WO2016173320A1 - 一种视频质量评估的方法和装置 - Google Patents

一种视频质量评估的方法和装置 Download PDF

Info

Publication number
WO2016173320A1
WO2016173320A1 PCT/CN2016/075489 CN2016075489W WO2016173320A1 WO 2016173320 A1 WO2016173320 A1 WO 2016173320A1 CN 2016075489 W CN2016075489 W CN 2016075489W WO 2016173320 A1 WO2016173320 A1 WO 2016173320A1
Authority
WO
WIPO (PCT)
Prior art keywords
video
frame
rate
packet loss
obtaining
Prior art date
Application number
PCT/CN2016/075489
Other languages
English (en)
French (fr)
Inventor
高山
杨付正
徐子强
Original Assignee
华为技术有限公司
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 华为技术有限公司 filed Critical 华为技术有限公司
Priority to EP16785751.5A priority Critical patent/EP3282698B1/en
Publication of WO2016173320A1 publication Critical patent/WO2016173320A1/zh
Priority to US15/796,035 priority patent/US10530675B2/en

Links

Images

Classifications

    • HELECTRICITY
    • H04ELECTRIC COMMUNICATION TECHNIQUE
    • H04LTRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L43/00Arrangements for monitoring or testing data switching networks
    • H04L43/08Monitoring or testing based on specific metrics, e.g. QoS, energy consumption or environmental parameters
    • H04L43/0823Errors, e.g. transmission errors
    • H04L43/0829Packet loss
    • HELECTRICITY
    • H04ELECTRIC COMMUNICATION TECHNIQUE
    • H04LTRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L65/00Network arrangements, protocols or services for supporting real-time applications in data packet communication
    • H04L65/60Network streaming of media packets
    • HELECTRICITY
    • H04ELECTRIC COMMUNICATION TECHNIQUE
    • H04LTRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L65/00Network arrangements, protocols or services for supporting real-time applications in data packet communication
    • H04L65/80Responding to QoS
    • 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

