能作为新的基准帧。 判 断一帧是否存在丟包的方法是: 该帧对应的最后一个网络包的序号减去第 一个网络包的序号如等于该帧对应的网络包的个数, 则认为该帧是无丟包 的。 在实际应用中, 可根据业务性质、 评测精度要求、 及评测实时性要求 等确定 G的取值。
当然, 本实施例中的视频流信息的获取方法的实现不局限于所述流程, 只要能够准确获取到视频流信息即可。 例如可以通过与系统中的封装协议 的相配合, 来定位当前视频帧。 例如, 如使用 RTP协议直接封装视频帧, 且 RTP封装时采用了分片单元( FU, Fragmentation Unit )方法, 则可通过 FU中的标志位来定位当前帧。
对于所述步骤 S104中的根据视频流信息确定基准帧的质量评分, 本实 施例提供了一种优选实施方式, 即根据所述当前视频帧的编码比特流判断 所述当前视频帧是否为基准帧; 如果是基准帧, 则对该基准帧执行块效应 评价操作以及模糊评价操作; 对所述块效应评价操作的结果以及所述模糊 评价操作的结果进行加权平均, 得到基准帧的质量评分。 该优选实施方式 中, 引入基准帧进行质量评分, 提高了视频流质量的监测精度。
实时视频业务主要的空间域损伤为块效应和模糊, 图 3是本发明实施 例的基准帧的质量评分流程图, 如图 3 所示, 该流程包括以下步骤(步骤 S302-步骤 S310 ):
步骤 S302, 根据视频流信息中的编码比特流确定当前视频帧是否为基 准帧; 具体为, 判断当前视频帧的编码比特流是否存在, 如果存在, 则当 前视频帧为基准帧; 如果不存在, 则当前视频帧为非基准帧;
步骤 S304, 对当前视频帧进行视频解码;
步骤 S306, 对当前视频帧的重建图像进行块效应评价;
这里, 具体可以采用以下方法实现: 使用一种简单的块效应级别统计 法进行确定。 将重建图像的亮度分量分割成互不重叠的 8 X 8块。 在一个 8 X 8亮度块(即有 8行 8歹 ij )中, 若有大于或等于 7行的值是相同的, 就认 为该块是 "块效应比较严重"; 除边缘外, 每个 8 x 8 亮度块都有与之邻接 的 4个 8 X 8亮度块, 若 4条邻接边中有 3条或四条发生邻接的 8个灰度值 中多于 5个之间的差值大于阈值, 就认为该块的 "块效应严重"; 若 4条邻 接边中有 2条或 1条发生邻接的 8个灰度值有多于 5个之间的差值大于阈 值, 就认为该块的 "块效应一般"。 根据所述方法对所有 8 x 8块进行分类, 分别统计出当前视频帧中各种块效应程度块的所占比例, 最后求出块效应 对该图像的影响。
步骤 S308, 对当前帧的重建图像进行模糊评价;
这里, 具体可以采用一种简单的平均边缘宽度法进行确定: 使用 Sobel 竖直方向算子得到边缘图, 对于边缘图中的某个边缘点, 找到图像的对应 点附近(例如同一行) 的极值点, 极值点间的距离就是该边缘的宽度, 对 边缘图中所有边缘点求边缘宽度, 所有边缘宽度的和与边缘点数量的比值, 即为图像的模糊评价。
步骤 S310, 对块效应评价和模糊评价的结果加权平均, 即为基准帧的 质量评分 Qo。
当然, 所述块效应评价和模糊评价的方式并不仅限于所述实施例中列 举的方式, 只要能够准确获取基准帧的质量评分即可。 例如可根据业务性 质、 评测精度要求、 及评测实时性要求等选择基准帧的质量评分的具体实 施方法。 例如块效应可使用基于奇异值分解的块边界差异评测、 或基于局 部块差异与 HVS的块效应评价等方法进行评测, 模糊可使用边缘轮廓梯度 统计、 局部峰态统计等方法进行评测。
在本实施例中, 为了提高视频流的监测精度, 可以将一个或多个图像 组(GoP, Group of Picture ) 中的第一个无损 I帧作为基准帧, 从视频流中 提取基准帧的编码比特流, 将基准帧完全解码后进行空间域无参考监测。 用这一监测结果作为这一个或多个 GoP中的其他后续帧的评分基准。 在实 际应用中, 可根据业务性质、 监测精度要求、 以及监测实时性要求等选择 基准帧出现的间隔。
对于所述步骤 S104中的根据视频流信息, 确定当前视频帧的编码质量 评分, 本实施例提供了一种优选实施方式, 即根据当前视频帧的比特数和 尺寸信息, 得到当前视频帧的每像素比特数; 根据所述每像素比特数以及 所述量化参数, 采取与当前视频帧的编码类型相应的确定操作, 得到时间 和空间复杂度; 根据时间和空间复杂度以及所述量化参数, 得到当前视频 帧的编码质量评分。 通过该优选实施方式, 提高了视频流质量的监测精度。
在实际应用中, 量化参数越大, 量化步长就越大, 视频质量越低, 这 说明视频质量和量化参数之间存在反比的关系, 这种关系在这里用线性模 型来表示。 同时编码质量还跟时间和空间复杂度有关, 不同的时间和空间 复杂度会影响该线性模型的斜率。 时间和空间复杂度可由残差帧的性质来 表征, 这里采用该帧的量化参数以及编码后的每像素比特数来体现。 因此, 当前视频帧的编码质量可根据帧编码类型、 量化参数、 每像素比特数以及 时间和空间复杂度来确定。
基于此, 在所述实施方式中, 当前视频帧的编码质量除了使用传统的 量化参数(QP, Quantization Parameter )来衡量外, 还进一步地考虑了视频 帧的时间和空间复杂度。 视频帧的时间和空间复杂度具有以下特点: 视频 帧的时间和空间复杂度使用视频帧的量化参数 QP 以及帧每像素编码比特 数简单有效地表征, 并根据该帧所采用的不同编码类型而采取不同的确定 方法。 由于实时视频业务对时延要求很高,所以一般仅采用帧内预测 I帧和 前向预测 P帧, 而不采用引入时延的双向预测 B帧, 因此此处仅考虑两种 编码类型, I帧或 P帧。
图 4是本发明实施例的编码质量评分流程图, 如图 4所示, 该流程包 括以下步骤(步骤 S402-步骤 S406 ):
步骤 S402, 根据视频流信息中的当前视频帧的比特数 B和尺寸(宽度 W和高度 H ), 确定当前视频帧的每像素比特数 R;
这里, 所述每像素比特数 R的确定过程具体可以采用以下公式实现: π Β
R = 。
W x H
