JP4490374B2 - Video quality evaluation apparatus and method - Google Patents

Description

  The present invention relates to a communication quality evaluation technique, and more particularly to a video quality evaluation technique for evaluating a quality (subjective quality) perceived by a viewer for a video distributed via a network.

  In recent years, commercial video distribution services have become widespread with the broadbandization of Internet access lines. There are roughly two techniques for distributing video to viewers using such a network. One is a download technology for transferring video data from the video distribution server on the network to the viewer terminal and playing the video after the transfer is completed. Another method is to download the video data from the video distribution server on the network. It is a streaming technology that plays while playing. In video distribution, the latter technique is now widely recognized, and many video distribution systems such as Windows Media (registered trademark), Real, and QuickTime (registered trademark) employ this technology.

  When distributing video to viewers using streaming technology, if the network between the video distribution system and the viewer terminal is congested, network transmission bandwidth sufficient to distribute high-quality video cannot be used. Data may be missing. As a result, the encoded video cannot be decoded correctly, and block-like distortion or freezing occurs in the reproduced video, so that the subjective video quality perceived by the viewer is lowered.

In order to alleviate this quality degradation, the video distribution system performs processing such as creating encoded video of multi-stage quality and distributing video of optimal encoding speed according to the available network bandwidth. Yes.
Specifically, information on the amount of data (packet loss) that could not be correctly received from the video distribution system at the viewer terminal, information on temporal fluctuation of data arrival (delay fluctuation amount), or already arrived at the viewer terminal Information on the amount of data (buffering amount) before reproduction is returned to the video distribution system, and the degree of network congestion between the video distribution system and the viewer terminal is estimated from the information. When the video distribution system and the viewer terminal are congested, the video distribution system decreases the transmission band of the video for the viewer terminal by one level.

  As a result, the required network transmission bandwidth is reduced, and although video quality degradation due to a reduction in coding speed occurs, video distortion / freezing due to data loss (packet loss), sound interruptions, etc. It is possible to avoid a serious quality deterioration phenomenon. For example, a streaming media server that can automatically set an appropriate connection speed parameter in accordance with an actual communication environment in a streaming reception application for each client has been devised.

  When such a video distribution system is used, the video distribution bit rate fluctuates with time, so the quality perceived by the viewer also fluctuates with time. Further, even in a system in which the distribution bit rate is not dynamically changed, the quality deterioration occurs non-steadyly, so that the video quality is temporally changed as in the case described above. Therefore, in order to evaluate the quality perceived by the viewer with respect to the video distributed using the network, how to evaluate the temporally varying quality becomes an issue. In conventional quality evaluation technology, it is common to view a video of about 10 seconds and evaluate its quality. (1) Considering that time fluctuation of the delivery bit rate occurs in units of several seconds or more. Then, since it is impossible to evaluate various time fluctuations by simulating various images with a time of about 10 seconds, and (2) a certain amount of time is required for the viewer to understand the meaning of the video content. Quality assessment in units is required.

  Conventionally, as an example of a method for solving such a problem, an arbitrary evaluation is performed in consideration of the influence of quality fluctuations and the forgetting characteristic of the viewer's quality deterioration from the time series data of the video quality evaluation obtained every unit time. Techniques for estimating video quality within a period have been proposed (see, for example, Patent Document 1 and Non-Patent Document 1-3). Further, Document 2 shows that even if the quality evaluation value for 3 minutes is estimated using the average value of the quality evaluation values of the video for about 10 seconds, the quality evaluation estimation accuracy is very poor.