Definitions

  • the present invention relates to the field of digital signal processing technologies, and in particular, to a video data quality assessment method and apparatus.
  • the network planning layer planning model predicts the quality of the transmitted video according to the parameters of the network and the bearer video.
  • the model is usually applied to network planning to predict the video quality that can be obtained by transmitting video streams with different encoding parameters in the case of known network conditions. Therefore, the planning layer model can guide us to make reasonable network parameter planning and obtain better service quality. However, for a particular video stream, such models can also be used to quality monitor a particular video stream if its specific parameters are available.
  • the statistical data can be obtained through the de-packet header for quality evaluation, and the model calculates the encoding of the video by using parameters including the video content complexity factor (V_CCF) of the video.
  • V_CCF video content complexity factor
  • the model can also add the impact of packet loss to the model, so that the packet loss quality of the video is used as the video quality, and the video packet loss quality V_MOSP is:
  • V_MOSP V_MOSC-V_DP
  • V_DP is the distortion of the video after packet loss, and the packet header needs to be decompressed when calculating V_DP.
  • the header information analyzes the packet loss information, and the following packet loss parameters are obtained: the average video frame damage rate V_AIRF, the sequence damage rate V_IR, and the packet loss event frequency V_PLEF.
  • the model may further obtain the re-buffer quality V_MOSR of the video based on the encoding quality of the video and/or the packet loss quality of the video, thereby using the re-buffer quality as the video quality.
  • a specific video stream cannot be obtained. Therefore, specific statistical parameters related to a specific video stream cannot be obtained, such as but not limited to parameters such as V_CCF, V_PLEF, V_AIRF, and V_IR, so For applications where specific video streams are not available and video data quality assessment is required, video quality assessment requires new models.
  • Embodiments of the present invention provide a method and apparatus for video data quality evaluation, which are used to solve the problem that video data quality assessment cannot be performed in a specific video stream application.
  • a first aspect of the present invention provides a method for video data quality assessment, which may include:
  • the parameters for obtaining video quality of the video including content complexity of the video, average video frame impairment rate of the video, sequence of the video At least one of a damage rate and a frequency of packet loss events of the video;
  • the obtaining quality parameters for obtaining video quality of the video includes:
  • the calculating the content complexity by using the code rate comprises:
  • the content complexity is calculated according to the following formula:
  • V_CCF is the content complexity
  • V_BR is the code rate
  • a is a preset value
  • b is a preset value
  • c is a preset value
  • Threshold is a preset value
  • Long ride The value obtained by the wide calculation of the video.
  • the calculating by using the packet loss ratio, and the average number of packets included in each frame,
  • the average video frame impairment rate includes:
  • the average video frame damage rate is calculated according to the following formula:
  • V_AIRF is the average video frame impairment rate
  • V_PktpF is the average number of packets per frame
  • V_LossRate is the packet loss rate
  • the utilizing the length of the group of pictures and the frame loss Rate calculate the sequence damage rate including:
  • the damage rate of the sequence is calculated according to the following formula:
  • V_IR is the damage rate of the sequence
  • V_LossRateFrame is the frame loss rate
  • Goplength is the length of the picture group.
  • the utilizing the sequence Calculating the frequency of the packet loss event and the packet loss rate include:
  • V_PLEF is the frequency of the packet loss event
  • TotalPktNum is the total number of packets of the sequence
  • V_LossRate is the packet loss rate
  • V_Burst is the burst packet length of the video
  • V_PLEF is the frequency of the packet loss event
  • TotalPktNum is the total number of packets of the sequence
  • V_LossRate is the packet loss rate
  • V_Burst is the burst packet length of the video
  • V_PktpF is the packet included in the average frame per frame. number.
  • the average number of packets included in each frame is a total number of packets of the sequence of the video and a total number of frames of the sequence of the video The ratio.
  • the frame loss rate includes:
  • V_LossRateFrame 1-(1-V_LossRate) V_PktpF
  • V_PktpF is the average number of packets included in each frame
  • V_LossRate is the packet loss rate
  • V_LossRateFrame is the frame loss rate of the video.
  • a second aspect of the present invention provides an apparatus for video data quality assessment, which may include:
  • An obtaining module configured to acquire a parameter for obtaining a video quality of the video, where the parameter for obtaining a video quality of the video includes a content complexity of the video, an average video frame damage rate of the video, At least one of a sequence damage rate of the video and a packet loss event frequency of the video;
  • a processing module configured to obtain, according to the parameter used to obtain the video quality of the video, the quality of the video
  • the acquiring module acquires parameters for obtaining video quality of the video, including:
  • the obtaining, by the acquiring module, the calculating the content complexity by using the code rate comprises:
  • the obtaining module calculates the content complexity according to the following formula:
  • V_CCF is the content complexity
  • V_BR is the code rate
  • a is a preset value
  • b is a preset value
  • c is a preset value
  • Threshold is a preset value
  • Long multiply the value obtained by the width of the video.
  • the acquiring module by using the packet loss ratio, and the average per frame included The number of packets, the average video frame damage rate calculated includes:
  • the obtaining module calculates the average video frame damage rate according to the following formula:
  • V_AIRF is the average video frame impairment rate
  • V_PktpF is the average number of packets per frame
  • V_LossRate is the packet loss rate
  • the acquiring module is configured to utilize a length of the group of pictures and The frame loss rate, and calculating the sequence damage rate includes:
  • the obtaining module calculates the damage rate of the sequence according to the following formula:
  • V_IR is the damage rate of the sequence
  • V_LossRateFrame is the frame loss rate
  • Goplength is the length of the picture group.
  • Calculating the frequency of the packet loss event by using the total number of packets of the sequence and the packet loss rate includes:
  • the obtaining module calculates the frequency of the packet loss event according to the following formula:
  • V_PLEF is the frequency of the packet loss event
  • TotalPktNum is the total number of packets of the sequence
  • V_LossRate is the packet loss rate
  • V_Burst is the burst packet length of the video
  • V_PLEF is the frequency of the packet loss event
  • TotalPktNum is the total number of packets of the sequence
  • V_LossRate is the packet loss rate
  • V_Burst is the burst packet length of the video
  • V_PktpF is the packet included in the average frame per frame. number.
  • the average per frame includes The number of packets is the ratio of the total number of packets of the sequence of video to the total number of frames of the sequence of video.
  • Obtaining the frame loss rate of the video includes:
  • the obtaining module obtains the packet loss rate and the average number of packets included in each frame, and calculates the frame loss according to the following formula:
  • V_LossRateFrame 1-(1-V_LossRate) V_PktpF
  • V_PktpF is the average number of packets included in each frame
  • V_LossRate is the packet loss rate
  • V_LossRateFrame is the frame loss rate of the video.
  • the present invention implements calculation of statistical parameters related to a specific video stream based on the fitting model and the probability model, so that an application where the specific video stream cannot be obtained can perform video data quality evaluation.
  • 1 is a method for evaluating video data quality according to an embodiment of the present invention
  • FIG. 3 is a device for evaluating video data quality according to an embodiment of the present invention
  • FIG. 4 is another apparatus for evaluating video data quality according to an embodiment of the present invention.
  • FIG. 1 is a method for evaluating video data quality according to an embodiment of the present invention.
  • the parameters in the method may be, but are not limited to, parameters in a model of P.1202.1.
  • the method steps are as follows:
  • the encoding quality of the video may be obtained according to the following formula:
  • V_MOSC MOS_MAX-V_DC
  • MOS_MAX is the maximum average opinion value of the video
  • V_DC is the distortion introduced by the video due to the encoding
  • V_MOSC is the encoding quality of the video.
  • the encoding quality of the video may also be obtained according to the following formula:
  • the V_CCF is the content complexity of the video
  • the videoFrameRate is the frame rate of the video
  • v1 is a preset coefficient value.
  • the coding quality of the video may be obtained according to the following formula:
  • V_MOSC MOS_MAX-V_DC
  • the encoding quality of the video can be obtained according to the following formula:
  • MOS_MIN is the minimum average opinion value of the video
  • V_NBR is the normalized code rate
  • v3, v4, v5, and v6 are preset coefficient values.
  • the two parameters may be calculated according to the code rate V_BR and the average code number V_ABIF of the I frame, and the calculation formula is as follows:
  • the encoding quality V_MOSC of the video may directly be used as the video.
  • the video quality V_MOS, or the average opinion value of the video, is used for the evaluation of video data quality.
  • S104 Calculate a packet loss quality of the video by using an encoding quality of the video and a distortion introduced by the packet due to packet loss.
  • the packet loss quality of the video can be obtained by using the following formula:
  • V_MOSP V_MOSC-V_DP
  • V_DP is the distortion introduced by the packet due to packet loss
  • V_MOSP is the packet loss quality of the video
  • V_AIRF is the average video frame damage rate of the video
  • V_IR is the sequence damage rate of the video
  • V_PLEF is the packet loss event frequency of the video
  • v7, v8, v9, v10, v11, v12 are preset. Coefficient value.
  • the distortion introduced by the packet loss can also be obtained by the following formula:
  • the distortion introduced by the packet loss may be obtained by using the following formula:
  • the average video frame impairment rate V_AIRF for the video may be obtained according to the following formula:
  • V_NDF is the number of damaged frames of the video
  • V_IRpFi is the average video frame damage rate in the i-th interval in the preset video frame set.
  • sequence damage rate V_IR for the video may be obtained according to the following formula:
  • V_TNF is the total number of video frames in the video stream of the video.
  • the frequency of the packet loss event V_PLEF can be analyzed according to a specific packet. When there is a packet loss in a frame, the frequency value of the packet loss event can be increased by one, and finally the frequency of the packet loss event is counted.
  • the packet loss quality V_MOSP of the video may be directly used as the video quality V_MOS of the video, or the average opinion value of the video is used for the evaluation of the video data quality.
  • S106 Calculate a re-buffer quality of the video by using an encoding quality of the video and/or a packet loss quality of the video.
  • the re-buffering model uses the number of re-buffering times, the re-buffering duration, and the re-buffering interval at the time of multiple re-buffering to measure the video quality affected by the buffering, and the re-buffer quality V_MOSR of the video can be calculated according to the following formula:
  • V_MOSR Video_Quality-V_DR
  • Video_Quality indicates the current video quality
  • NRE is the number of re-buffer events
  • ARL is the average length of the re-buffer
  • MREEF is the multi-buffer event influence factor
  • v13, v14, v15, v16, v17, v18 are preset coefficients. value.
  • the current video quality needs to be determined according to the channel condition, and when there is a packet loss situation and a re-buffering situation, the value of Video_Quality is the packet loss quality of the video; otherwise, the Video_Quality The value is the encoding quality of the video.
  • the re-buffer quality V_MOSR of the video may be used directly as the video quality V_MOS of the video, or as an average opinion value of the video for evaluation of video data quality.
  • the invention divides the model into a coding quality model, a lost packet quality model and a re-buffer quality model.
  • the way can effectively reflect the impact of coding quality on video quality.
  • the embodiment of the present invention provides another method for video data quality assessment, as shown in FIG. 2, which may be considered to adjust the model of P.1202.1 according to the planning layer.
  • the parameters that can be used to find a simplified model that is, the parameters in P.1202.1 that are not available, such as the content complexity of the video, the average video frame damage rate of the video, the sequence damage rate of the video, and the frequency of the packet loss event of the video.
  • the parameters in P.1201.1 are derived and mapped to achieve quality assessment.
  • the method steps are as follows:
  • a parameter for obtaining a video quality of the video includes a content complexity of the video, an average video frame impairment rate of the video, and the video. At least one of a sequence damage rate, a frequency of packet loss events of the video.
  • the process of acquiring the content complexity of the video may be specifically: obtaining a code rate of the video, and calculating the content complexity by using the code rate.
  • the content complexity can be calculated according to the following formula:
  • V_CCF is the content complexity
  • V_BR is the code rate
  • a is a preset value
  • b is a preset value
  • c is a preset value
  • Threshold is a preset value