步骤 S404,根据视频流信息中的当前视频帧的编码类型和量化参数 QP, 确定当前帧的时间和空间复杂度 ;
这里, 若当前帧为 I 帧, 则时间和空间复杂度确定如下: 二^^ + ^^ 。 若当前帧为 P 帧, 则时间和空间复杂度确定如下: δ二 (a2 x R + b2 Qp。 其中, 所述公式中的 、 bi、 β2和 b2为常数值的模型参 数。
步骤 S406, 根据量化参数 QP, 以及时间和空间复杂度, 确定当前视频 帧的编码质量 Qc;
这里, 具体可以利用以下公式实现: ec = a3 (l - ( ) ) x ^ + 63。 其中, a4
所述公式中的 β3、 、 β4和 b4为常数值的模型参数。
在确定出基准帧的质量评分, 以及当前视频帧的编码质量评分之后, 便可确定源编码质量评分。 错误传播损伤评价用于衡量错误传播对当前视 频帧的损伤影响, 并确定源编码质量评分 QE。 如果当前帧为 I帧, 该帧的 质量完全由自身决定, 不受错误传播的影响。 如果当前帧为 P帧, 该帧的 质量并不完全由自身决定, 还受其参考帧质量的影响, 如果其参考帧的质 量产生了衰退, 这种衰退会传播到该帧。 实时视频业务对时延要求严格, 所以一般不使用多参考帧技术, 因此此处仅考虑上一帧对当前帧的错误传 播影响。
对于所述步骤 S104中的源编码质量评分的确定操作, 本实施例提供了 一种优选实施方式, 即根据基准帧的质量评分、 当前视频帧的编码质量评 分、 以及当前视频帧的编码类型, 确定源编码质量评分包括:
如果当前视频帧的编码类型为帧内预测 I帧, 且当前视频帧是基准帧, 则源编码质量评分等于基准帧的质量评分;
如果当前视频帧的编码类型为帧内预测 I帧, 当前视频帧不是基准帧, 且存在基准帧的质量评分, 则根据基准帧的质量评分、 当前视频帧的编码 质量评分, 以及当前视频帧与基准帧之间的编码质量差异, 确定源编码质 量评分;
如果当前视频帧的编码类型为帧内预测 I帧, 当前视频帧不是基准帧, 且不存在基准帧的质量评分, 则源编码质量评分等于当前视频帧的编码质 量评分; 如果当前视频帧的编码类型为前向预测 P帧, 则获取上一视频帧的质 量监测评分; 根据上一视频帧的质量监测评分、 当前视频帧的编码质量评 分、 以及当前视频帧与上一视频帧之间的编码质量差异, 确定源编码质量 评分。
图 5是本发明实施例的源编码质量评分流程图, 如图 5所示, 该流程 包括以下步骤(步骤 S502-步骤 S506 ):
步骤 S502, 若当前视频帧为 I帧, 且为基准帧 (判断当前视频帧的编 码比特流是否存在, 如存在则当前视频帧为基准帧), 则源编码质量评分
QE就是基准帧的质量评分 Qo, 即& = β。;
步骤 S504, 若当前视频帧为 I帧, 且不为基准帧, 且基准帧的质量评 分 Q。已存在, 则源编码质量评分 QE以基准帧的质量评分 。为基础, 并考 虑当前视频帧与基准帧之间的编码质量差异, 如下式: QE 。 其中,
Figure imgf000018_0001
Qc,o的确定同所述实施例中步骤 S406中 Qc的确定方式。
若当前视频帧为 I帧, 且不为基准帧, 但基准帧的质量评分 Qo (在评 测最开始,还未出现一个无损 I帧),则源编码质量评分 QE为当前视频帧的 编码质量评分 Qc, 即 = &。
步骤 S506, 若当前视频帧为 P帧, 考虑上一视频帧对当前视频帧的错 误传播, 即源编码质量评分 QE以上一视频帧的最终质量动态评分 Qprc为参 考, 并考虑当前视频帧与上一视频帧之间的编码质量差异, 如下式:
& = ¾re x-^ '其中, Qc,pre的确定同所述实施例中步骤 S406中 Qc的确定 方式。
对于所述实施例的步骤 S106中根据视频流信息确定网络损伤评分, 网 络损伤除了考虑丟包因素外, 还考虑了时延抖动因素对视频质量的影响。 此处, 没有考虑时延对视频质量的影响, 因为时延太大会导致丟包, 仅考 虑丟包即可。 视频业务流的整体时延表现为初始緩冲, 可通过一些附加手 段如广告等加以緩解。
在考虑丟包对视频质量的影响时, 可以使用实时平均丟包率替代常用 的实时丟包率, 使动态评分结果更为稳定。 实时丟包率为当前视频帧到达 时刻之前的一段时间内的丟包率。 实时平均丟包率为当前视频帧到达时刻 之前的一段时间内的丟包率平均值。
另外, 还可以确定丟包的鲁棒性因子以提高视频流质量监测的精度。 视频流实时平均比特率和实时平均帧率不同时对丟包的容忍程度不同。 实 时平均比特率或实时帧率越大, 对丟包的容忍程度越低。 实时比特率为当 前视频帧到达时刻之前的一段时间内通过网络节点的视频流速率。 实时平 均比特率为当前视频帧到达时刻之前的一段时间内通过网络节点的视频流 速率。 由于本实施例应用于网络侧, 帧率指一定时间内通过网络节点的视 频帧数量。 实时帧率为当前视频帧到达时刻之前的一段时间内通过网络节 点的视频帧数量。 实时平均帧率为当前视频帧到达时刻之前的一段时间内 通过网络节点的视频帧平均数量。
在考虑时延抖动对视频质量的影响时, 可以利用实时平均帧率的变化 表征时延抖动。 实时平均帧率的变化以前后视频帧到达时刻的实时平均帧 率的差值来表征。 由于某一视频帧相对于上一视频帧的时延抖动往往可在 终端侧通过终端緩冲器加以緩解, 但对于终端来说, 实时平均帧率的变化 会造成视频播放忽快忽慢的效果, 影响终端用户的主观体验。
基于所述多重考虑, 本实施例提供了一种优选实施方式, 即根据视频 流信息, 确定网络损伤评分包括: 设定第一时间窗口长度; 其中, 第一时 间窗口的结束点为当前视频帧, 第一时间窗口的开始点为与当前视频帧的 时间间隔不大于第一时间窗口长度的视频帧; 根据第一时间窗口内的各个 视频帧的实时丟包率, 确定实时平均丟包率; 根据第一时间窗口内的各个 视频帧的实时比特率, 确定实时平均比特率; 根据第一时间窗口内的各个 视频帧的实时帧率, 确定实时平均帧率; 根据实时平均比特率和实时平均 帧率, 确定丟包鲁棒性因子; 根据丟包鲁棒性因子以及实时平均丟包率, 确定丟包损伤评分; 根据当前视频帧到达时刻的实时平均帧率, 以及上一 视频帧到达时刻的实时平均帧率, 确定时延抖动评分; 其中, 网络损伤评 分包括丟包损伤评分和时延抖动评分。 通过该优选实施方式, 提高了视频 流质量监测的精度。