JP 2004-172753 A Kawaguchi Ginga, Hayashi Takanori, "Subjective quality assessment of the effect of bit rate fluctuation in video distribution services", July 2002, The Institute of Electronics, Information and Communication Engineers, Communication Quality Study Group, CQ2002-71 Kawaguchi Ginga, Hayashi Takanori, Okamoto Satoshi, Takahashi Satoshi, "Subjective Quality Estimation Model for Video Distribution Services Under QoS Control", 2003, The Institute of Electronics, Information and Communication Engineers, General Conference SB-7-1 Kawaguchi Ginga, Hayashi Takanori, Okamoto Satoshi, Takahashi Satoshi, "Examination of QoS Control Method for Multi-bitrate Video Distribution Service", September 2004, The Institute of Electronics, Information and Communication Engineers, Communication Quality Study Group, CQ2004-78

  However, in such a conventional technique, even when the video encoding speed fluctuates with time, the video quality perceived by the user can be evaluated with a certain degree of accuracy. Considering the difference between quality evaluation characteristics when viewing (quality evaluation characteristics during short-time viewing) and quality evaluation characteristics when viewing videos with an evaluation time longer than unit time (quality evaluation characteristics during long-time viewing) Therefore, there is a problem that the quality estimation accuracy is lowered depending on the size of the quality evaluation value.

  As shown in FIG. 10, when it is determined that the quality for a unit time of about 10 seconds is relatively good, for example, an image having the same quality level continues for an evaluation period longer than the unit time, for example, about 3 minutes. When viewed, the evaluation value tends to be evaluated higher than the quality evaluation value per unit time. On the other hand, if it is determined that the quality of the video for a unit time of about 10 seconds is relatively poor, the evaluation when the video of the same quality level is continuously viewed for an evaluation period longer than the unit time, for example, about 3 minutes. The value tends to be evaluated lower than the quality evaluation value per unit time. In the prior art, such characteristics cannot be reflected, and there is a problem that the estimation accuracy is lowered in a region where the quality evaluation value is high and a region where the quality evaluation value is low.

  The present invention is for solving such a problem, and considers the difference between the quality evaluation characteristic at the time of viewing of the viewer unit time and the quality evaluation characteristic at the time of viewing the evaluation period, and the quality perceived by the viewer. An object of the present invention is to provide a video quality evaluation apparatus and method that can be estimated with high accuracy.

  In order to achieve such an object, a video quality evaluation apparatus according to the present invention is a video quality evaluation apparatus that estimates and evaluates the quality perceived by a viewer for distribution video distributed via a network, A unit-time video quality evaluation value estimation unit that estimates the video quality evaluation value of the input distribution video as a unit-time video quality evaluation value every unit time and stores it in the storage unit, and obtains it from the storage unit within a desired evaluation period. Each unit time video quality evaluation value obtained, and from these unit time video quality evaluation values, an overall video quality evaluation value estimation unit for estimating a total video quality evaluation value in the evaluation period, and a viewer's unit time viewing time An overall video quality evaluation value correction unit that corrects the overall video quality evaluation value based on a correction function corresponding to a difference between the quality evaluation characteristic and the quality evaluation characteristic during viewing during the evaluation period;

  At this time, if the total video quality evaluation value estimated by the total video quality evaluation value estimation unit is high in the total video quality evaluation value correction unit, it is higher and the total video quality estimated by the total video quality evaluation value estimation unit If the evaluation value is low, it may be corrected so as to be lower.

  Also, the video quality evaluation method according to the present invention is used in a video quality evaluation apparatus that estimates and evaluates the quality perceived by a viewer for a distribution video distributed via a network, and is obtained from the distribution video every unit time. Based on each unit time video quality evaluation value obtained during the desired evaluation period, the total video quality evaluation value indicates the quality perceived by the viewer for the distribution video during the evaluation period. Is a video quality evaluation method that estimates the total video quality evaluation value based on a correction function according to the difference between the quality evaluation characteristics of the viewer during unit time viewing and the quality evaluation characteristics during viewing of the evaluation period. A video quality evaluation value correction step is provided.

  Another video quality evaluation method according to the present invention is a video quality evaluation method used in a video quality evaluation apparatus for estimating and evaluating the quality perceived by a viewer for a distribution video distributed via a network. The unit time video quality evaluation value estimating unit estimates the video quality evaluation value of the input distribution video as a unit time video quality evaluation value for each unit time and stores it in the storage unit; and Each unit time video quality evaluation value obtained within a desired evaluation period is acquired from the storage unit by the total video quality evaluation value estimation unit, and the total video quality evaluation value in the evaluation period is obtained from these unit time video quality evaluation values. The total video quality evaluation value estimation step to be estimated and the total video quality evaluation value correction unit, the quality evaluation characteristics of the viewer during unit time viewing and the quality evaluation characteristics during the evaluation period viewing On the basis of the correction function according to the had and a comprehensive image quality evaluation value correction step of correcting the overall image quality evaluation value.