  • the video is utilized. Long multiply the value obtained by the width of the video.
  • the values of a, b, c, and Threshold can be obtained through fitting training, or can be obtained according to the query in Table 1.
  • the value of the Threshold can also be calculated by using the length of the video of the video, and the value of the Threshold can also be calculated by using the resolution of the video.
  • the calculation formula can be specifically as follows:
  • Width is the width of the video
  • Height is the length of the video
  • Resolution is the resolution of the video
  • d is a preset coefficient value
  • the value of d can be obtained by fitting training, or according to Table 2 The query is obtained.
  • the process of obtaining an average frame frame impairment rate of the video may be specifically: obtaining a packet loss rate of the video and an average number of packets included in each frame of the video, using the packet loss rate and the The average video frame damage rate is calculated by averaging the number of packets per frame.
  • the average video frame damage rate may be calculated according to the following formula:
  • V_AIRF is the average video frame impairment rate
  • V_PktpF is the average number of packets per frame
  • V_LossRate is the packet loss rate
  • the process of obtaining the sequence damage rate of the video may be specifically: obtaining a length of the picture group of the video, obtaining a frame loss rate of the video, using a length of the picture group, and the frame loss rate.
  • the damage rate of the sequence can be calculated according to the following formula:
  • V_IR is the damage rate of the sequence
  • V_LossRateFrame is the frame loss rate
  • Goplength is the length of the picture group.
  • the process of obtaining the frequency of the packet loss event of the video may be specifically: obtaining a total number of packets of the sequence of the video and a packet loss rate of the video, using the total number of packets of the sequence, and the lost The packet rate calculates the frequency of the packet loss event.
  • the frequency of the packet loss event may be calculated according to the following formula:
  • V_PLEF is the frequency of the packet loss event
  • TotalPktNum is the total number of packets of the sequence
  • V_LossRate is the packet loss rate
  • V_Burst is the burst packet length of the video
  • V_PLEF is the frequency of the packet loss event
  • TotalPktNum is the total number of packets of the sequence
  • V_LossRate is the packet loss rate
  • V_Burst is the burst packet length of the video
  • V_PktpF is the packet included in the average frame per frame. number.
  • the ratio of the burst packet length of the video to the average number of packets included in each frame is set to V_ratio, and when V_ratio is less than 1, the following formula may be calculated.
  • the frequency of the packet loss event is set to V_ratio, and when V_ratio is less than 1, the following formula may be calculated.
  • the frequency of the packet loss event can be calculated according to the following formula:
  • the average number of packets V_PktpF included in each frame may be a ratio of a total number of packets TotalPktNum of the sequence of the video to a total number of frames TotalFrameNum of the sequence of the video.
  • the value of the total number of frames of the sequence of the video TotalFrameNum may be a value of the product of the frame rate videoFrameRate of the video and the duration MeasureTime of the video.
  • the process of obtaining the frame loss rate of the video may be specifically: obtaining the packet loss rate and the average number of packets included in each frame, and calculating the frame loss according to the following formula:
  • V_LossRateFrame 1-(1-V_LossRate) V_PktpF
  • V_PktpF is the average number of packets included in each frame
  • V_LossRate is the packet loss rate
  • V_LossRateFrame is the frame loss rate of the video.
  • the invention provides an effective method for objective measurement of video quality, which overcomes a series of problems such as high complexity, high cost, and possible harm to the human body.
  • the invention can effectively reflect the influence of the coding quality on the video quality, and fully utilizes the channel parameters, converts the channel parameters or the data available in the network planning layer into the code stream parameters necessary for evaluating the video quality, and uses the probability model. To calculate the impact of packet loss on the video, so that applications that cannot obtain a specific video stream can perform video data quality assessment and are more realistic. Time.
  • the invention fully considers the psychological feelings when people watch video, and the evaluation method is simple and reliable, and can actively guide the service provider to provide better video quality service.
  • an embodiment of the present invention further provides a video data quality evaluation apparatus 300, which may include:
  • the obtaining module 302 is configured to obtain a parameter for obtaining a video quality of the video, where the parameter for obtaining a video quality of the video includes a content complexity of the video, and an average video frame damage rate of the video. At least one of a sequence impairment rate of the video and a packet loss event frequency of the video.
  • the obtaining, by the obtaining module 302, the content complexity of the video may be: obtaining a code rate of the video, and calculating the content complexity by using the code rate.
  • the content complexity can be calculated according to the following formula:
  • V_CCF is the content complexity
  • V_BR is the code rate
  • a is a preset value
  • b is a preset value
  • c is a preset value
  • Threshold is a preset value
  • Long multiply the value obtained by the width of the video.
  • the obtaining, by the obtaining module 302, the average video frame impairment rate may be: obtaining a packet loss rate of the video, and an average number of packets included in each frame of the video, using the packet loss rate and the The average video frame damage rate is calculated by averaging the number of packets per frame.
  • the average video frame damage rate may be calculated according to the following formula:
  • V_AIRF is the average video frame impairment rate
  • V_PktpF is the average number of packets per frame
  • V_LossRate is the packet loss rate
  • the obtaining, by the obtaining module 302, the sequence damage rate may be: obtaining a length of the picture group of the video, obtaining a frame loss rate of the video, using a length of the picture group, and the frame loss rate. Calculate the sequence damage rate.
  • the damage rate of the sequence can be calculated according to the following formula:
  • V_IR is the damage rate of the sequence
  • V_LossRateFrame is the frame loss rate
  • Goplength is the length of the picture group.
  • the obtaining, by the obtaining module 302, the frequency of the packet loss event may be: obtaining a total number of packets of the sequence of the video and a packet loss rate of the video, using the total number of packets in the sequence, and the lost The packet rate calculates the frequency of the packet loss event.
  • the frequency of the packet loss event may be calculated according to the following formula:
  • V_PLEF is the frequency of the packet loss event
  • TotalPktNum is the total number of packets of the sequence
  • V_LossRate is the packet loss rate
  • V_Burst is the burst packet length of the video
  • V_PLEF is the frequency of the packet loss event
  • TotalPktNum is the total number of packets of the sequence
  • V_LossRate is the packet loss rate
  • V_Burst is the burst packet length of the video
  • V_PktpF is the packet included in the average frame per frame. number.
  • the average number of packets V_PktpF included in each frame may be a ratio of a total number of packets TotalPktNum of the sequence of the video to a total number of frames TotalFrameNum of the sequence of the video.
  • the obtaining, by the obtaining module 302, the frame loss rate of the video may be: obtaining the packet loss rate and the average number of packets included in each frame, and calculating the frame loss according to the following formula:
  • V_LossRateFrame 1-(1-V_LossRate) V_PktpF
  • V_PktpF is the average number of packets included in each frame
  • V_LossRate is the packet loss rate
  • V_LossRateFrame is the frame loss rate of the video.
  • the processing module 304 is configured to obtain the quality of the video according to the obtained parameter for obtaining the video quality of the video obtained by the obtaining module 302.
  • the video data quality evaluation device 300 can be any device that needs to output and play video, such as a notebook computer, a tablet computer, a personal computer, a mobile phone, and the like.
  • the invention provides an effective video quality objective measuring device, which overcomes a series of problems such as subjective evaluation with high complexity, high cost and possible damage to the human body.
  • the video data quality evaluation apparatus 300 provided by the embodiment of the present invention can effectively reflect the influence of the coding quality on the video quality, and fully utilizes the channel parameters, and converts the channel parameters or the data available in the network planning layer into the evaluation video quality.
  • the necessary code stream parameters and the probability model are used to calculate the impact of the packet loss on the video, so that the application where the specific video stream cannot be obtained can perform the video data quality evaluation and is more realistic.
  • FIG. 4 is a schematic diagram of a video data quality evaluation apparatus 400 according to an embodiment of the present invention.
  • the video data quality evaluation apparatus 400 may include at least one bus 401, at least one processor 402 connected to the bus 401, and connected to the bus 401.
  • the processor 402 calls a code or an instruction stored in the memory 403 via the bus 401 for acquiring a parameter for obtaining a video quality of the video, where the parameter for obtaining a video quality of the video includes the At least one of a content complexity of the video, an average video frame impairment rate of the video, a sequence impairment rate of the video, and a packet loss event frequency of the video.
  • the quality of the video is obtained according to the parameters for obtaining the video quality of the video.
  • the processor 402 may be configured to: obtain a code rate of the video, and calculate the content complexity by using the code rate.
  • the content complexity can be calculated according to the following formula:
  • V_CCF is the content complexity
  • V_BR is the code rate
  • a is a preset value
  • b is a preset value
  • c is a preset value
  • Threshold is a preset value
  • Long multiply the value obtained by the width of the video.
  • the processor 402 is configured to: obtain a packet loss rate of the video, and an average number of packets included in each frame of the video, and use the packet loss rate.
  • the average video frame damage rate is calculated by the average number of packets included in each frame.
  • the average video frame damage rate may be calculated according to the following formula:
  • V_AIRF is the average video frame impairment rate
  • V_PktpF is the average number of packets per frame
  • V_LossRate is the packet loss rate
  • the processor 402 in acquiring the sequence damage rate of the video, is configured to: obtain a length of the picture group of the video, obtain a frame loss rate of the video, utilize a length of the picture group, and The frame loss rate is calculated, and the sequence damage rate is calculated. In the specific implementation process, it can be based on the following Calculate the damage rate of the sequence:
  • V_IR is the damage rate of the sequence
  • V_LossRateFrame is the frame loss rate
  • Goplength is the length of the picture group.
  • the processor 402 is configured to: obtain a total number of packets of the sequence of the video, and a packet loss rate of the video, by using a total number of packets of the sequence.
  • the packet loss rate calculates the frequency of the packet loss event.
  • the frequency of the packet loss event may be calculated according to the following formula:
  • V_PLEF is the frequency of the packet loss event
  • TotalPktNum is the total number of packets of the sequence
  • V_LossRate is the packet loss rate
  • V_Burst is the burst packet length of the video
  • V_PLEF is the frequency of the packet loss event
  • TotalPktNum is the total number of packets of the sequence
  • V_LossRate is the packet loss rate
  • V_Burst is the burst packet length of the video
  • V_PktpF is the packet included in the average frame per frame. number.
  • the average number of packets V_PktpF included in each frame may be a ratio of a total number of packets TotalPktNum of the sequence of the video to a total number of frames TotalFrameNum of the sequence of the video.
  • the processor 402 is configured to: obtain the packet loss rate and the average number of packets included in each frame, and calculate the frame loss according to the following formula:
  • V_LossRateFrame 1-(1-V_LossRate) V_PktpF
  • V_PktpF is the average number of packets included in each frame
  • V_LossRate is the packet loss rate
  • V_LossRateFrame is the frame loss rate of the video.
  • the video data quality evaluation device 400 can be any device that needs to output and play video, such as a laptop, a tablet, a personal computer, a hand. Machine and other equipment.
  • the invention provides an effective video quality objective measuring device, which overcomes a series of problems such as subjective evaluation with high complexity, high cost and possible damage to the human body.
  • the video data quality evaluation apparatus 400 provided by the embodiment of the present invention can effectively reflect the influence of the coding quality on the video quality, and fully utilizes the channel parameters, and converts the channel parameters or the data available in the network planning layer into the evaluation video quality.
  • the necessary code stream parameters and the probability model are used to calculate the impact of the packet loss on the video, so that the application where the specific video stream cannot be obtained can perform the video data quality evaluation and is more realistic.
  • the embodiment of the present invention further provides a computer storage medium, wherein the computer storage medium can store a program, and the program includes some or all of the steps of any one of the image prediction methods described in the foregoing method embodiments.
  • the storage medium may be a magnetic disk, an optical disk, a read-only memory (ROM), or a random access memory (RAM).