图 6是本发明实施例的网络损伤评分流程图, 如图 6所示, 该流程包 括以下步骤(步骤 S602至步骤 S620 ):
步骤 S602, 设定实时丟包率、 实时平均丟包率、 实时比特率、 实时平 均比特率、 实时帧率和实时平均帧率确定时使用的时间窗口;
这里, 可以设定时间窗口长度1 时间窗口的结束点为当前视频帧, 时 间窗口的开始点为第一个与当前视频帧的到达时刻间隔不大于 T的视频帧。
步骤 S604, 确定当前视频帧到达时刻的实时丟包率 PLRr;
这里, 确定 PLRr的方法为: 当前视频帧对应的最后一个网络数据包的 序号减去时间窗口内第一个视频帧对应的第一个网络数据包的序号的差 值, 再减去时间窗口内所有视频帧对应的所有网络数据包个数, 得到的差 值除以当前视频帧对应的最后一个网络数据包的序号与时间窗口内第一个 视频帧对应的第一个网络数据包的序号的差值。
步骤 S606, 确定当前视频帧到达时刻的实时平均丟包率 PLRa;
这里, 确定 PLRa的方法为: 确定时间窗口内所有视频帧到达时刻的实 时丟包率的平均。 使用实时平均丟包率替代实时丟包率, 可使评测结果更 为平滑。
步骤 S608, 确定当前视频帧到达时刻的实时比特率 BRr; 这里, 确定 BRr的方法为: 时间窗口内所有视频帧的比特数和, 除以 当前视频帧的最晚到达时刻与时间窗口内第一个视频帧的最早到达时刻的 差值。
步骤 S610, 确定当前视频帧到达时刻的实时平均比特率 BRa;
这里, 确定的 81^方法为: 确定时间窗口内所有视频帧到达时刻的实 时比特率的平均。 使用实时平均比特率替代实时比特率, 可使评测结果更 为平滑。
步骤 S612, 确定当前视频帧到达时刻的实时帧率 FRr;
这里, 确定的 FRr方法为: 时间窗口内的所有视频帧的个数, 除以当前 视频帧的最晚到达时刻与时间窗口内第一个视频帧的最早到达时刻的差 值。
步骤 S614, 确定当前视频帧到达时刻的实时平均帧率 FRa;
这里, 确定 FRa的方法为: 确定时间窗口内所有视频帧到达时间的实 时帧率的平均。 使用实时平均帧率替代实时帧率, 可使评测结果更为平滑。
步骤 S616, 确定当前视频帧到达时刻的丟包鲁棒性因子 Dplv;
这里, 根据实时平均比特率 BRa和实时平均帧率 FRa, 利用下式确定
BRa FRa
Dplv: Dplv = e ^ 。 其中, 该公式中的 a5和 b5为常数值的模型参数。
步骤 S618, 确定当前视频帧到达时刻的丟包损伤评分 PL; 这里, 确定 PL的方法为: PL二 e ' 。
步骤 S620, 确定当前视频帧到达时刻的时延抖动评分 DF。
这里,确定 DF的方法为: 平均帧率抖动为当前视频帧到达时刻的实时 平均帧率减去上一帧到达时刻的实时平均帧率。时延抖动评分 DF是平均帧 率抖动的线性模型, 如下式: DF = kx \ FRa - FRa p
在确定出源编码质量评分和网络损伤评分之后, 综合源编码质量评分、 网络损伤评分中丟包损伤评分和时延抖动评分的结果, 利用下式确定当前 视频帧的动态评分: Q = GE X PL _ DF。
为使评测结果更为稳定, 可采用时间窗口平滑技术。 设定时间窗口长 度 Tq, 时间窗口的结束点为当前视频帧, 时间窗口的开始点为第一个与当 前视频帧的时间间隔不大于 Tq的视频帧。 通过下式确定时间窗口内的所有 视频帧动态评分的平均值 作为最终的视频流的质量监测评分: = ρ。 基于所述视频流的质量监测评分的实施过程, 本实施例提供了一种优 选实施方式, 即 居源编码质量评分和网络损伤评分, 确定视频流的质量 监测评分包括: 根据源编码质量评分、 丟包损伤评分、 时延抖动评分, 确 定当前视频帧的动态评分; 设定第二时间窗口长度; 其中, 第二时间窗口 的结束点为当前视频帧, 第二时间窗口的开始点为与当前视频帧的时间间 隔不大于第二时间窗口长度的视频帧; 对第二时间窗口内的各个视频帧的 动态评分取平均值, 得到视频流的质量监测评分。 通过该优选实施方式, 提高了视频流质量监测结果的精度和稳定性。
对应于所述实施例介绍的视频流质量监测方法, 本实施例提供了一种 视频流质量监测装置, 该装置可以设置在网络侧, 用以实现所述实施例。 图 7是本发明实施例的视频流质量监测装置一的结构示意图, 如图 7所示, 该装置包括: 信息获取模块 10、 源编码质量评分模块 20、 网络损伤评分模 块 30和质量监测评分模块 40。 下面对该结构进行详细介绍。
信息获取模块 10, 配置为获取网络侧的视频流信息;
源编码质量评分模块 20, 配置为根据视频流信息, 确定基准帧的质量 评分, 以及当前视频帧的编码质量评分; 根据基准帧的质量评分、 当前视 频帧的编码质量评分、 以及当前视频帧的编码类型, 确定源编码质量评分; 网络损伤评分模块 30, 配置为根据视频流信息, 确定网络损伤评分; 其中, 网络损伤是由丟包以及时延抖动引起的视频流质量损伤; 质量监测评分模块 40, 配置为根据源编码质量评分和网络损伤评分, 确定视频流的质量监测评分。
通过所述方法, 在信息获取模块 10获取到网络侧的视频流信息后, 源 编码质量评分模块 20和网络损伤评分模块 30根据该视频流信息确定源编 码质量评分和网络损伤评分, 其中, 在源编码质量评分的确定过程中使用 基准帧, 在网络损伤评分的确定过程中考虑了丟包以及时延抖动引起的视 频流质量损伤, 最后质量监测评分模块 40根据所述源编码质量评分和所述 网络损伤评分, 确定视频流的质量监测评分, 解决了相关技术中网络侧的 视频流质量监测方法监测精度不高、 实现较复杂的问题, 较为准确地反映 出视频质量的动态变化, 提高了视频流质量监测的准确度, 且复杂度不高, 能够很方便地应用于网络侧视频质量实时监测环境中。