  At this time, in the total video quality evaluation value correction step, if the total video quality evaluation value estimated in the total video quality evaluation value estimation step is high, it is higher and the total video quality estimated in the total video quality evaluation value estimation step You may correct | amend so that it may become still lower when an evaluation value is low.

  According to the present invention, the video quality evaluation value of the input video signal is estimated in time series as the unit time video quality evaluation value every unit time, and each unit time video quality evaluation obtained within a desired evaluation period is obtained. The total video quality evaluation value during the evaluation period is estimated from the value, and the difference between the quality evaluation characteristics during viewing of the viewer in unit time and the quality evaluation characteristics during viewing during the evaluation period is considered for the obtained total video quality evaluation value Thus, the quality perceived by the viewer can be estimated with high accuracy in consideration of the difference between the quality evaluation characteristics during the unit time viewing of the viewer and the quality evaluation characteristics during the viewing of the evaluation period.

Next, embodiments of the present invention will be described with reference to the drawings.
[First Embodiment]
First, the video quality evaluation apparatus according to the first embodiment of the present invention will be described with reference to FIG. FIG. 1 is a block diagram showing the configuration of the video quality evaluation apparatus according to the first embodiment of the present invention.
The video quality evaluation apparatus 1 is composed of an information processing apparatus having a communication function such as a personal computer. The video quality evaluation apparatus 1 has a quality (subjective quality) perceived by a viewer for a video distributed from a distribution server 2 via a network 3 such as the Internet. Has a function to evaluate.

  In this embodiment, the video quality evaluation value of the input video signal is estimated in time series as the unit time video quality evaluation value every unit time, and each unit time video quality evaluation value obtained within a desired evaluation period The overall video quality evaluation value during the evaluation period is estimated from the above, and the difference between the quality evaluation characteristics during viewing of the viewer in unit time and the quality evaluation characteristics during viewing during the evaluation period is considered for the obtained total video quality evaluation value. Correction is performed.

  Next, the configuration of the video quality evaluation apparatus according to the present embodiment will be described in detail with reference to FIG. The video quality evaluation apparatus 1 includes, as main functional units, a unit time video quality evaluation value estimation unit 11, a unit time video quality evaluation value storage unit 12, a total video quality evaluation value estimation unit 13, and a total video quality evaluation value correction. A portion 14 is provided.

Among these functional units, the unit time video quality evaluation value estimation unit 11, the total video quality evaluation value estimation unit 13, and the total video quality evaluation value correction unit 14 are configured by arithmetic processing units that perform various arithmetic processes. The arithmetic processing unit may be composed of a dedicated arithmetic processing circuit, but the CPU and its peripheral circuits are used to read the program from the CPU and its peripheral circuit memory and execute it, thereby executing the above hardware and program. You may implement | achieve the said function part by cooperating.
The unit time video quality evaluation value storage unit 12 includes a storage device such as a hard disk or a memory.

The unit time video quality evaluation value estimation unit 11 has a function of estimating the video quality evaluation value of the input video signal as a unit time video quality evaluation value for each unit time (short time) in time series, and unit time video quality evaluation. And a function of storing in the value storage unit 12.
The total video quality evaluation value estimation unit 13 obtains each unit time video quality evaluation value obtained within a desired evaluation period from the unit time video quality evaluation value storage unit 12, and from these unit time video quality evaluation values. And a function of estimating an overall video quality evaluation value during the evaluation period.
The total video quality evaluation value correction unit 14 has a function of correcting the estimated total video quality evaluation value based on a predetermined correction function.