Landscapes

  • Engineering & Computer Science (AREA)
  • Signal Processing (AREA)
  • Multimedia (AREA)
  • Computer Networks & Wireless Communication (AREA)
  • Environmental & Geological Engineering (AREA)
  • Health & Medical Sciences (AREA)
  • Biomedical Technology (AREA)
  • General Health & Medical Sciences (AREA)
  • Compression Or Coding Systems Of Tv Signals (AREA)
  • Testing, Inspecting, Measuring Of Stereoscopic Televisions And Televisions (AREA)

Abstract

本发明涉及数字信号处理技术领域,公开了一种视频数据质量评估方法和装置,其中视频数据质量评估方法包括:获取用于获得所述视频的视频质量的参数,所述用于获得所述视频的视频质量的参数包括所述视频的内容复杂度、所述视频的平均视频帧损伤率、所述视频的序列损伤率、所述视频的丢包事件频率中至少一项;根据所述用于获得所述视频的视频质量的参数,获得所述视频的质量。使用本发明,可以使得不能获得具体的视频流的应用场合对视频数据进行质量评估。

Description

一种视频质量评估的方法和装置
本申请要求于2015年04月28日提交中国专利局、申请号为201510208427.5、发明名称为“一种视频质量评估的方法和装置”的中国专利申请的优先权,其全部内容通过引用结合在本申请中。
技术领域
本发明涉及数字信号处理技术领域,尤其是一种视频数据质量评估方法和装置。
背景技术
随着通信技术和计算机网络技术的发展,网络已经广泛应用于人们的日常工作生活中,越来越多的人们通过网络来随时随地的发送接收文本、语音以及图像视频信息,而由于视频信息能给人们以直观、生动的形象,包含最丰富的内容,所以网络视频业务得到了飞速发展,视频电话(Videophone)、视频会议(Video Conference)、网络电视(IPTV)等网络视频业务在人们的工作生活中扮演着越来越重要的角色。为了能够提供更好的视频服务质量,对网络视频质量的客观评价标准也得到了研究人员的重视。
网络规划层规划模型根据网络及承载视频的参数预测传输视频的质量,该模型通常应用于网络规划,在已知网络状况的情况下预测传输不同编码参数的视频流可以得到的视频质量。所以规划层模型可以指导我们进行合理的网络参数规划,获取更好的服务质量。然而对于某个特定的视频码流,如能获得其具体参数,该类模型也可用于对特定视频流进行质量监控。
现有技术中,可以通过得到包头信息,从而可以通过解包头来获取统计数据进行质量评估,其模型通过使用包含视频的内容复杂度(Video content complexity factor,V_CCF)在内的参数计算视频的编码质量V_MOSC,并将编码质量作为视频质量。
考虑到视频质量还受丢包的影响,因此该模型还可以将丢包的影响加入到模型中,从而将视频的丢包质量作为视频质量,视频的丢包质量V_MOSP为:
V_MOSP=V_MOSC-V_DP
其中V_DP为丢包后视频的失真,在计算V_DP时需要解包头,根据包 头信息分析出丢包信息,统计得到如下的丢包参数:平均的视频帧损伤率V_AIRF,序列的损伤率V_IR,丢包事件频率V_PLEF。
该模型还可进一步基于视频的编码质量和/或视频的丢包质量获得视频的重缓冲质量V_MOSR,从而将重缓冲质量作为视频质量。
对于包括网络规划层在内的一些应用场合中,并不能获得具体的视频流,因此不能得到具体的涉及具体视频流的统计参数,例如但不限于V_CCF、V_PLEF、V_AIRF以及V_IR等参数,所以针对不能获得具体的视频流且需要进行视频数据质量评估的应用场合,视频质量评估需要新的模型。
发明内容
本发明实施例提供了一种视频数据质量评估的方法和装置,用于解决不能获得具体的视频流应用场合无法进行视频数据质量评估的问题。
本发明第一方面提供了一种视频数据质量评估的方法,可包括:
获取用于获得所述视频的视频质量的参数,所述用于获得所述视频的视频质量的参数包括所述视频的内容复杂度、所述视频的平均视频帧损伤率、所述视频的序列损伤率、所述视频的丢包事件频率中至少一项;
根据所述用于获得所述视频的视频质量的参数,获得所述视频的质量;
所述获取用于获得所述视频的视频质量的质量参数包括:
获得所述视频的码率,利用所述码率计算所述内容复杂度;和/或
获得所述视频的丢包率和所述视频的平均每帧包含的包数,利用所述丢包率和所述平均每帧包含的包数,计算所述平均视频帧损伤率;和/或
获得所述视频的画面组的长度,获得所述视频的丢帧率,利用所述画面组的长度和所述丢帧率,计算所述序列损伤率;和/或
获得所述视频的序列的总包数和所述视频的丢包率,利用所述序列的总包数和所述丢包率计算所述丢包事件频率。
结合第一方面,在第一方面的第一种可能的实现方式中,所述利用所述码率计算所述内容复杂度包括:
根据如下公式计算所述内容复杂度:
Figure PCTCN2016075489-appb-000001
其中,V_CCF为所述内容复杂度,V_BR为所述码率,a为预设的值,b为预设的值,c为预设的值,Threshold为预设的值或利用所述视频的长乘所 述视频的宽计算得到的值。
结合第一方面或第一方面第一种可能的实现方式,在第一方面的第二种可能的实现方式中,所述利用所述丢包率和所述平均每帧包含的包数,计算
所述的平均视频帧损伤率包括:
根据如下公式计算所述的平均视频帧损伤率:
Figure PCTCN2016075489-appb-000002
其中,V_AIRF为所述的平均视频帧损伤率,V_PktpF为所述平均每帧包含的包数,V_LossRate为所述丢包率。
结合第一方面或第一方面的第一种或第二种可能的实现方式中,在第一方面的第三中可能的实现方式中,所述利用所述画面组的长度和所述丢帧率,计算所述序列损伤率包括:
根据如下公式计算所述序列的损伤率:
Figure PCTCN2016075489-appb-000003
其中,V_IR为所述序列的损伤率,V_LossRateFrame为所述丢帧率,Goplength为所述画面组的长度。
结合第一方面或第一方面的第一种至第三种可能的实现方式中的任意一种可能的实现方式,在第一方面的第四种可能的实现方式中,所述利用所述序列的总包数和所述丢包率计算所述丢包事件频率包括:
根据以下公式计算所述丢包事件频率:
Figure PCTCN2016075489-appb-000004
其中,V_PLEF为所述丢包事件频率;TotalPktNum为所述序列的总包数,V_LossRate为所述丢包率,V_Burst为所述视频的突发包长度;或
根据以下公式计算所述丢包事件频率:
Figure PCTCN2016075489-appb-000005
其中,V_PLEF为所述丢包事件频率;TotalPktNum为所述序列的总包数,V_LossRate为所述丢包率,V_Burst为所述视频的突发包长度,V_PktpF为所述平均每帧包含的包数。
结合第一方面或第一方面的第一种至第四种可能的实现方式中的任意 一种可能的实现方式,在第一方面的第五种可能的实现方式中,所述平均每帧包含的包数为所述视频的序列的总包数与所述视频的序列的总帧数的比值。
结合第一方面或第一方面的第一种至第五种可能的实现方式中的任意一种可能的实现方式,在第一方面的第六种可能的实现方式中,所述获得所述视频的丢帧率包括:
获得所述丢包率和所述平均每帧包含的包数,根据以下公式计算所述丢帧:
V_LossRateFrame=1-(1-V_LossRate)V_PktpF
其中,V_PktpF为所述平均每帧包含的包数,V_LossRate为所述丢包率,V_LossRateFrame为所述视频的丢帧率。
本发明第二方面提供了一种视频数据质量评估的装置,可包括:
获取模块,用于获取用于获得所述视频的视频质量的参数,所述用于获得所述视频的视频质量的参数包括所述视频的内容复杂度、所述视频的平均视频帧损伤率、所述视频的序列损伤率、所述视频的丢包事件频率中至少一项;
处理模块,用于根据所述获取模块获取得到的所述用于获得所述视频的视频质量的参数,获得所述视频的质量;
所述获取模块获取用于获得所述视频的视频质量的参数包括:
获得所述视频的码率,利用所述码率计算所述内容复杂度;和/或
获得所述视频的丢包率和所述视频的平均每帧包含的包数,利用所述丢包率和所述平均每帧包含的包数,计算所述平均视频帧损伤率;和/或
获得所述视频的画面组的长度,获得所述视频的丢帧率,利用所述画面组的长度和所述丢帧率,计算所述序列损伤率;和/或
获得所述视频的序列的总包数和所述视频的丢包率,利用所述序列的总包数和所述丢包率计算所述丢包事件频率。
结合第二方面,在第二方面的第一种可能的实现方式中,所述获取模块所述利用所述码率计算所述内容复杂度包括:
所述获取模块根据如下公式计算所述内容复杂度:
Figure PCTCN2016075489-appb-000006
其中,V_CCF为所述内容复杂度,V_BR为所述码率,a为预设的值,b为预设的值,c为预设的值,Threshold为预设的值或利用所述视频的长乘所述视频的宽计算得到的值。
结合第二方面或第二方面第一种可能的实现方式,在第二方面的第二种可能的实现方式中,所述获取模块所述利用所述丢包率和所述平均每帧包含的包数,计算所述的平均视频帧损伤率包括:
所述获取模块根据如下公式计算所述的平均视频帧损伤率:
Figure PCTCN2016075489-appb-000007
其中,V_AIRF为所述的平均视频帧损伤率,V_PktpF为所述平均每帧包含的包数,V_LossRate为所述丢包率。
结合第二方面或第二方面的第一种或第二种可能的实现方式中,在第二方面的第三种可能的实现方式中,所述获取模块所述利用所述画面组的长度和所述丢帧率,计算所述序列损伤率包括:
所述获取模块根据如下公式计算所述序列的损伤率:
Figure PCTCN2016075489-appb-000008
其中,V_IR为所述序列的损伤率,V_LossRateFrame为所述丢帧率,Goplength为所述画面组的长度。
结合第二方面或第二方面的第一种至第三种可能的实现方式中的任意一种可能的实现方式,在第二方面的第四种可能的实现方式中,所述获取模块所述利用所述序列的总包数和所述丢包率计算所述丢包事件频率包括:
所述获取模块根据以下公式计算所述丢包事件频率:
Figure PCTCN2016075489-appb-000009
其中,V_PLEF为所述丢包事件频率;TotalPktNum为所述序列的总包数,V_LossRate为所述丢包率,V_Burst为所述视频的突发包长度;或
根据以下公式计算所述丢包事件频率:
Figure PCTCN2016075489-appb-000010
其中,V_PLEF为所述丢包事件频率;TotalPktNum为所述序列的总包数,V_LossRate为所述丢包率,V_Burst为所述视频的突发包长度,V_PktpF为所述平均每帧包含的包数。