优选地, 所述装置还可以包括网络数据获取模块, 所述网络数据获取 模块配置为从网络上截取视频流。 对于所述信息获取模块 10的视频流信息 的获取操作,本实施例提供了一种优选实施方式,即信息获取模块 10包括: 视频流获取子单元, 网络数据获取模块对网络侧的视频业务网络数据进行 解封装和解复用, 得到视频流; 时刻确定子单元, 配置为解析视频流, 定 位当前视频帧和上一视频帧在视频流中的起始点和结束点; 将当前视频帧 的起始点和结束点, 以及上一视频帧的起始点和结束点, 分别反向映射到 视频业务网络数据包, 分别得到当前视频帧和上一视频帧的所有网络包的 到达时刻; 才艮据该到达时刻分别确定当前视频帧到达时刻和上一视频帧到 达时刻; 比特数确定子单元, 配置为根据上一视频帧的起始点以及当前视 频帧的起始点, 确定当前视频帧的比特数; 编码信息确定子单元, 配置为 确定当前视频帧的实时丟包率、 实时比特率以及实时帧率; 解析当前视频 帧的编码比特流, 确定当前视频帧的编码类型、 尺寸信息以及量化参数。 通过该优选实施方式, 为后续确定源编码质量评分和网络损伤评分提供了 计算基础, 提到了视频流质量的监测精度和准确度。 前面对视频流信息的 获取过程进行了详细介绍, 在此不再赘述。
优选地, 所述信息获取模块 10还可具有以下特点: 所提取的视频编码 信息包括, 基准帧的编码比特流; 帧的编码类型 (I帧或 P帧); 帧的尺寸 (宽度和高度); 帧的量化参数 QP; 每帧所含的比特数; 帧的到达时刻。 所 提取的网络包信息包括, 用于封装一个视频帧的所有网络包的序号、 到达 时间。
对于所述源编码质量评分模块 20根据视频流信息确定基准帧的质量评 分, 本实施例提供了一种优选实施方式, 即源编码质量评分模块 20包括基 准帧的质量评分单元, 配置为根据视频流信息, 确定基准帧的质量评分; 其中, 基准帧的质量评分单元包括: 基准帧判断子单元, 配置为根据当前 视频帧的编码比特流判断当前视频帧是否为基准帧; 评价子单元, 配置为 如果是基准帧, 则对基准帧执行块效应评价操作以及模糊评价操作; 基准 帧的质量评分子单元, 配置为对块效应评价操作的结果以及模糊评价操作 的结果进行加权平均, 得到基准帧的质量评分。 该优选实施方式中, 引入 基准帧进行质量评分, 提高了视频流质量的监测精度。 前面对基准帧的质 量评分过程进行了详细介绍, 在此不再赘述。
优选地, 所述基准帧的质量评分单元还可具有以下特点: 针对实时视 频业务中最明显的两种空间损伤: 块效应和模糊, 分别采用复杂度较低的 空间域质量评测方法进行评测, 综合后得到基准帧质量评价结果。 在保证 效果的同时, 计算量增加不多。
对于所述源编码质量评分模块 20根据视频流信息, 确定当前视频帧的 编码质量评分, 本实施例提供了一种优选实施方式, 即源编码质量评分模 块 20包括编码质量评分单元, 配置为根据视频流信息, 确定当前视频帧的 编码质量评分; 其中, 编码质量评分单元包括: 每像素比特数确定子单元, 配置为根据当前视频帧的比特数和尺寸信息, 得到当前视频帧的每像素比 特数; 时间和空间复杂度确定子单元, 配置为根据每像素比特数以及量化 参数, 采取与当前视频帧的编码类型相应的确定操作, 得到时间和空间复 杂度; 编码质量评分子单元, 配置为根据时间和空间复杂度以及量化参数, 得到当前视频帧的编码质量评分。 通过该优选实施方式, 提高了视频流质 量的监测精度。 前面对编码质量评分过程进行了详细介绍, 在此不再赘述。
对于所述源编码质量评分模块 20的源编码质量评分的确定操作, 本实 施例提供了一种优选实施方式, 即源编码质量评分模块 20包括源编码质量 评分单元, 配置为根据基准帧的质量评分、 当前视频帧的编码质量评分、 以及当前视频帧的编码类型, 确定源编码质量评分; 其中, 源编码质量评 分单元包括:
第一评分子单元, 配置为在当前视频帧的编码类型为帧内预测 I帧,且 当前视频帧是基准帧的情况下, 源编码质量评分等于基准帧的质量评分; 第二评分子单元, 配置为在当前视频帧的编码类型为帧内预测 I帧, 当 前视频帧不是基准帧, 且存在基准帧的质量评分的情况下, 根据基准帧的 质量评分、 当前视频帧的编码质量评分, 以及当前视频帧与基准帧之间的 编码质量差异, 确定源编码质量评分;
第三评分子单元, 配置为在当前视频帧的编码类型为帧内预测 I帧, 当 前视频帧不是基准帧, 且不存在基准帧的质量评分的情况下, 源编码质量 评分等于当前视频帧的编码质量评分;
第四评分子单元, 配置为在当前视频帧的编码类型为前向预测 P帧的 情况下, 获取上一视频帧的质量监测评分; 根据上一视频帧的质量监测评 分、 当前视频帧的编码质量评分、 以及当前视频帧与上一视频帧之间的编 码质量差异, 确定源编码质量评分。 前面对源编码质量评分过程进行了详 细介绍, 在此不再赘述。 优选地, 所述源编码质量评分单元还可具有以下特点: 根据帧编码类 型判断是否引入错误传播机制, 只有 P帧才会受到错误传播的影响。 而某 帧的质量受到损伤后, 该帧以后的、在同一个 GoP的其他帧都会受到影响。 I帧虽不考虑错误传播, 但为了提高评测精度, 以基准帧为基础进行源编码 质量评价。
对于所述实施例的网络损伤评分模块 30根据视频流信息获取网络损伤 评分, 网络损伤除了考虑丟包因素外, 还考虑了时延抖动因素对视频质量 的影响。本实施例提供了一种优选实施方式,即网络损伤评分模块 30包括: 第一设定单元, 配置为设定第一时间窗口长度; 其中, 第一时间窗口的结 束点为当前视频帧, 第一时间窗口的开始点为与当前视频帧的时间间隔不 大于第一时间窗口长度的视频帧; 平均单元, 配置为根据第一时间窗口内 的各个视频帧的实时丟包率, 确定实时平均丟包率; 根据第一时间窗口内 的各个视频帧的实时比特率, 确定实时平均比特率; 根据第一时间窗口内 的各个视频帧的实时帧率, 确定实时平均帧率; 丟包损伤评分单元, 配置 为根据实时平均比特率和实时平均帧率, 确定丟包鲁棒性因子; 根据丟包 鲁棒性因子以及实时平均丟包率, 确定丟包损伤评分; 时延抖动评分单元, 配置为根据当前视频帧到达时刻的实时平均帧率, 以及上一视频帧到达时 刻的实时平均帧率, 确定时延抖动评分; 其中, 网络损伤评分包括丟包损 伤评分和时延抖动评分。 通过该优选实施方式, 提高了视频流质量监测的 精度。 前面对网络损伤评分过程进行了详细介绍, 在此不再赘述。