[Operation of First Embodiment]
Next, the operation of the video quality evaluation apparatus according to the first embodiment of the present invention will be described with reference to FIG. FIG. 2 is a flowchart showing the video quality evaluation process of the video quality evaluation apparatus according to the first embodiment of the present invention.
The video quality evaluation apparatus 1 executes the video quality evaluation process of FIG. 2 when evaluating the quality (subjective quality) that the viewer feels about the video distributed from the distribution server 2 via the network 3.

In the video quality evaluation process of FIG. 2, the video quality evaluation apparatus 1 first performs a time series of unit time video quality evaluation values from the distribution video every unit time, for example, every 10 seconds, by the unit time video quality evaluation value estimation unit 11. The unit time video quality evaluation values are stored in the unit time video quality evaluation value storage unit 12 (step 100).
Next, the total video quality evaluation value estimation unit 13 acquires each unit time video quality evaluation value obtained within a desired evaluation period from the unit time video quality evaluation value storage unit 12, and the unit time video quality evaluation The total video quality evaluation value in the evaluation period is estimated from the value (step 101).
Subsequently, the total video quality evaluation value correction unit 14 corrects the estimated total video quality evaluation value based on a predetermined correction function (step 102), and the series of video quality evaluation processing ends.

  Next, with reference to FIGS. 3 to 5, each processing step constituting the video quality evaluation process of FIG. 2 will be described in detail. FIG. 3 is a flowchart showing unit time video quality evaluation value estimation processing. FIG. 4 is a flowchart showing the comprehensive video quality evaluation value estimation process. FIG. 5 is a flowchart showing the comprehensive video quality evaluation value correction process.

First, the unit time video quality evaluation value estimation processing will be described with reference to FIG. In step 100 of the video quality evaluation process of FIG. 2, the unit time video quality evaluation value estimation unit 11 executes the unit time video quality evaluation value estimation process of FIG.
In the unit time video quality evaluation value estimation process of FIG. 3, the unit time video quality evaluation value estimation unit 11 first obtains the encoding bit rate of the input distribution video (step 110), and this encoding bit rate. The video quality evaluation value for each unit time, for example, every 10 seconds, that is, the unit time video quality evaluation value is sequentially calculated (step 111).

  When calculating the unit time video quality evaluation value, the correspondence relationship between the encoding bit rate and the unit time video quality evaluation value is obtained in advance, and the encoding bit rate may be converted into the unit time video quality evaluation value. Specifically, in a video distribution application to be subjected to quality evaluation, the subjective quality evaluation characteristics of the video whose coding bit rate has been changed are obtained in advance, for example, a database. Video quality uses the mean opinion score (MOS) for each of the five levels of quality rating scales: “very good”, “good”, “normal”, “bad”, and “very bad”. 5 to 1 points, and the average value of the scores voted by 16 or more evaluators is called the average opinion score).

  For details of the subjective evaluation experiment method, see International Standardization Organization ITU-T (International Telecommunication Union Telecommunication Standardization Sector) Recommendation 910. Even for videos with the same coding bit rate, the quality achieved depends on the amount of motion and fineness of the video, so the evaluation results for many evaluation videos are used, and the average value is used as the representative value for subjective quality evaluation values. It can be obtained by defining. This unit time video quality evaluation value is derived, for example, by evaluating a video for 10 seconds. In addition, ITU-T recommendation J.I. It is also possible to derive the correspondence between the encoding bit rate and the video quality evaluation value per unit time using the video quality objective evaluation method described in No. 144.

Thereafter, the unit time video quality evaluation value estimation unit 11 stores the obtained unit time video quality evaluation values in the unit time video quality evaluation value storage unit 12 in time series (step 112), and a series of unit time video quality evaluations. The evaluation value estimation process is terminated.
These unit time video quality evaluation values are the original data for estimating the total video quality evaluation value. In the present embodiment, the desired evaluation period is set to 3 minutes, and the total video quality evaluation value in the evaluation period is estimated from each unit time video quality evaluation value obtained within this evaluation period. In step 111 of FIG. 3, the unit time is set to 10 seconds, and the unit time video quality evaluation value is obtained for each encoding bit rate of this unit time. Here, the video quality of 1 second is also set to the video quality of 10 seconds. Assuming that they are equivalent, 180 unit time video quality evaluation values are stored for 3 minutes.