结合第二方面或第二方面的第一种至第四种可能的实现方式中的任意一种可能的实现方式,在第二方面的第五种可能的实现方式中,所述平均每帧包含的包数为所述视频的序列的总包数与所述视频的序列的总帧数的比值。
结合第二方面或第二方面的第一种至第五种可能的实现方式中的任意一种可能的实现方式,在第二方面的第六种可能的实现方式中,所述获取模块所述获得所述视频的丢帧率包括:
所述获取模块获得所述丢包率和所述平均每帧包含的包数,根据以下公式计算所述丢帧:
V_LossRateFrame=1-(1-V_LossRate)V_PktpF
其中,V_PktpF为所述平均每帧包含的包数,V_LossRate为所述丢包率,V_LossRateFrame为所述视频的丢帧率。
从本发明提供的技术方案可以看出,本发明基于拟合模型以及概率模型实现了对涉及具体视频流的统计参数进行计算,使得不能获得具体的视频流的应用场合可以进行视频数据质量评估。
附图说明
为了更清楚地说明本发明实施例中的技术方案,下面将对实施例描述中所需要使用的附图作简单地介绍,显而易见地,下面描述中的附图仅仅是本发明的一些实施例,对于本领域普通技术人员来讲,在不付出创造性劳动性的前提下,还可以根据这些附图获得其他的附图。
图1是本发明实施例提供的一种视频数据质量评估的方法;
图2是本发明实施例提供的另一种视频数据质量评估的方法;
图3是本发明实施例提供的一种视频数据质量评估的装置;
图4是本发明实施例提供的另一种视频数据质量评估的装置。
具体实施方式
下面将结合本发明实施例中的附图,对本发明实施例中的技术方案进行清楚、完整地描述。
需要说明,本发明提供的具体实施方式中,如无特别说明,对于用同一符号表示的参数可表示同一参数。
图1是本发明实施例提供的一种视频数据质量评估的方法,该方法中的参数可以但不限于为P.1202.1的模型中的参数,该方法步骤如下:
S102,利用所述视频的最大平均意见值和所述视频由于编码引进的失真,计算得到所述视频的编码质量。
可选的,可以根据以下公式获得所述视频的编码质量:
V_MOSC=MOS_MAX-V_DC
其中,MOS_MAX为所述视频的最大平均意见值,V_DC为所述视频由于编码引进的失真,V_MOSC为所述视频的编码质量
可选的,也可以根据以下公式获得所述视频的编码质量:
Figure PCTCN2016075489-appb-000011
其中,V_CCF为所述视频的内容复杂度,videoFrameRate为所述视频的帧率,v1为预设的系数值。
可选的,在具体实现过程中,在所述视频的帧率大于等于24时,可以根据以下公式获得所述视频的编码质量:
V_MOSC=MOS_MAX-V_DC
在所述视频的帧率小于24时,可以根据以下公式获得所述视频的编码质量:
Figure PCTCN2016075489-appb-000012
可选的,对于由于编码引进的失真V_DC,可以根据以下公式获得:
Figure PCTCN2016075489-appb-000013
其中,MOS_MIN为所述视频的最小平均意见值,V_NBR为标准化后的码率,v3、v4、v5、v6为预设的系数值。
可选的,对于内容复杂度V_CCF,可根据码率V_BR和I帧的平均码数V_ABIF这两个参数计算出来,计算公式如下:
Figure PCTCN2016075489-appb-000014
需要说明的是,所述视频的编码质量V_MOSC可以直接作为所述视频 的视频质量V_MOS,或者称为所述视频的平均意见值用于视频数据质量的评估。
S104,利用所述视频的编码质量和所述视频的由于丢包引进的失真,计算所述视频的丢包质量。
可选的,可以通过以下公式获得所述视频的丢包质量:
V_MOSP=V_MOSC-V_DP
其中,V_DP为所述视频的由于丢包引进的失真,V_MOSP为所述视频的丢包质量。
可选的,对于由于丢包引进的失真V_DP,可以根据以下公式获得:
Figure PCTCN2016075489-appb-000015
其中,V_AIRF为所述视频的平均视频帧损伤率,V_IR为所述视频的序列损伤率,V_PLEF为所述视频的丢包事件频率,v7、v8、v9、v10、v11、v12为预设的系数值。
可选的,还可以通过以下公式获得所述由于丢包引进的失真:
Figure PCTCN2016075489-appb-000016
可选的,在具体实现过程中,在所述视频的错误掩盖类型VideoPLC为SLICING时,可通过以下公式获得所述由于丢包引进的失真:
Figure PCTCN2016075489-appb-000017
在所述视频的错误掩盖类型VideoPLC为FREEZING时,可通过以下公式获得所述由于丢包引进的失真:
Figure PCTCN2016075489-appb-000018
可选的,对于所述视频的平均视频帧损伤率V_AIRF,可以根据以下公式获得:
Figure PCTCN2016075489-appb-000019
其中,V_NDF为所述视频的损伤帧数量,V_IRpFi为在预设的视频帧集合中的第i区间内平均视频帧损伤率。
可选的,对于所述视频的序列损伤率V_IR,可以根据以下公式获得:
Figure PCTCN2016075489-appb-000020
其中,V_TNF为所述视频的视频流中视频帧的总个数。
其中,对于丢包事件频率V_PLEF,可以根据具体的包来分析,当一帧内有丢包时,可以将丢包事件频率值进行加1,最后统计出丢包事件频率。
需要说明的是,所述视频的丢包质量V_MOSP可以直接作为所述视频的视频质量V_MOS,或者称为所述视频的平均意见值用于视频数据质量的评估。
S106,利用所述视频的编码质量和/或所述视频的丢包质量,计算所述视频的重缓冲质量。
可选的,重缓冲模型使用重缓冲次数、重缓冲时长和多次重缓冲时的重缓冲间隔来衡量视频质量受缓冲的影响,可以根据以下公式计算所述视频的重缓冲质量V_MOSR:
V_MOSR=Video_Quality-V_DR
Figure PCTCN2016075489-appb-000021
其中,Video_Quality表示的是当前的视频质量,NRE为重缓冲事件次数,ARL为重缓冲平均长度,MREEF为多重重缓冲事件影响因子,v13、v14、v15、v16、v17、v18为预设的系数值。
可选的,在具体实现过程中,需要根据信道情况来确定当前的视频质量,在即存在丢包情况又存在重缓冲情况时,则Video_Quality的值为所述视频的丢包质量;反之,Video_Quality的值为所述视频的编码质量。
所述视频的重缓冲质量V_MOSR可以直接作为所述视频的视频质量V_MOS,或者称为所述视频的平均意见值用于视频数据质量的评估。
本发明将模型分为编码质量模型、丢包质量模型和重缓冲质量模型。该 方式能够有效地体现编码质量对视频质量的影响。
结合图1中的视频数据质量评估方法,本发明实施例提供另一种视频数据质量评估的方法,如图2所示,该方法可以被认为是对P.1202.1的模型进行调整,根据规划层可以使用的参数来寻找简化的模型,即针对得不到的P.1202.1中的参数,例如视频的内容复杂度、视频的平均视频帧损伤率、视频的序列损伤率、视频的丢包事件频率,使用规划层的参数同时结合概率模型,推导映射出P.1201.1中的参数,进而实现质量评估。该方法步骤如下:
S202,获取用于获得所述视频的视频质量的参数,所述用于获得所述视频的视频质量的参数包括所述视频的内容复杂度、所述视频的平均视频帧损伤率、所述视频的序列损伤率、所述视频的丢包事件频率中至少一项。
可选的,获取所述视频的内容复杂度的过程可以具体为:获得所述视频的码率,利用所述码率计算所述内容复杂度。在具体实现过程中,可以根据如下公式计算所述内容复杂度:
Figure PCTCN2016075489-appb-000022
其中,V_CCF为所述内容复杂度,V_BR为所述码率,a为预设的值,b为预设的值,c为预设的值,Threshold为预设的值或利用所述视频的长乘所述视频的宽计算得到的值。其中,a、b、c、Threshold的值可以通过拟合训练获得,也可以根据表1中查询得到。Threshold的值也可以利用所述视频的长乘所述视频的宽计算得到,Threshold的值也可以利用所述视频的分辨率计算得到,计算公式可以具体为:
Threshold=Width×Height×d=Resolution×d
其中,Width为所述视频的宽,Height为所述视频的长,Resolution为所述视频的分辨率,d为预设的系数值,d的值可以通过拟合训练获得,也可以根据表2中查询得到。
Figure PCTCN2016075489-appb-000023
Figure PCTCN2016075489-appb-000024
表1.内容复杂度V_CCF的系数
  H264 MPEG4
d 0.02 0.012
表2.内容复杂度V_CCF的系数d
可选的,获取所述视频的平均帧帧损伤率的过程可以具体为:获得所述视频的丢包率和所述视频的平均每帧包含的包数,利用所述丢包率和所述平均每帧包含的包数,计算所述平均视频帧损伤率。在具体实现过程中,可以根据根据如下公式计算所述的平均视频帧损伤率:
Figure PCTCN2016075489-appb-000025
其中,V_AIRF为所述的平均视频帧损伤率,V_PktpF为所述平均每帧包含的包数,V_LossRate为所述丢包率。
可选的,获取所述视频的序列损伤率的过程可以具体为:获得所述视频的画面组的长度,获得所述视频的丢帧率,利用所述画面组的长度和所述丢帧率,计算所述序列损伤率。在具体实现过程中,可以根据如下公式计算所述序列的损伤率:
Figure PCTCN2016075489-appb-000026
其中,V_IR为所述序列的损伤率,V_LossRateFrame为所述丢帧率,Goplength为所述画面组的长度。
可选的,获取所述视频的丢包事件频率的过程可以具体为:获得所述视频的序列的总包数和所述视频的丢包率,利用所述序列的总包数和所述丢包率计算所述丢包事件频率。在具体实现过程中,可以根据以下公式计算所述丢包事件频率:
Figure PCTCN2016075489-appb-000027
其中,V_PLEF为所述丢包事件频率,TotalPktNum为所述序列的总包数,V_LossRate为所述丢包率,V_Burst为所述视频的突发包长度;或
根据以下公式计算所述丢包事件频率:
Figure PCTCN2016075489-appb-000028
其中,V_PLEF为所述丢包事件频率;TotalPktNum为所述序列的总包数,V_LossRate为所述丢包率,V_Burst为所述视频的突发包长度,V_PktpF为所述平均每帧包含的包数。
可选的,在具体实现过程中,可以将所述视频的突发包长度与所述平均每帧包含的包数的比值V_Burst/V_PktpF设置为V_ratio,在V_ratio小于1时,可以根据以下公式计算所述丢包事件频率:
Figure PCTCN2016075489-appb-000029
在V_ratio大于等于1时,可以根据以下公式计算所述丢包事件频率:
Figure PCTCN2016075489-appb-000030
可选的,所述平均每帧包含的包数V_PktpF可以为所述视频的序列的总包数TotalPktNum与所述视频的序列的总帧数TotalFrameNum的比值。所述视频的序列的总帧数TotalFrameNum的值可以是所述视频的帧率videoFrameRate与所述视频的时长MeasureTime的乘积的值。