优选地, 网络损伤评分模块 30还可具有以下特点: 确定时使用的实时 平均丟包率、 实时平均比特率和实时平均帧率是当前视频帧到达时刻之前 一段时间之内的实时丟包率、 实时比特率和实时帧率的平均值。 这样可使 评测结果更为稳定。 实时丟包率、 实时比特率和实时帧率也是在当前视频 帧到达时刻之前一段时间之内确定的。 优选地, 网络损伤评分模块 30还可具有以下特点: 为确定当前视频帧 到达时刻之前一段时间之内的实时丟包率、 实时比特率和实时帧率, 可使 用时间窗口概念, 以及一个存放时间窗口内所有视频帧信息的 "先进先出" 的队列。 时间窗口长度为确定实时丟包率、 实时比特率和实时帧率的时间 长度。 当前视频帧到达后, 将当前视频帧信息加入队列, 将队列头上所有 与当前视频帧到达时刻相差超过时间窗口长度的视频帧的信息全部删除。 也可以为实时丟包率和实时帧率建立队列, 配置为确定实时平均丟包率和 实时平均帧率。
基于质量监测评分模块 40的视频流质量监测评分的实施过程, 本实施 例提供了一种优选实施方式,即质量监测评分模块 40包括:动态评分单元, 配置为根据源编码质量评分、 丟包损伤评分、 时延抖动评分, 确定得到当 前视频帧的动态评分;
第二设定单元, 配置为设定第二时间窗口长度; 其中, 第二时间窗口 的结束点为当前视频帧, 第二时间窗口的开始点为与当前视频帧的时间间 隔不大于第二时间窗口长度的视频帧; 质量监测评分单元, 配置为对第二 时间窗口内的各个视频帧的动态评分取平均值, 得到视频流的质量监测评 分。 通过该优选实施方式, 提高了视频流质量监测结果的精度和稳定性。
优选地, 质量监测评分模块 40配置为综合源编码质量评价和网络损伤 评价的结果, 确定当前视频帧的动态评分结果, 可使用时间窗口对当前视 频帧的动态评分进行平滑, 输出更为稳定的动态评分结果。
图 8是本发明实施例的视频流质量监测装置二的结构示意图, 如图 8 所示, 所述视频流质量监测装置包括网络数据获取模块 81、 信息获取模块 82、 源编码质量评分模块 83、 网络损伤评分模块 84和质量监测评分模块 85, 网络数据获取模块 81负责在网络侧获取网络数据, 并根据协议和端口 号分离出待评测的视频业务网络数据。 继而由信息获取模块 82从视频业务 网络数据中提取评测相关信息, 包括视频编码信息及网络包信息; 源编码 质量评分模块 83和网络损伤评分模块 84分别根据上述相关信息进行分项 评测。 源编码质量评分模块 83继续细分为三个单元, 基准帧的质量评分单 元 831、 编码质量评分单元 832和源编码质量评分单元 833 ; 网络损伤评分 模块 84进行丟包损伤评价和时延抖动损伤评价; 质量监测评分模块 85综 合源编码质量评价和网络损伤评价的结果, 给出最终的动态评分。
从以上的描述中可以看出, 本发明实施例在网络侧获取指定业务和会 话的实时视频流, 对该实时视频流进行解析, 对其源编码的质量进行评分。 并评价由于错误传播因素造成的视频质量损伤, 以及评价由于网络性能造 成的视频质量损伤, 最后根据评价结果得出该实时视频流质量的动态评分。 本发明实施例对网络侧获取的视频流, 每解析得到一个视频帧的编码比特 流, 就可以给出该帧到达时刻的业务质量动态评分。 某视频帧到达时刻的 业务质量动态评分不仅考虑了当前帧的情况, 还考虑了当前帧之前一段时 间内其他帧的情况, 使评分既具有动态性, 也具备一定的稳定性。
本发明实施例中提出的视频流质量监测装置中的网络数据获取模块、 信息获取模块、 源编码质量评分模块、 网络损伤评分模块和质量监测评分 模块, 以及各模块中的各子单元, 都可以通过基站控制器中的处理器来实 现; 所述基站控制器可以是无线网络控制器 (RNC, Radio Network Controller )或基站控制器(BSC, Base Station Controller )等; 在实际应用 中, 所述处理器可以为中央处理器(CPU, Central Processing Unit ), 微处 理器(MPU, Micro Processor Unit )、 数字信号处理器(DSP, Digital Signal Processor )或现场可编程门阵列 ( FPGA, Field Programmable Gate Array ) 等。 另外, 本发明实施例中提出的视频流质量监测装置也可通过具体的逻 辑电路实现。
本发明实施例同时考虑源视频编码和网络性能两方面对视频质量的影 响, 在源视频编码评价上创新地使用了基准帧, 且衡量编码质量时考虑了 量化参数、 时间和空间复杂度、 帧编码类型、 错误传播等多个因素, 网络 性能方面除了考虑丟包外, 还考虑时延抖动带来的质量损伤, 提高了评测 精度。 另外, 本发明实施例的确定过程很简单, 也能监测实时业务的动态 质量, 可很好地应用于网络侧实时视频业务质量在线监测。
尽管为示例目的, 已经公开了本发明的优选实施例, 本领域的技术人 员将意识到各种改进、 增加和取代也是可能的, 因此, 本发明的范围应当 不限于所述实施例。 工业实用性
本发明实施例在获取到网络侧的视频流信息后, 根据该视频流信息获 取源编码质量评分和网络损伤评分, 其中, 在源编码质量评分的确定过程 中使用基准帧, 在网络损伤评分的确定过程中考虑了丟包以及时延抖动引 起的视频流质量损伤, 最后根据所述源编码质量评分和所述网络损伤评分, 确定视频流的质量监测评分; 如此, 本发明实施例提供的技术方案, 能够 解决相关技术中网络侧的视频流质量监测方法监测精度不高、 实现较复杂 的问题, 较为准确地反映出视频质量的动态变化, 提高了视频流质量监测 的准确度, 且复杂度不高, 能够很方便地应用于网络侧视频质量实时监测 环境中。