Next, the overall video quality evaluation value estimation process will be described with reference to FIG. In step 102 of the video quality evaluation process of FIG. 2, the total video quality evaluation value estimation unit 13 executes the total video quality evaluation value estimation process of FIG.
In the overall video quality evaluation value estimation process of FIG. 4, the overall video quality evaluation value estimation unit 13 first obtains 180 unit times obtained from the unit time video quality evaluation value recording unit 12 within the evaluation period of the past 3 minutes. A video quality evaluation value is acquired (step 120).

Subsequently, the total video quality evaluation value estimator 13 considers the influence of quality fluctuations and the forgetting characteristics of the viewer's quality degradation, and thus estimates the total video quality evaluation value of the distribution video during the evaluation period. A total quality evaluation value obtained by weighted averaging the data group of the unit time video quality evaluation value by the quality degradation amount and the elapsed time length is derived.
Specifically, first, the average opinion score at time t is set to MOS (t), and the average value is calculated by the equation (1) using the MOS (t) at time t = −180 three minutes before the current time t = 0. The opinion score MOS (t) is converted into Q (t) which is a psychological distance measure (step 121).

Next, the total video quality evaluation value Q _est is calculated on the psychological distance scale for the data group of Q (t) by the equation (2) (step 122), and the total video quality evaluation value correction is performed. Output to the unit 14 (step 123), and the series of comprehensive video quality evaluation value estimation processing ends.

  Here, w and T represent constants and take values of w = 0.355 and T = 441, respectively. Exp {−wQ (t)} on the right side of Equation (2) represents a term that takes into consideration that the greater the deterioration, that is, the smaller the Q (t), the stronger the influence on the overall video quality. Yes. Also, exp (t / T) represents a term that takes into account that the influence of deterioration that occurred later in time is greater than that that occurred earlier.

Next, the overall video quality evaluation value correction process will be described with reference to FIG. In step 103 of the video quality evaluation process of FIG. 2, the total video quality evaluation value correction unit 14 executes the total video quality evaluation value correction process of FIG.
In the total video quality evaluation value correction process of FIG. 5, the total video quality evaluation value correction unit 14 first acquires the total video quality evaluation value Q _est in the evaluation period of the past 3 minutes from the unit time video quality evaluation value estimation unit 13. (Step 130).

Subsequently, the total video quality evaluation value correction unit 140 corrects the total video quality evaluation value correction by correcting the difference between the quality evaluation characteristic at the time of viewing the viewer unit time and the quality evaluation characteristic at the time of viewing the evaluation period according to the equation (3). The value Q _{est_c} is calculated (step 131). FIG. 6 is an explanatory diagram showing the difference in video quality evaluation characteristics before and after correction, and Equation (3) is a function for correcting the difference in these characteristics. In Expression (3), c1 to c3 represent constants, and take values of c ₁ = 0.117, c ₂ = 0.145, and c ₃ = 1.049, for example. Here, when the total video quality evaluation value estimated by the total video quality evaluation value estimation unit 140 is high, the total video quality evaluation value is corrected to be higher, and the total video quality evaluation value estimated by the total video quality evaluation value estimation unit 140 is If it is lower, correct it to be lower.

Next, the total video quality evaluation value correction unit 140 calculates the total video quality evaluation value correction value Q _{est_c} expressed on the psychological distance scale by the equation (4) on the original evaluation scale, here, the average opinion. The total video quality evaluation value MOS _all at the score is converted to MOS _all (step 132), and output as a desired total video quality evaluation value 15 corresponding to the distribution video in the evaluation period (step 133), and a series of total video quality evaluation value correction The process ends.

FIG. 7 is an explanatory diagram showing the estimation accuracy of the video quality evaluation value (with correction) obtained from the distribution video. FIG. 8 is an explanatory diagram showing the estimation accuracy of the video quality evaluation value (without correction) obtained from the distribution video.
In FIG. 7, the horizontal axis indicates the subjective quality evaluation value that the viewer feels by viewing the video for the evaluation period of 3 minutes, and the vertical axis indicates the subjective quality evaluation value obtained by the video quality evaluation apparatus according to the present embodiment. _Is an estimated value (MOS _all ). Here, the estimation results when applied to two types of contents A and B are plotted.

On the other hand, FIG. 8 shows the estimation of the quality evaluation value obtained by performing the inverse conversion to the MOS value by substituting Q _est obtained in Equation 2 as Q _{est_c} as it is into Equation 4 without _performing the correction of Equation 3 described above. The results are plotted.
The average value MCI (Mean of 95% Confidence Interval) of the statistical variation (95% confidence interval) of the subjective quality evaluation value itself is 0.23, whereas the estimation error (RMSE: Root Mean Square) in FIG. Error) is 0.19 smaller than MCI, and the estimated error in FIG. 8 is 0.25 larger than MCI. Therefore, the video quality evaluation apparatus according to the present embodiment has a smaller estimated error than MCI. It can be seen that the video quality can be estimated with sufficient accuracy for practical use.

  As described above, in this embodiment, the unit time video quality evaluation value estimation unit 11 estimates the video quality evaluation value of the input distribution video as the unit time video quality evaluation value for each unit time. The total video quality evaluation value in the evaluation period is estimated by the total video quality evaluation value estimation unit 13 from each unit time video quality evaluation value obtained within the desired evaluation period among the quality evaluation values. Since the total video quality evaluation value correction unit 14 corrects the total video quality evaluation value based on a correction function according to the difference between the quality evaluation characteristic at the time and the quality evaluation characteristic during viewing during the evaluation period, video encoding is performed. Even when the speed fluctuates over time, the quality perceived by the viewer can be estimated with high accuracy by taking into account the difference between the quality evaluation characteristics of the viewer during unit time viewing and the quality evaluation characteristics during viewing of the evaluation period.

[Second Embodiment]
Next, a video quality evaluation apparatus according to the second embodiment of the present invention will be described with reference to FIG. FIG. 9 is a flowchart showing unit time video quality evaluation value estimation processing of the video quality evaluation apparatus according to the second embodiment of the present invention.

  In the first embodiment, the unit time video quality evaluation value estimator 11 receives the encoded bit rate of the distribution video and inputs the video quality evaluation value per unit time from the relationship between the encoding bit rate and the video quality evaluation value. Was estimated and output. In the present embodiment, a case will be described in which the quality evaluation value variation is obtained as a time series from the distribution video signal. Note that the configuration of the video quality evaluation apparatus is the same as that shown in FIG. 1, and a detailed description thereof is omitted here.

In step 100 of the video quality evaluation process of FIG. 2, the unit time video quality evaluation value estimation unit 11 executes the unit time video quality evaluation value estimation process of FIG.
In the unit time video quality evaluation value estimation process of FIG. 9, the unit time video quality evaluation value estimation unit 11 first acquires the video signal of the input distribution video (step 200), and from this video signal for each unit time For example, video quality objective evaluation values every 10 seconds are sequentially calculated (step 201). At this time, the video quality objective evaluation value is, for example, ITU-T recommendation J.E. The video quality objective evaluation method described in 144 may be used.

Thereafter, the unit time video quality evaluation value estimation unit 11 stores the obtained video quality objective evaluation values in the unit time video quality evaluation value storage unit 12 in time series (step 202), and a series of unit time video quality evaluations. The value estimation process ends.
Thereby, when estimating a unit time video quality evaluation value, the fluctuation of the quality evaluation value can be obtained as a time series from the distribution video signal.

[Extended embodiment]
In each of the above embodiments, the case where the video quality evaluation apparatus 1 is configured by an apparatus separate from the viewer terminal has been described as an example. However, the present invention is not limited to this, and is executed by the viewer terminal. You may implement | achieve each function part of the video quality evaluation apparatus 1 with an application program. In this case, an application program for realizing each functional unit of the video quality evaluation apparatus 1 may be recorded on a recording medium such as a CD-ROM and distributed.

It is a block diagram which shows the structure of the video quality evaluation apparatus concerning the 1st Embodiment of this invention. It is a flowchart which shows the video quality evaluation process of the video quality evaluation apparatus concerning the 1st Embodiment of this invention. It is a flowchart which shows a unit time image | video quality evaluation value estimation process. It is a flowchart which shows a comprehensive image quality evaluation value estimation process. It is a flowchart which shows a comprehensive image quality evaluation value correction process. It is explanatory drawing which shows the difference in the video quality evaluation characteristic before and behind correction | amendment. It is explanatory drawing which shows the estimation precision of a video quality evaluation value (with correction | amendment). It is explanatory drawing which shows the estimation precision of a video quality evaluation value (no correction | amendment). It is a flowchart which shows the unit time video quality evaluation value estimation process of the video quality evaluation apparatus concerning the 2nd Embodiment of this invention. It is explanatory drawing which shows the difference of the quality evaluation characteristic at the time of a viewer's unit time viewing, and the quality evaluation characteristic at the time of evaluation period viewing.

Explanation of symbols

  DESCRIPTION OF SYMBOLS 1 ... Video quality evaluation apparatus, 11 ... Unit time video quality evaluation value estimation part, 12 ... Unit time video quality evaluation value memory | storage part, 13 ... Total video quality evaluation value estimation part, 14 ... Total video quality evaluation value correction part, 15 ... Total video quality evaluation value, 2 ... distribution server, 3 ... network.

Claims (4)

A video quality evaluation apparatus that estimates and evaluates the quality perceived by viewers for distribution video distributed via a network,
A unit time video quality evaluation value estimator for estimating the video quality evaluation value of the input distribution video as a unit time video quality evaluation value for each unit time,
Total video quality evaluation value estimation for obtaining each unit time video quality evaluation value obtained within a desired evaluation period from the storage unit and estimating the total video quality evaluation value in the evaluation period from these unit time video quality evaluation values And
A total video quality evaluation value correction unit that corrects the total video quality evaluation value based on a correction function according to a difference between a quality evaluation characteristic during viewing of the viewer unit time and a quality evaluation characteristic during viewing during the evaluation period. A video quality evaluation device characterized by The video quality evaluation apparatus according to claim 1,
The total video quality evaluation value correction unit is higher when the total video quality evaluation value estimated by the total video quality evaluation value estimation unit is higher, and the total video quality estimated by the total video quality evaluation value estimation unit A video quality evaluation apparatus, wherein when the evaluation value is low, correction is performed so that the evaluation value is further reduced. It is used in a video quality evaluation device that estimates and evaluates the quality perceived by viewers for distribution video distributed via a network. Among the unit time video quality evaluation values obtained for each unit time from the distribution video, a desired value Based on each unit time video quality evaluation value obtained during the evaluation period, a video quality evaluation method for estimating the quality perceived by the viewer for the distribution video in the evaluation period as an overall video quality evaluation value,
A comprehensive video quality evaluation value correction step for correcting the total video quality evaluation value based on a correction function according to a difference between a quality evaluation characteristic at the time of viewing by the viewer and a quality evaluation characteristic at the time of viewing the evaluation period; Characteristic video quality evaluation method. A video quality evaluation method used in a video quality evaluation apparatus for estimating and evaluating the quality perceived by a viewer with respect to distribution video distributed via a network,
A unit time video quality evaluation value estimating unit for estimating a video quality evaluation value of the input distribution video as a unit time video quality evaluation value for each unit time and storing it in the storage unit;
Each unit time video quality evaluation value obtained within a desired evaluation period is acquired from the storage unit by the total video quality evaluation value estimation unit, and the total video quality evaluation value in the evaluation period is obtained from these unit time video quality evaluation values. A total video quality evaluation value estimation step for estimating
Comprehensive video quality evaluation value correction unit corrects the comprehensive video quality evaluation value based on a correction function according to a difference between a quality evaluation characteristic during viewing of a viewer unit time and a quality evaluation characteristic during evaluation period viewing by a viewer. A video quality evaluation method comprising: a quality evaluation value correction step.

Images