可选的,获得所述视频的丢帧率的过程可以具体为:获得所述丢包率和所述平均每帧包含的包数,根据以下公式计算所述丢帧:
V_LossRateFrame=1-(1-V_LossRate)V_PktpF
其中,V_PktpF为所述平均每帧包含的包数,V_LossRate为所述丢包率,V_LossRateFrame为所述视频的丢帧率。
S204,根据所述用于获得所述视频的视频质量的参数,获得所述视频的质量。
根据所述视频的视频质量的参数,获得所述视频的质量的详细过程已在图1的具体实施方式中详细描述,在此不再赘述。
本发明提供一种有效的视频质量客观度量方法,克服了主观评价复杂度高、代价大、可能对人体造成伤害等一系列问题。
本发明能够有效地体现编码质量对视频质量的影响,同时充分利用了信道参数,将其信道参数或是网络规划层中可用的数据转换成了评价视频质量必须的码流参数,并使用概率模型来计算丢包对视频的影响,从而使得不能获得具体的视频流的应用场合可以进行视频数据质量评估,且更加符合实 际。
同时本发明充分考虑了人们观看视频时的心理感受,评价方法简单、可靠,可积极指导服务提供者提供更好的视频质量服务。
下面还提供用于实施上述方案的相关装置。
参见图3,本发明实施例还提供一种视频数据质量评估装置300,可包括:
获取模块302,用于获取用于获得所述视频的视频质量的参数,所述用于获得所述视频的视频质量的参数包括所述视频的内容复杂度、所述视频的平均视频帧损伤率、所述视频的序列损伤率、所述视频的丢包事件频率中至少一项。
可选的,所述获取模块302获取所述视频的内容复杂度可以是:获得所述视频的码率,利用所述码率计算所述内容复杂度。在具体实现过程中,可以根据如下公式计算所述内容复杂度:
Figure PCTCN2016075489-appb-000031
其中,V_CCF为所述内容复杂度,V_BR为所述码率,a为预设的值,b为预设的值,c为预设的值,Threshold为预设的值或利用所述视频的长乘所述视频的宽计算得到的值。
可选的,所述获取模块302获取所述平均视频帧损伤率可以是:获得所述视频的丢包率和所述视频的平均每帧包含的包数,利用所述丢包率和所述平均每帧包含的包数,计算所述平均视频帧损伤率。在具体实现过程中,可以根据根据如下公式计算所述的平均视频帧损伤率:
Figure PCTCN2016075489-appb-000032
其中,V_AIRF为所述的平均视频帧损伤率,V_PktpF为所述平均每帧包含的包数,V_LossRate为所述丢包率。
可选的,所述获取模块302获取所述序列损伤率可以是:获得所述视频的画面组的长度,获得所述视频的丢帧率,利用所述画面组的长度和所述丢帧率,计算所述序列损伤率。在具体实现过程中,可以根据如下公式计算所述序列的损伤率:
Figure PCTCN2016075489-appb-000033
其中,V_IR为所述序列的损伤率,V_LossRateFrame为所述丢帧率,Goplength为所述画面组的长度。
可选的,所述获取模块302获取所述丢包事件频率可以是:获得所述视频的序列的总包数和所述视频的丢包率,利用所述序列的总包数和所述丢包率计算所述丢包事件频率。在具体实现过程中,可以根据以下公式计算所述丢包事件频率:
Figure PCTCN2016075489-appb-000034
其中,V_PLEF为所述丢包事件频率,TotalPktNum为所述序列的总包数,V_LossRate为所述丢包率,V_Burst为所述视频的突发包长度;或
根据以下公式计算所述丢包事件频率:
Figure PCTCN2016075489-appb-000035
其中,V_PLEF为所述丢包事件频率;TotalPktNum为所述序列的总包数,V_LossRate为所述丢包率,V_Burst为所述视频的突发包长度,V_PktpF为所述平均每帧包含的包数。
可选的,所述平均每帧包含的包数V_PktpF可以为所述视频的序列的总包数TotalPktNum与所述视频的序列的总帧数TotalFrameNum的比值。
可选的,所述获取模块302获得所述视频的丢帧率可以是:获得所述丢包率和所述平均每帧包含的包数,根据以下公式计算所述丢帧:
V_LossRateFrame=1-(1-V_LossRate)V_PktpF
其中,V_PktpF为所述平均每帧包含的包数,V_LossRate为所述丢包率,V_LossRateFrame为所述视频的丢帧率。
处理模块304,用于根据所述获取模块302获取得到的所述用于获得所述视频的视频质量的参数,获得所述视频的质量。
可以理解的是,本实施例的视频数据质量评估装置300的各功能单元的功能可根据上述方法实施例中的方法具体实现,其具体实现过程可以参照上述方法实施例的相关描述,此处不再赘述。视频数据质量评估装置300可为任何需要输出、播放视频的装置,如笔记本电脑,平板电脑、个人电脑、手机等设备。
本发明提供一种有效的视频质量客观度量装置,克服了主观评价复杂度高、代价大、可能对人体造成伤害等一系列问题。
本发明实施例提供的视频数据质量评估装置300能够有效地体现编码质量对视频质量的影响,同时充分利用了信道参数,将其信道参数或是网络规划层中可用的数据转换成了评价视频质量必须的码流参数,并使用概率模型来计算丢包对视频的影响,从而使得不能获得具体的视频流的应用场合可以进行视频数据质量评估,且更加符合实际。
参见图4,图4为本发明实施例提供的视频数据质量评估装置400的示意图,视频数据质量评估装置400可包括至少一个总线401、与总线401相连的至少一个处理器402以及与总线401相连的至少一个存储器403。
其中,处理器402通过总线401调用存储器403中存储的代码或者指令以用于,获取用于获得所述视频的视频质量的参数,所述用于获得所述视频的视频质量的参数包括所述视频的内容复杂度、所述视频的平均视频帧损伤率、所述视频的序列损伤率、所述视频的丢包事件频率中至少一项。根据所述用于获得所述视频的视频质量的参数,获得所述视频的质量。
可选的,在获取所述视频的内容复杂度方面,处理器402可用于:获得所述视频的码率,利用所述码率计算所述内容复杂度。在具体实现过程中,可以根据如下公式计算所述内容复杂度:
Figure PCTCN2016075489-appb-000036
其中,V_CCF为所述内容复杂度,V_BR为所述码率,a为预设的值,b为预设的值,c为预设的值,Threshold为预设的值或利用所述视频的长乘所述视频的宽计算得到的值。
可选的,在获取所述视频的平均帧帧损伤率方面,处理器402可用于:获得所述视频的丢包率和所述视频的平均每帧包含的包数,利用所述丢包率和所述平均每帧包含的包数,计算所述平均视频帧损伤率。在具体实现过程中,可以根据根据如下公式计算所述的平均视频帧损伤率:
Figure PCTCN2016075489-appb-000037
其中,V_AIRF为所述的平均视频帧损伤率,V_PktpF为所述平均每帧包含的包数,V_LossRate为所述丢包率。
可选的,在获取所述视频的序列损伤率方面,处理器402可用于:获得所述视频的画面组的长度,获得所述视频的丢帧率,利用所述画面组的长度和所述丢帧率,计算所述序列损伤率。在具体实现过程中,可以根据如下公 式计算所述序列的损伤率:
Figure PCTCN2016075489-appb-000038
其中,V_IR为所述序列的损伤率,V_LossRateFrame为所述丢帧率,Goplength为所述画面组的长度。
可选的,在获取所述视频的丢包事件频率方面,处理器402可用于:获得所述视频的序列的总包数和所述视频的丢包率,利用所述序列的总包数和所述丢包率计算所述丢包事件频率。在具体实现过程中,可以根据以下公式计算所述丢包事件频率:
Figure PCTCN2016075489-appb-000039
其中,V_PLEF为所述丢包事件频率,TotalPktNum为所述序列的总包数,V_LossRate为所述丢包率,V_Burst为所述视频的突发包长度;或
根据以下公式计算所述丢包事件频率:
Figure PCTCN2016075489-appb-000040
其中,V_PLEF为所述丢包事件频率;TotalPktNum为所述序列的总包数,V_LossRate为所述丢包率,V_Burst为所述视频的突发包长度,V_PktpF为所述平均每帧包含的包数。
可选的,所述平均每帧包含的包数V_PktpF可以为所述视频的序列的总包数TotalPktNum与所述视频的序列的总帧数TotalFrameNum的比值。
可选的,在获得所述视频的丢帧率方面,处理器402可用于:获得所述丢包率和所述平均每帧包含的包数,根据以下公式计算所述丢帧:
V_LossRateFrame=1-(1-V_LossRate)V_PktpF
其中,V_PktpF为所述平均每帧包含的包数,V_LossRate为所述丢包率,V_LossRateFrame为所述视频的丢帧率。
根据所述视频的视频质量的参数,获得所述视频的质量的详细过程已在图1和图2的具体实施方式中详细描述,在此不再赘述。
可以理解的是,本实施例的视频数据质量评估装置400的各功能单元的功能可根据上述方法实施例中的方法具体实现,其具体实现过程可以参照上述方法实施例的相关描述,此处不再赘述。视频数据质量评估装置400可为任何需要输出、播放视频的装置,如笔记本电脑,平板电脑、个人电脑、手 机等设备。
本发明提供一种有效的视频质量客观度量装置,克服了主观评价复杂度高、代价大、可能对人体造成伤害等一系列问题。
本发明实施例提供的视频数据质量评估装置400能够有效地体现编码质量对视频质量的影响,同时充分利用了信道参数,将其信道参数或是网络规划层中可用的数据转换成了评价视频质量必须的码流参数,并使用概率模型来计算丢包对视频的影响,从而使得不能获得具体的视频流的应用场合可以进行视频数据质量评估,且更加符合实际。
本发明实施例还提供一种计算机存储介质,其中,该计算机存储介质可存储有程序,该程序执行时包括上述方法实施例中记载的任意一种图像预测方法的部分或全部步骤。
在上述实施例中,对各个实施例的描述都各有侧重,某个实施例中没有详述的部分,可以参见其他实施例的相关描述。
需要说明的是,对于前述的各方法实施例,为了简单描述,故将其都表述为一系列的动作组合,但是本领域技术人员应该知悉,本发明并不受所描述的动作顺序的限制,因为依据本发明,某些步骤可以采用其他顺序或者同时进行。其次,本领域技术人员也应该知悉,说明书中所描述的实施例均属于优选实施例,所涉及的动作和模块并不一定是本发明所必须的。
上述装置的各模块之间的信息交互、执行过程等内容,由于与本发明方法实施例基于同一构思,具体内容可参见本发明方法实施例中的叙述,此处不再赘述。
本领域普通技术人员可以理解实现上述实施例方法中的全部或部分流程,是可以通过计算机程序来指令相关的硬件来完成,上述的程序可存储于一计算机可读取存储介质中,该程序在执行时,可包括如上述各方法的实施例的流程。其中,上述的存储介质可为磁碟、光盘、只读存储记忆体(ROM:Read-Only Memory)或随机存储记忆体(RAM:Random Access Memory)等。
本文中应用了具体个例对本发明的原理及实施方式进行了阐述,以上实施例的说明只是用于帮助理解本发明的方法及其思想;同时,对于本领域的一般技术人员,依据本发明的思想,在具体实施方式及应用范围上均会有改变之处,综上所述,本说明书内容不应理解为对本发明的限制。