Claims

权利要求书
1、 一种视频流质量监测方法, 包括:
获取网络侧的视频流信息;
根据所述视频流信息确定基准帧的质量评分以及当前视频帧的编码质 量评分; 根据所述基准帧的质量评分、 所述当前视频帧的编码质量评分、 以及所述当前视频帧的编码类型确定源编码质量评分;
根据所述视频流信息确定获取网络损伤评分; 其中, 所述网络损伤是 由丟包和时延抖动引起的视频流质量损伤;
根据所述源编码质量评分和所述网络损伤评分, 确定所述视频流的质 量监测评分。
2、 如权利要求 1所述的方法, 其中, 所述获取网络侧的视频流信息包 括:
对网络侧的视频业务网络数据进行解封装和解复用, 得到视频流; 解析所述视频流, 定位当前视频帧和上一视频帧在所述视频流中的起 始点和结束点; 将当前视频帧的起始点和结束点, 以及所述上一视频帧的 起始点和结束点, 分别反向映射到所述视频业务网络数据包, 分别得到所 述当前视频帧和上一视频帧的所有网络包的到达时刻; 居该到达时刻分 别确定当前视频帧到达时刻和上一视频帧到达时刻;
根据所述上一视频帧的结束点以及所述当前视频帧的结束点, 确定所 述当前视频帧的比特数;
确定所述当前视频帧的实时丟包率、 实时比特率以及实时帧率; 解析 所述当前视频帧的编码比特流, 确定所述当前视频帧的编码类型、 尺寸信 息以及量化参数。
3、 如权利要求 2所述的方法, 其中, 所述根据所述视频流信息确定基 准帧的质量评分, 包括:
根据所述当前视频帧的编码比特流判断所述当前视频帧是否为基准 帧;
确定所述当前视频帧为基准帧时, 对所述基准帧执行块效应评价操作 以及模糊评价操作;
对所述块效应评价操作的结果以及所述模糊评价操作的结果进行加权 平均, 得到所述基准帧的质量评分。
4、 如权利要求 2所述的方法, 其中, 所述根据所述视频流信息确定当 前视频帧的编码质量评分, 包括:
根据所述当前视频帧的比特数和尺寸信息, 得到所述当前视频帧的每 像素比特数;
根据所述每像素比特数, 以及所述量化参数, 采取与所述当前视频帧 的编码类型相应的确定操作, 得到时间和空间复杂度;
根据所述时间和空间复杂度, 以及所述量化参数, 得到所述当前视频 帧的编码质量评分。
5、如权利要求 1所述的方法,其中, 所述根据所述基准帧的质量评分、 所述当前视频帧的编码质量评分、 以及所述当前视频帧的编码类型, 确定 源编码质量评分包括:
如果所述当前视频帧的编码类型为帧内预测 I帧,且所述当前视频帧是 基准帧, 则所述源编码质量评分等于所述基准帧的质量评分;
如果所述当前视频帧的编码类型为帧内预测 I帧,所述当前视频帧不是 基准帧, 且存在所述基准帧的质量评分, 则根据所述基准帧的质量评分、 所述当前视频帧的编码质量评分, 以及所述当前视频帧与基准帧之间的编 码质量差异, 确定所述源编码质量评分;
如果所述当前视频帧的编码类型为帧内预测 I帧,所述当前视频帧不是 基准帧, 且不存在所述基准帧的质量评分, 则所述源编码质量评分等于所 述当前视频帧的编码质量评分;
如果所述当前视频帧的编码类型为前向预测 P帧, 则获取上一视频帧 的质量监测评分; 根据所述上一视频帧的质量监测评分、 所述当前视频帧 的编码质量评分、 以及所述当前视频帧与所述上一视频帧之间的编码质量 差异, 确定所述源编码质量评分。
6、 如权利要求 2至 5中任一项所述的方法, 其中, 根据所述视频流信 息, 确定网络损伤评分包括:
设定第一时间窗口长度; 其中, 所述第一时间窗口的结束点为所述当 前视频帧, 所述第一时间窗口的开始点为与所述当前视频帧的时间间隔不 大于所述第一时间窗口长度的视频帧;
根据所述第一时间窗口内的各个视频帧的实时丟包率, 确定实时平均 丟包率; 根据所述第一时间窗口内的各个视频帧的实时比特率, 确定实时 平均比特率; 根据所述第一时间窗口内的各个视频帧的实时帧率, 确定实 时平均帧率;
根据所述实时平均比特率和所述实时平均帧率, 确定丟包鲁棒性因子; 根据所述当前视频帧到达时刻的所述实时平均帧率, 以及所述上一视 频帧到达时刻的实时平均帧率, 确定时延抖动评分; 其中, 所述网络损伤 评分包括所述丟包损伤评分和所述时延抖动评分。
7、 如权利要求 6所述的方法, 其中, 所述根据所述源编码质量评分和 所述网络损伤评分, 确定所述视频流的质量监测评分, 包括:
根据所述源编码质量评分、 所述丟包损伤评分、 所述时延抖动评分, 确定当前视频帧的动态评分;
设定第二时间窗口长度; 其中, 所述第二时间窗口的结束点为所述当 前视频帧, 所述第二时间窗口的开始点为与所述当前视频帧的时间间隔不 大于所述第二时间窗口长度的视频帧;
对所述第二时间窗口内的各个视频帧的动态评分取平均值, 得到所述 视频流的质量监测评分。
8、 一种视频流质量监测装置, 所述装置包括:
信息获取模块, 配置为获取网络侧的视频流信息;
源编码质量评分模块, 配置为根据所述视频流信息确定基准帧的质量 评分以及当前视频帧的编码质量评分; 根据所述基准帧的质量评分、 所述 当前视频帧的编码质量评分、 以及所述当前视频帧的编码类型, 确定源编 码质量评分;
网络损伤评分模块, 配置为根据所述视频流信息确定网络损伤评分; 其中, 所述网络损伤是由丟包和时延抖动引起的视频流质量损伤;
质量监测评分模块, 配置为根据所述源编码质量评分和所述网络损伤 评分, 确定所述视频流的质量监测评分。
9、 如权利要求 8所述的装置, 其中, 所述信息获取模块包括: 视频流获取子单元, 配置为对网络侧的视频业务网络数据进行解封装 和解复用, 得到视频流;
时刻确定子单元, 配置为解析所述视频流, 定位当前视频帧和上一视 频帧在所述视频流中的起始点和结束点; 将当前视频帧的起始点和结束点, 以及所述上一视频帧的起始点和结束点, 分别反向映射到所述视频业务网 络数据包, 分别得到所述当前视频帧和上一视频帧的所有网络包的到达时 刻; 才艮据该到达时刻分别确定当前视频帧到达时刻和上一视频帧到达时刻; 比特数确定子单元, 配置为根据所述上一视频帧的结束点以及所述当 前视频帧的结束点, 确定所述当前视频帧的比特数;
编码信息确定子单元, 配置为确定所述当前视频帧的实时丟包率、 实 时比特率以及实时帧率; 解析所述当前视频帧的编码比特流, 确定所述当 前视频帧的编码类型、 尺寸信息以及量化参数。
10、 如权利要求 9所述的装置, 其中, 所述源编码质量评分模块包括: 基准帧的质量评分单元, 配置为根据所述视频流信息, 确定基准帧的 质量评分; 其中, 所述基准帧的质量评分单元包括:
基准帧判断子单元, 配置为根据所述当前视频帧的编码比特流判断所 述当前视频帧是否为基准帧;
评价子单元, 配置为确定所述当前视频帧为基准帧时, 对所述基准帧 执行块效应评价操作以及模糊评价操作;
基准帧的质量评分子单元, 配置为对所述块效应评价操作的结果以及 所述模糊评价操作的结果进行加权平均, 得到所述基准帧的质量评分。
11、 如权利要求 9所述的装置, 其中, 所述源编码质量评分模块包括: 编码质量评分单元, 配置为根据所述视频流信息, 确定当前视频帧的 编码质量评分; 其中, 所述编码质量评分单元包括:
每像素比特数确定子单元, 配置为根据所述当前视频帧的比特数和尺 寸信息, 得到所述当前视频帧的每像素比特数;
时间和空间复杂度确定子单元, 配置为根据所述每像素比特数, 以及 所述量化参数, 采取与所述当前视频帧的编码类型相应的确定操作, 得到 时间和空间复杂度;
编码质量评分子单元, 配置为根据所述时间和空间复杂度, 以及所述 量化参数, 得到所述当前视频帧的编码质量评分。
12、 如权利要求 9所述的装置, 其中, 所述源编码质量评分模块包括: 源编码质量评分单元, 配置为根据所述基准帧的质量评分、 所述当前 视频帧的编码质量评分、 以及所述当前视频帧的编码类型, 确定源编码质 量评分; 其中, 所述源编码质量评分单元包括: 第一评分子单元, 配置为在所述当前视频帧的编码类型为帧内预测 I 帧, 且所述当前视频帧是基准帧的情况下, 所述源编码质量评分等于所述 基准帧的质量评分;
第二评分子单元, 配置为在所述当前视频帧的编码类型为帧内预测 I 帧, 所述当前视频帧不是基准帧, 且存在所述基准帧的质量评分的情况下, 根据所述基准帧的质量评分、 所述当前视频帧的编码质量评分, 以及所述 当前视频帧与基准帧之间的编码质量差异, 确定所述源编码质量评分; 第三评分子单元, 配置为在所述当前视频帧的编码类型为帧内预测 I 帧, 所述当前视频帧不是基准帧, 且不存在所述基准帧的质量评分的情况 下, 所述源编码质量评分等于所述当前视频帧的编码质量评分;
第四评分子单元, 配置为在所述当前视频帧的编码类型为前向预测 p 帧的情况下, 获取上一视频帧的质量监测评分; 根据所述上一视频帧的质 量监测评分、 所述当前视频帧的编码质量评分、 以及所述当前视频帧与所 述上一视频帧之间的编码质量差异, 确定所述源编码质量评分。
13、 如权利要求 9至 12中任一项所述的装置, 其中, 所述网络损伤评 分模块包括:
第一设定单元, 配置为设定第一时间窗口长度; 其中, 所述第一时间 窗口的结束点为所述当前视频帧, 所述第一时间窗口的开始点为与所述当 前视频帧的时间间隔不大于所述第一时间窗口长度的视频帧;
平均单元, 配置为根据所述第一时间窗口内的各个视频帧的实时丟包 率, 确定实时平均丟包率; 根据所述第一时间窗口内的各个视频帧的实时 比特率, 确定实时平均比特率; 根据所述第一时间窗口内的各个视频帧的 实时帧率, 确定实时平均帧率;
丟包损伤评分单元, 配置为根据所述实时平均比特率和所述实时平均 帧率, 确定丟包鲁棒性因子; 根据所述丟包鲁棒性因子以及所述实时平均 丟包率, 确定丟包损伤评分;
时延抖动评分单元, 配置为根据所述当前视频帧到达时刻的所述实时 平均帧率, 以及所述上一视频帧到达时刻的实时平均帧率, 确定时延抖动 评分; 其中, 所述网络损伤评分包括所述丟包损伤评分和所述时延抖动评 分。
14、 如权利要求 13所述的装置, 其中, 所述质量监测评分模块包括: 动态评分单元, 配置为 艮据所述源编码质量评分、 所述丟包损伤评分、 所述时延抖动评分, 确定当前视频帧的动态评分;
第二设定单元, 配置为设定第二时间窗口长度; 其中, 所述第二时间 窗口的结束点为所述当前视频帧, 所述第二时间窗口的开始点为与所述当 前视频帧的时间间隔不大于所述第二时间窗口长度的视频帧;
质量监测评分单元, 配置为对所述第二时间窗口内的各个视频帧的动 态评分取平均值, 得到所述视频流的质量监测评分。
PCT/CN2013/084795 2013-01-22 2013-09-30 一种视频流质量监测方法及装置 WO2014114109A1 (zh)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN201310022730.7A CN103945213B (zh) 2013-01-22 2013-01-22 一种视频流质量监测方法及装置
CN201310022730.7 2013-01-22

Publications (1)

Publication Number Publication Date
WO2014114109A1 true WO2014114109A1 (zh) 2014-07-31

Family

ID=51192660

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2013/084795 WO2014114109A1 (zh) 2013-01-22 2013-09-30 一种视频流质量监测方法及装置

Country Status (2)

Country Link
CN (1) CN103945213B (zh)
WO (1) WO2014114109A1 (zh)