Claims (14)

  1. 一种视频数据质量评估的方法,其特征在于,包括:
    获取用于获得所述视频的视频质量的参数,所述用于获得所述视频的视频质量的参数包括所述视频的内容复杂度、所述视频的平均视频帧损伤率、所述视频的序列损伤率、所述视频的丢包事件频率中至少一项;
    根据所述用于获得所述视频的视频质量的参数,获得所述视频的质量;
    所述获取用于获得所述视频的视频质量的质量参数包括:
    获得所述视频的码率,利用所述码率计算所述内容复杂度;和/或
    获得所述视频的丢包率和所述视频的平均每帧包含的包数,利用所述丢包率和所述平均每帧包含的包数,计算所述平均视频帧损伤率;和/或
    获得所述视频的画面组的长度,获得所述视频的丢帧率,利用所述画面组的长度和所述丢帧率,计算所述序列损伤率;和/或
    获得所述视频的序列的总包数和所述视频的丢包率,利用所述序列的总包数和所述丢包率计算所述丢包事件频率。
  2. 根据权利要求1所述的方法,其特征在于,所述利用所述码率计算所述内容复杂度包括:
    根据如下公式计算所述内容复杂度:
    Figure PCTCN2016075489-appb-100001
    其中,V_CCF为所述内容复杂度,V_BR为所述码率,a为预设的值,b为预设的值,c为预设的值,Threshold为预设的值或利用所述视频的长乘所述视频的宽计算得到的值。
  3. 根据权利要求1或2所述的方法,其特征在于,所述利用所述丢包率和所述平均每帧包含的包数,计算所述的平均视频帧损伤率包括:
    根据如下公式计算所述的平均视频帧损伤率:
    Figure PCTCN2016075489-appb-100002
    其中,V_AIRF为所述的平均视频帧损伤率,V_PktpF为所述平均每帧包含的包数,V_LossRate为所述丢包率。
  4. 根据权利要求1至3任一项所述的方法,其特征在于,所述利用所述画面组的长度和所述丢帧率,计算所述序列损伤率包括:
    根据如下公式计算所述序列的损伤率:
    Figure PCTCN2016075489-appb-100003
    其中,V_IR为所述序列的损伤率,V_LossRateFrame为所述丢帧率,Goplength为所述画面组的长度。
  5. 根据权利要求1至4任一项所述的方法,其特征在于,所述利用所述序列的总包数和所述丢包率计算所述丢包事件频率包括:
    根据以下公式计算所述丢包事件频率:
    Figure PCTCN2016075489-appb-100004
    其中,V_PLEF为所述丢包事件频率,TotalPktNum为所述序列的总包数,V_LossRate为所述丢包率,V_Burst为所述视频的突发包长度;或
    根据以下公式计算所述丢包事件频率:
    Figure PCTCN2016075489-appb-100005
    其中,V_PLEF为所述丢包事件频率;TotalPktNum为所述序列的总包数,V_LossRate为所述丢包率,V_Burst为所述视频的突发包长度,V_PktpF为所述平均每帧包含的包数。
  6. 根据权利要求1至5任一项所述的方法,其特征在于,所述平均每帧包含的包数为所述视频的序列的总包数与所述视频的序列的总帧数的比值。
  7. 根据权利要求1至6任一项所述的方法,其特征在于,所述获得所述视频的丢帧率包括:
    获得所述丢包率和所述平均每帧包含的包数,根据以下公式计算所述丢帧:
    V_LossRateFrame=1-(1-V_LossRate)V_PktpF
    其中,V_PktpF为所述平均每帧包含的包数,V_LossRate为所述丢包率,V_LossRateFrame为所述视频的丢帧率。
  8. 一种视频数据质量评估的装置,其特征在于,包括:
    获取模块,用于获取用于获得所述视频的视频质量的参数,所述用于获得所述视频的视频质量的参数包括所述视频的内容复杂度、所述视频的平均 视频帧损伤率、所述视频的序列损伤率、所述视频的丢包事件频率中至少一项;
    处理模块,用于根据所述获取模块获取得到的所述用于获得所述视频的视频质量的参数,获得所述视频的质量;
    所述获取模块获取用于获得所述视频的视频质量的参数包括:
    获得所述视频的码率,利用所述码率计算所述内容复杂度;和/或
    获得所述视频的丢包率和所述视频的平均每帧包含的包数,利用所述丢包率和所述平均每帧包含的包数,计算所述平均视频帧损伤率;和/或
    获得所述视频的画面组的长度,获得所述视频的丢帧率,利用所述画面组的长度和所述丢帧率,计算所述序列损伤率;和/或
    获得所述视频的序列的总包数和所述视频的丢包率,利用所述序列的总包数和所述丢包率计算所述丢包事件频率。
  9. 根据权利要求8所述的方法,其特征在于,所述获取模块所述利用所述码率计算所述内容复杂度包括:
    所述获取模块根据如下公式计算所述内容复杂度:
    Figure PCTCN2016075489-appb-100006
    其中,V_CCF为所述内容复杂度,V_BR为所述码率,a为预设的值,b为预设的值,c为预设的值,Threshold为预设的值或利用所述视频的长乘所述视频的宽计算得到的值。
  10. 根据权利要求8或9所述的方法,其特征在于,所述获取模块所述利用所述丢包率和所述平均每帧包含的包数,计算所述的平均视频帧损伤率包括:
    所述获取模块根据如下公式计算所述的平均视频帧损伤率:
    Figure PCTCN2016075489-appb-100007
    其中,V_AIRF为所述的平均视频帧损伤率,V_PktpF为所述平均每帧包含的包数,V_LossRate为所述丢包率。
  11. 根据权利要求8至10任一项所述的方法,其特征在于,所述获取模块所述利用所述画面组的长度和所述丢帧率,计算所述序列损伤率包括:
    所述获取模块根据如下公式计算所述序列的损伤率:
    Figure PCTCN2016075489-appb-100008
    其中,V_IR为所述序列的损伤率,V_LossRateFrame为所述丢帧率,Goplength为所述画面组的长度。
  12. 根据权利要求8至11任一项所述的方法,其特征在于,所述获取模块所述利用所述序列的总包数和所述丢包率计算所述丢包事件频率包括:
    所述获取模块根据以下公式计算所述丢包事件频率:
    Figure PCTCN2016075489-appb-100009
    其中,V_PLEF为所述丢包事件频率;TotalPktNum为所述序列的总包数,V_LossRate为所述丢包率,V_Burst为所述视频的突发包长度;或
    根据以下公式计算所述丢包事件频率:
    Figure PCTCN2016075489-appb-100010
    其中,V_PLEF为所述丢包事件频率;TotalPktNum为所述序列的总包数,V_LossRate为所述丢包率,V_Burst为所述视频的突发包长度,V_PktpF为所述平均每帧包含的包数。
  13. 根据权利要求8至12任一项所述的方法,其特征在于,所述平均每帧包含的包数为所述视频的序列的总包数与所述视频的序列的总帧数的比值。
  14. 根据权利要求8至13任一项所述的方法,其特征在于,所述获取模块所述获得所述视频的丢帧率包括:
    所述获取模块获得所述丢包率和所述平均每帧包含的包数,根据以下公式计算所述丢帧:
    V_LossRateFrame=1-(1-V_LossRate)V_PktpF
    其中,V_PktpF为所述平均每帧包含的包数,V_LossRate为所述丢包率,V_LossRateFrame为所述视频的丢帧率。
PCT/CN2016/075489 2015-04-28 2016-03-03 一种视频质量评估的方法和装置 WO2016173320A1 (zh)