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US10944993B2 (en) 2018-05-25 2021-03-09 Carrier Corporation Video device and network quality evaluation/diagnostic tool
US11902547B2 (en) 2021-07-15 2024-02-13 Google Llc Low-delay two-pass frame-level rate control using an adjusted complexity

Families Citing this family (16)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN104486092B (zh) * 2014-12-15 2018-04-06 北京国双科技有限公司 数据监控方法及装置
CN105721862B (zh) * 2016-02-01 2017-11-21 杭州电子科技大学 一种无人机航拍视频质量确定方法
CN106210926B (zh) * 2016-07-11 2018-12-04 天津大学 基于模糊控制的视频质量自适应控制方法
CN111432275B (zh) * 2016-10-08 2023-11-17 华为技术有限公司 评估视频质量的方法和设备
CN106851261A (zh) * 2017-02-23 2017-06-13 中国矿业大学 一种基于gop的视频质量评价方法
CN108696751B (zh) * 2017-04-11 2020-07-28 中国移动通信有限公司研究院 一种视频处理方法和装置
CN108965934B (zh) 2017-05-27 2021-01-29 华为技术有限公司 一种故障检测方法、监控设备及网络设备
CN107105226B (zh) * 2017-06-22 2019-01-01 迪爱斯信息技术股份有限公司 一种视频质量评价装置
CN107613315B (zh) * 2017-10-30 2020-05-12 Oppo广东移动通信有限公司 视频画质调整方法、装置、终端设备及存储介质
CN109413413B (zh) * 2018-09-26 2020-06-05 央视国际网络无锡有限公司 视频质量评价方法及装置
CN111314691B (zh) * 2018-12-11 2022-09-16 中国移动通信集团广东有限公司 一种视频通话质量评估方法和装置
CN114449253A (zh) * 2020-10-30 2022-05-06 中国移动通信有限公司研究院 视频质量评估方法、装置及网络设备
CN115412776B (zh) * 2021-05-28 2024-10-18 华为技术有限公司 一种近场场景下视频传输中的网络质量评估方法及设备
CN113905197A (zh) * 2021-09-29 2022-01-07 深圳市天视通视觉有限公司 一种网络录像机拷机装置及系统
CN114915846B (zh) * 2022-05-10 2024-06-21 中移(杭州)信息技术有限公司 数据处理方法、装置、设备及计算机可读存储介质
CN115150641A (zh) * 2022-06-28 2022-10-04 炫彩互动网络科技有限公司 一种云游戏视频编码参数动态调整方法及装置

Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN101227604A (zh) * 2007-01-18 2008-07-23 上海未来宽带技术及应用工程研究中心有限公司 一种通过网络损伤度检测网络视频质量的方法
CN101448175A (zh) * 2008-12-25 2009-06-03 华东师范大学 一种无参考的流视频质量评估方法
US20090153668A1 (en) * 2007-12-14 2009-06-18 Yong Gyoo Kim System and method for real-time video quality assessment based on transmission properties
CN101626506A (zh) * 2008-07-10 2010-01-13 华为技术有限公司 一种视频码流的质量评估方法、装置及系统
CN101635846A (zh) * 2008-07-21 2010-01-27 华为技术有限公司 一种视频质量评估方法、系统及装置
CN102158729A (zh) * 2011-05-05 2011-08-17 西北工业大学 无参考的视频序列编码质量客观评价方法

Patent Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN101227604A (zh) * 2007-01-18 2008-07-23 上海未来宽带技术及应用工程研究中心有限公司 一种通过网络损伤度检测网络视频质量的方法
US20090153668A1 (en) * 2007-12-14 2009-06-18 Yong Gyoo Kim System and method for real-time video quality assessment based on transmission properties
CN101626506A (zh) * 2008-07-10 2010-01-13 华为技术有限公司 一种视频码流的质量评估方法、装置及系统
CN101635846A (zh) * 2008-07-21 2010-01-27 华为技术有限公司 一种视频质量评估方法、系统及装置
CN101448175A (zh) * 2008-12-25 2009-06-03 华东师范大学 一种无参考的流视频质量评估方法
CN102158729A (zh) * 2011-05-05 2011-08-17 西北工业大学 无参考的视频序列编码质量客观评价方法

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US10944993B2 (en) 2018-05-25 2021-03-09 Carrier Corporation Video device and network quality evaluation/diagnostic tool
US11902547B2 (en) 2021-07-15 2024-02-13 Google Llc Low-delay two-pass frame-level rate control using an adjusted complexity

Also Published As

Publication number Publication date
CN103945213A (zh) 2014-07-23
CN103945213B (zh) 2016-12-28

Similar Documents

Publication Publication Date Title
WO2014114109A1 (zh) 一种视频流质量监测方法及装置
KR101414435B1 (ko) 비디오 스트림 품질 평가 방법 및 장치
JP5302342B2 (ja) ビデオ符号ストリームの品質を評価する方法、装置およびシステム
US9253063B2 (en) Bi-directional video compression for real-time video streams during transport in a packet switched network
JP5277177B2 (ja) ビデオテレフォニー品質評価方法及び装置
US9723329B2 (en) Method and system for determining a quality value of a video stream
US9497459B2 (en) Method and apparatus for detecting frame types
US9756356B2 (en) Application-assisted spatio-temporal error concealment for RTP video
US8249141B1 (en) Method and system for managing bandwidth based on intraframes
Adeyemi-Ejeye et al. Packet loss visibility across SD, HD, 3D, and UHD video streams
JP4914400B2 (ja) 品質推定方法、品質推定システム、ユーザ端末、品質管理端末およびプログラム
Khan et al. Impact of video content on video quality for video over wireless networks
JP4861371B2 (ja) 映像品質推定装置、方法、およびプログラム
JP2012244566A (ja) 伝送ビットレートに基づいて映像フレームレートを制御させる受信装置、システム、プログラム及び方法
CN101860471A (zh) 一种媒体流切换方法、装置和系统
Uhrina et al. The impact of bitrate and packet loss on the video quality of H. 264/AVC compression standard
Ascenso et al. Packet-header based no-reference quality metrics for H. 264/AVC video transmission
JP2011004354A (ja) 映像品質推定装置、映像品質推定方法および映像品質推定装置の制御プログラム
Kuo et al. QoS Improvement for H. 264 Video Transmission Based on Structural Similarity
Wenzel et al. Deduction of Video Quality Degradation in Transport Networks
Liu et al. Simulation and analysis of video stream transmission based on decodable frame ratio model in wireless network

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 13872976

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 13872976

Country of ref document: EP

Kind code of ref document: A1