Priority Applications (2)

Application Number Priority Date Filing Date Title
EP16785751.5A EP3282698B1 (en) 2015-04-28 2016-03-03 Method and apparatus for evaluating video quality
US15/796,035 US10530675B2 (en) 2015-04-28 2017-10-27 Video quality estimation method and apparatus

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN201510208427.5A CN104811694B (zh) 2015-04-28 2015-04-28 一种视频数据质量评估的方法和装置
CN201510208427.5 2015-04-28

Related Child Applications (1)

Application Number Title Priority Date Filing Date
US15/796,035 Continuation US10530675B2 (en) 2015-04-28 2017-10-27 Video quality estimation method and apparatus

Publications (1)

Publication Number Publication Date
WO2016173320A1 true WO2016173320A1 (zh) 2016-11-03

Family

ID=53696145

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2016/075489 WO2016173320A1 (zh) 2015-04-28 2016-03-03 一种视频质量评估的方法和装置

Country Status (4)

Country Link
US (1) US10530675B2 (zh)
EP (1) EP3282698B1 (zh)
CN (1) CN104811694B (zh)
WO (1) WO2016173320A1 (zh)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN111629746A (zh) * 2017-11-21 2020-09-04 北卡罗来纳州立大学 具有超快响应的用于葡萄糖触发胰岛素递送的电荷可切换聚合物储库

Families Citing this family (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN104811694B (zh) 2015-04-28 2017-03-08 华为技术有限公司 一种视频数据质量评估的方法和装置
CN106507101B (zh) * 2016-11-21 2018-12-28 中国联合网络通信集团有限公司 一种视频业务质量评价方法及装置
CN108401263B (zh) * 2017-02-07 2021-09-17 大唐移动通信设备有限公司 一种语音质量的评估方法及装置
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
CN107734300A (zh) * 2017-10-19 2018-02-23 歌尔科技有限公司 图像传输方法及设备
WO2022155844A1 (zh) * 2021-01-21 2022-07-28 华为技术有限公司 一种视频传输质量的评估方法及装置

Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN101742353A (zh) * 2008-11-04 2010-06-16 工业和信息化部电信传输研究所 无参考视频质量评价方法
WO2011072414A1 (en) * 2009-12-15 2011-06-23 Thomson Licensing Method and apparatus for subjective assessment of video quality
CN103414915A (zh) * 2013-08-22 2013-11-27 合一网络技术(北京)有限公司 一种网站上传视频质量评价方法及装置
CN103634594A (zh) * 2012-08-21 2014-03-12 华为技术有限公司 一种获得视频编码压缩质量的方法及装置
CN104811694A (zh) * 2015-04-28 2015-07-29 华为技术有限公司 一种视频数据质量评估的方法和装置

Family Cites Families (8)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP2005244851A (ja) * 2004-02-27 2005-09-08 Kddi Corp ネットワーク性能評価装置およびその方法
RU2402885C2 (ru) * 2005-03-10 2010-10-27 Квэлкомм Инкорпорейтед Классификация контента для обработки мультимедийных данных
CN101227604B (zh) * 2007-01-18 2010-05-19 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
US20090046595A1 (en) * 2007-07-16 2009-02-19 Alan Clark Method and system for content estimation of packet video streams
JP5172440B2 (ja) * 2008-01-08 2013-03-27 日本電信電話株式会社 映像品質推定装置、方法およびプログラム
EP2106154A1 (en) 2008-03-28 2009-09-30 Deutsche Telekom AG Audio-visual quality estimation
CN104883563B (zh) * 2011-04-11 2017-04-12 华为技术有限公司 一种视频数据质量评估方法和装置
CN103379360B (zh) * 2012-04-23 2015-05-27 华为技术有限公司 一种视频质量评估方法和装置

Patent Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN101742353A (zh) * 2008-11-04 2010-06-16 工业和信息化部电信传输研究所 无参考视频质量评价方法
WO2011072414A1 (en) * 2009-12-15 2011-06-23 Thomson Licensing Method and apparatus for subjective assessment of video quality
CN103634594A (zh) * 2012-08-21 2014-03-12 华为技术有限公司 一种获得视频编码压缩质量的方法及装置
CN103414915A (zh) * 2013-08-22 2013-11-27 合一网络技术(北京)有限公司 一种网站上传视频质量评价方法及装置
CN104811694A (zh) * 2015-04-28 2015-07-29 华为技术有限公司 一种视频数据质量评估的方法和装置

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN111629746A (zh) * 2017-11-21 2020-09-04 北卡罗来纳州立大学 具有超快响应的用于葡萄糖触发胰岛素递送的电荷可切换聚合物储库
CN111629746B (zh) * 2017-11-21 2024-04-16 北卡罗来纳州立大学 具有超快响应的用于葡萄糖触发胰岛素递送的电荷可切换聚合物储库

Also Published As

Publication number Publication date
US10530675B2 (en) 2020-01-07
EP3282698A1 (en) 2018-02-14
CN104811694B (zh) 2017-03-08
EP3282698A4 (en) 2018-02-14
EP3282698B1 (en) 2021-05-26
CN104811694A (zh) 2015-07-29
US20180054369A1 (en) 2018-02-22

Similar Documents

Publication Publication Date Title
WO2016173320A1 (zh) 一种视频质量评估的方法和装置
JP5215288B2 (ja) 映像符号化のための時間的品質メトリック
US9419875B2 (en) Video quality assessment method and apparatus
US8908047B2 (en) Method and apparatus for obtaining video quality parameter, and electronic device
US9210419B2 (en) System and method for diagnostic modeling of audio and video quality of service
Yang et al. Content-adaptive packet-layer model for quality assessment of networked video services
Panayides et al. Open‐Source Telemedicine Platform for Wireless Medical Video Communication
KR20150027248A (ko) 비디오 인코딩 압축 품질을 획득하는 방법 및 장치
CN104702922A (zh) 一种视频传输方法和系统
KR20150003316A (ko) 미디어 품질 평가 방법 및 장치
WO2017084256A1 (zh) 一种视频质量评价方法及装置
US20150256822A1 (en) Method and Apparatus for Assessing Video Freeze Distortion Degree
US9723301B2 (en) Method and apparatus for context-based video quality assessment
US20150222905A1 (en) Method and apparatus for estimating content complexity for video quality assessment
EP2954677B1 (en) Method and apparatus for context-based video quality assessment
Usman et al. The impact of temporal impairment on quality of experience (QOE) in Video streaming: A no reference (NR) subjective and objective study
JP5405915B2 (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
CN104995914A (zh) 用于基于上下文的视频质量评估的方法和装置
Wang et al. A cross-layer based bandwidth and queue adaptations for wireless multimedia networks
CN104969548A (zh) 用于基于上下文的视频质量评估的方法和装置
Fajardo et al. Cross–layer cross–domain adaptation of mobile video services
JP2015065518A (ja) 映像品質推定装置、映像品質推定方法及びプログラム

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: 16785751

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE