JP2015122638A - Quality estimation device, method and program - Google Patents

Abstract

PROBLEM TO BE SOLVED: To estimate long time acoustic and image quality for encoding of sound, encoding and reproduction stop of an image in consideration of evaluation characteristics of a human such as an obliteration effect.SOLUTION: An image quality estimation device comprises: a short time AV encoding quality estimation part 310 which estimates short time AV encoding quality of an AV content; a long time AV encoding quality estimation part 320 which estimates long time AV encoding quality of the AV content on the basis of the short time AV encoding quality; a single communication deterioration amount estimation part 330 which estimates a single communication deterioration amount being an AV quality deterioration amount with respect to one reproduction stop of the AV content; a total communication deterioration amount estimation part 340 which estimates a total communication deterioration amount being a quality deterioration amount with respect to reproduction stop of the AV content on the basis of the single communication deterioration amount in a prescribed period; and an AV quality estimation part 350 which estimates the AV quality of the AV content on the basis of the long time AV encoding quality and the total communication deterioration amount.

Description

  The present invention relates to a quality estimation technique for distribution content, and more particularly to a quality estimation technique for estimating the quality of audio / video communication in an audio / video distribution service or an audio / video communication service provided via an IP (Internet Protocol) network. .

  In recent years, video distribution services and audio / music distribution services using network environments such as optical lines, 3G lines, and wireless LANs (Local Area Networks) have rapidly spread.

  Since the Internet does not necessarily guarantee the communication bandwidth, when communication is performed using audio / video media, if the network bandwidth between user terminals is narrow or the line is congested, the audio / video media will be used. On the other hand, the quality perceived by the user (user experience quality) decreases. For example, in a progressive download (PDL) type audio / video distribution service that plays audio / video in parallel with audio / video data download, if the communication bandwidth above the audio / video playback bit rate cannot be secured, The video is paused and the audio / video data is accumulated (rebuffered) in the buffer. At this time, sound and video pauses unintended by the user occur, so that the user experience quality deteriorates. It is known that such a reproduction stop has a great influence on the user experience quality (Non-Patent Document 1). Some PDL-type audio / video distribution services (adaptive streaming) dynamically prevent video from being stopped by video quality (resolution, frame rate, bit rate) and audio quality (bit rate) depending on the communication bandwidth. The quality control which switches is performed (nonpatent literature 2).

  As described above, in order to monitor the quality that changes depending on the state of the network or the like, a user experience quality estimation technique for quantifying the user experience quality is required. In addition, in order to perform quality control that dynamically maximizes the user experience quality, it is ideal to perform quality control by comparing user experience quality before and after switching the quality (bit rate or the like). Thus, in order to realize quality control based on the user experience quality, a user experience quality estimation technique that makes it possible to estimate the user experience quality is required. Since the quality of the PDL audio / video distribution service is affected by the video encoding quality, the audio encoding quality, and the playback stop, a user experience quality estimation technique considering these quality factors is required.

  As a technology for capturing such quality factors, objective evaluation of audio / video quality is made according to ITU-T (International Telecommunication Union-Telecommunication Standardization Sector) Recommendations P.1201.1 (Non-Patent Document 3) and P.1201.2 (Non-Patent Document 4). The law is stipulated. Recommendation P.1201.1 targets low resolution (QCIF, QVGA, HVGA) video, and Recommendation P.1201.2 targets high resolution (SD, HD) video. Recommendation P.1201.1 defines a model that can evaluate video quality against bit rate, frame rate, packet loss, and playback stop. Recommendation P.1201.2 defines video quality against changes in bit rate, frame rate, and packet loss. A model that can be evaluated is defined.

Japanese Unexamined Patent Publication No. 2004-172753

F. Dobrian et.al., "Understanding the Impact of Video Quality on User Engagement," in Proc. SIGCOMM, 2011. I. Sodagar, "The MPEG-DASH Standard for Multimedia Streaming Over the Internet," IEEE MultiMedia, vol.18, no.4, pp.62-67, Apr. 2011. ITU-T Rec. P.1201.1, "Parametric non-intrusive assessment of audiovisual media streaming quality-lower resolution application area," Oct. 2012. ITU-T Rec. P.1201.2, "Parametric non-intrusive assessment of audiovisual media streaming quality-higher resolution application area," Oct. 2012.

  Recommendations P.1201.1 (Non-Patent Document 3) and Recommendation P.1201.2 (Non-Patent Document 4) relate to technology for estimating short-time audio quality, video quality, and audio visual (AV) quality combining audio and video. According to this technology, it is impossible to estimate long-term sound quality, video quality, and AV quality for an audio / video distribution service in which sound quality / video quality dynamically changes. Also, in Recommendation P.1201.1, the average value of the stop time and the average value of the occurrence interval are used in the estimation of the quality degradation amount for the playback stop, and the forgetting effect that affects the user experience quality cannot be considered. Furthermore, Recommendation P.1201.2 does not define a method for estimating the amount of quality degradation with respect to playback stop.

  Patent Document 1 relates to a technique for estimating long-term sound quality and video quality in consideration of human evaluation characteristics such as forgetting effect from short-term quality. However, Patent Document 1 assumes a bit rate variation or the like as a quality factor, and does not take into account the influence of reproduction stop.

  In view of the above problems, the problem of the present invention is that it is possible to estimate long-term audio / video quality for audio encoding, video encoding and playback stop in consideration of human evaluation characteristics such as forgetting effect It is to provide quality estimation technology.

  In order to solve the above-described problem, an aspect of the present invention is a video quality estimation device that estimates video quality of video content, the short-time video coding quality that estimates short-time video coding quality of the video content. An estimation unit, a long-time video coding quality estimation unit for estimating a long-time video coding quality of the video content based on the short-time video coding quality, and a video for one stop of playback of the video content A single communication deterioration amount estimation unit that estimates a single communication deterioration amount that is a quality deterioration amount, and a total communication deterioration that is a quality deterioration amount with respect to the stop of playback of the video content based on the single communication deterioration amount in a predetermined period A total communication degradation amount estimation unit that estimates a video amount, and a video quality estimation unit that estimates the video quality of the video content based on the long-time video coding quality and the total communication degradation amount It relates to a video quality estimation apparatus to be.

  Another aspect of the present invention is an acoustic quality estimation apparatus that estimates the acoustic quality of acoustic content, the short-term acoustic encoding quality estimation unit that estimates the short-term acoustic encoding quality of the acoustic content, and the short A long-term acoustic coding quality estimation unit that estimates the long-term acoustic coding quality of the acoustic content based on the temporal acoustic coding quality; A single communication deterioration amount estimation unit that estimates a single communication deterioration amount, and a total communication that estimates a total communication deterioration amount that is a quality deterioration amount with respect to the reproduction stop of the acoustic content based on the single communication deterioration amount in a predetermined period An acoustic quality estimation apparatus comprising: a degradation amount estimation unit; and an acoustic quality estimation unit that estimates acoustic quality of the acoustic content based on the long-time acoustic coding quality and the total communication degradation amount About.

  Another aspect of the present invention is an AV quality estimation device that estimates AV quality of audiovisual (AV) content, and a short-time AV coding quality estimation unit that estimates short-time AV coding quality of the AV content. A long-term AV coding quality estimation unit that estimates long-term AV coding quality of the AV content based on the short-time AV coding quality, and AV quality degradation due to a single playback stop of the AV content A single communication deterioration amount estimation unit that estimates a single communication deterioration amount that is an amount, and a total communication deterioration amount that is a quality deterioration amount with respect to suspension of playback of the AV content based on the single communication deterioration amount in a predetermined period. The present invention relates to an AV quality estimation apparatus including an estimated total communication degradation amount estimation unit and an AV quality estimation unit that estimates AV quality of the AV content based on the long-time AV encoding quality and the total communication degradation amount.

  According to the present invention, it is possible to estimate the long-term audio / video quality for video encoding, audio encoding and playback stop taking into account human evaluation characteristics such as forgetting effect, and so on. It is possible to estimate the quality with higher accuracy than a simple average of audio and video quality.

FIG. 1 is a block diagram showing a configuration of a video quality estimation apparatus according to an embodiment of the present invention. FIG. 2 is a conceptual diagram related to estimation of short-time video coding quality and long-time video coding quality according to an embodiment of the present invention. FIG. 3 is a conceptual diagram regarding playback stop of a video quality estimation apparatus according to an embodiment of the present invention. FIG. 4 is a flowchart showing processing in the video quality estimation apparatus according to an embodiment of the present invention. FIG. 5 is a block diagram showing a configuration of a video quality estimation apparatus according to another embodiment of the present invention. FIG. 6 is a flowchart showing processing in the video quality estimation apparatus according to another embodiment of the present invention. FIG. 7 is a block diagram showing a configuration of an acoustic quality estimation apparatus according to an embodiment of the present invention. FIG. 8 is a conceptual diagram related to estimation of short-term acoustic coding quality and long-term acoustic coding quality according to an embodiment of the present invention. FIG. 9 is a conceptual diagram related to the stop of playback of the sound quality estimation apparatus according to an embodiment of the present invention. FIG. 10 is a flowchart showing processing in the sound quality estimation apparatus according to the embodiment of the present invention. FIG. 11 is a block diagram showing a configuration of an AV quality estimation apparatus according to an embodiment of the present invention. FIG. 12 is a conceptual diagram regarding short-time AV coding quality and long-time AV coding quality estimation according to an embodiment of the present invention. FIG. 13 is a conceptual diagram regarding playback stop of an AV quality estimation apparatus according to an embodiment of the present invention. FIG. 14 is a flowchart showing processing in the AV quality estimation apparatus according to an embodiment of the present invention. FIG. 15 is a block diagram showing a configuration of an AV quality estimation apparatus according to another embodiment of the present invention. FIG. 16 is a flowchart showing processing in the AV quality estimation apparatus according to another embodiment of the present invention. FIG. 17 is a block diagram showing a configuration of an AV quality estimation apparatus according to another embodiment of the present invention. FIG. 18 is a flowchart showing processing in the AV quality estimation apparatus according to another embodiment of the present invention.

  Hereinafter, embodiments of the present invention will be described with reference to the drawings.

[Video quality estimation device]
A video quality estimation apparatus according to a first embodiment of the present invention will be described with reference to FIGS.

  FIG. 1 is a block diagram showing a configuration of a video quality estimation apparatus according to an embodiment of the present invention. As shown in FIG. 1, the video quality estimation apparatus 100 includes a short-time video coding quality estimation unit 110, a long-time video coding quality estimation unit 120, a single communication degradation amount estimation unit 130, and a total communication degradation. A quantity estimation unit 140; and a video quality estimation unit 150.

The short time video coding quality estimation unit 110 estimates the short time video coding quality of the video content and provides the estimated short time video coding quality to the long time video coding quality estimation unit 120. In one embodiment, the short-time video coding quality estimation unit 110 calculates the short-time video coding quality V_MOSC _S based on the video resolution Rs, the video frame rate Fr, the video bit rate Br, and the video feature quantity Ir. The short time video coding quality V_MOSC _S is output to the long time video coding quality estimation unit 120.

  Here, the short-time video encoding quality is video quality considering deterioration due to encoding of video data for a short time (for example, a predetermined period of about several seconds to several tens of seconds). Here, it is assumed that short-time video encoding quality estimation section 110 estimates video quality for T seconds. The video resolution is the number of pixels per frame of the encoded video, the video frame rate is the number of frames per second of the encoded video, and the video bit rate is the encoded video data. It is the bit amount per second, and the video feature amount is an average bit amount of the I frame. These video resolution, video frame rate, video bit rate, and video feature quantity may be estimated and acquired from video data packets in a network or a video display terminal.

  The specific calculation of the short-time video coding quality in the short-time video coding quality estimation unit 110 is defined in, for example, ITU-T Recommendations P.1201.1 (Non-Patent Document 3) and P.1201.2 (Non-Patent Document 4). It can be calculated using the method that has been described.

In one embodiment, the short-time video coding quality V_MOSC _S may be calculated as follows using the following method that is an extension of ITU-T recommendation P.1201.1. First, using video bit rate Br and video feature amount (average bit amount of I frame) ABIF,

を算出する。ここで、Minは、引数として与えられた値のうち小さい方の値を返す関数であり、例えば、Min(1,2)=1である。上記映像特徴量ABIFは、パケットなどの情報から推定可能であり、何れか適切な方法が利用されてもよい。

Is calculated. Here, Min is a function that returns the smaller one of the values given as arguments, for example, Min (1,2) = 1. The video feature amount ABIF can be estimated from information such as a packet, and any appropriate method may be used.

  Next, the parameter NBr related to the bit amount per pixel is

により算出される。しかしながら、NBrは上記の式に限定されず、主観品質評価特性に合った式を用いればよく、例えば、

Is calculated by However, NBr is not limited to the above formula, and a formula suitable for the subjective quality evaluation characteristic may be used.

として算出されてもよいし、その他の式により算出されてもよい。パラメータv11, v12は、予め定められた係数であり、主観品質実験等によりサービス事業者が決めてよい。

Or may be calculated by other formulas. Parameters v11 and v12 are predetermined coefficients, and may be determined by the service provider through subjective quality experiments or the like.

  Next, the short time video coding quality estimation unit 110 calculates a coding degradation amount DC based on the video bit rate Br, the video frame rate Fr, the video resolution Rs, and the calculated video feature amount CCF. For example, the coding degradation amount DC is based on the NBr and CCF calculated above.

により算出されてもよい。ここで、パラメータA₁, A₂, v21-24は、予め定められる係数であり、サービス事業者が決めてよい。

May be calculated. Here, the parameters A ₁ , A ₂ , and v21-24 are predetermined coefficients, and may be determined by the service provider.

The short-time video encoding quality V_MOSC _S is calculated using the calculated encoding degradation amount DC and threshold values Th1 and Th2.

により算出されてもよい。ここで、パラメータA₃、閾値Th1, Th2及び係数v25-29は、所定の定数であり、主観品質実験等によりサービス事業者が決めてよい。だたし、上記の短時間映像符号化品質V_MOSC_Sの算出において、閾値Th1, Th2を用いることなく、

May be calculated. Here, the parameter A ₃ , the threshold values Th1, Th2 and the coefficient v25-29 are predetermined constants, and may be determined by the service provider through subjective quality experiments or the like. However, in the calculation of the short-time video encoding quality V_MOSC _S described above, without using the thresholds Th1 and Th2,

の何れかにより算出してもよい。

You may calculate by either.

  The long-time video coding quality estimation unit 120 estimates the long-time video coding quality of the video content based on the estimated short-time video coding quality, and the estimated long-time video coding quality is used as the video quality estimation unit 150. To provide. In one embodiment, the long-time video coding quality estimation unit 120 estimates the long-time video coding quality of video content based on the short-time video coding quality received from the short-time video coding quality estimation unit 110.

  Here, the long-time video encoding quality is, for example, video encoding quality of several minutes to several tens of minutes. Here, the long-time video encoding quality estimation unit 120 is configured for T seconds set for the short-time video encoding quality.

秒間の映像品質を推定するものとする（Nは１以上の整数）。長時間映像符号化品質V_MOSC_Lは、短時間映像符号化品質推定部１１０が所定の期間内に出力したM個の各T秒間の品質値

Assume that the video quality per second is estimated (N is an integer of 1 or more). The long-time video coding quality V_MOSC _L is a quality value for each of M T seconds output by the short-time video coding quality estimation unit 110 within a predetermined period.

のうち、直近のN個を単純に平均して、

Of these, simply average the last N,

により求められてもよい。

May be required.

FIG. 2 is a conceptual diagram related to estimation of short-time video coding quality and long-time video coding quality according to an embodiment of the present invention. As shown in FIG. 2, V_MOSC _S (1) is the recently output short-term video coding quality, and V_MOSC _S (i) is obtained from V_MOSC _S (1) when there is no playback stop.

秒前に出力された短時間映像符号化品質である。

This is the short-time video encoding quality output a second ago.

  However, the long-term video coding quality is not limited to the above formula, and in another embodiment, the long-time video coding quality estimation unit 120 has a forgetting effect based on the short-time video coding quality. Considering this, the long-term video encoding quality of the video content may be estimated. For example, long-term video coding quality takes into account the forgetting effect and quality evaluation characteristics for quality change (perceiving the change as deterioration when the quality changes)

により求められてもよい。ここで、第１項は、視聴直後のものはよく覚えているので影響は大きいが、時間が経過すると忘れていく忘却効果を示しており、第２項は、品質の変化が発生すると"変化"の絶対値が劣化として感じられ、それは視聴直後であるほど強いことを表している。パラメータA₄,A₅は、予め定められた係数であり、サービス事業者が決めてよい。また、w₁(i)は重み関数であり、例えば係数v30を用いて、

May be required. Here, the first term shows a forgetting effect that is greatly affected because it is remembered immediately after viewing, but the forgetting effect that is forgotten over time, and the second term is “change” when a change in quality occurs The absolute value of "is felt as a deterioration, which indicates that it is stronger as soon as it is viewed. The parameters A ₄ and A ₅ are predetermined coefficients and may be determined by the service provider. W ₁ (i) is a weighting function, for example, using a coefficient v30,

で計算されてもよい。

May be calculated by

The single communication degradation amount estimation unit 130 estimates a single communication degradation amount that is a video quality degradation amount for one stop of playback of video content, and the estimated single communication degradation amount is sent to the total communication degradation amount estimation unit 140. provide. In one embodiment, the single communication deterioration amount estimation unit 130 estimates the single communication deterioration amount based on the playback stop time length of the video content playback stop. For example, the playback stop time length that is the length of time that the video stops Based on RL, a single communication degradation amount DR _S (RL), which is a degradation amount of video quality due to the stop of reproduction, is estimated.

FIG. 3 is a conceptual diagram regarding playback stop of a video quality estimation apparatus according to an embodiment of the present invention. As shown in FIG. 3, the playback stop occurrence time is set to ST1, ST2,... From the side closer to the current time T _NOW , and the playback stop time length is also set to RL1, RL2,. Single communication degradation amount DR _S (RL), for example, using a threshold Th3,

として算出できる。ここで、パラメータv31-33は、予め定められた係数であり、サービス事業者が決めてよい。上記のRL<Th3であるときは、再生停止を知覚してからある秒数（例えば、１秒程度）までは急激に品質が劣化し、その後Th3（例えば、Th3=2秒）まではあまり品質が変化しないことを表している。他方、RL≧Th3であるときは、RLがTh3秒経過した後、緩やかに品質が低下していくことを表している。

Can be calculated as Here, the parameter v31-33 is a predetermined coefficient and may be determined by the service provider. When RL <Th3, the quality deteriorates rapidly until a certain number of seconds (eg, about 1 second) after the playback stop is perceived, and then the quality is too low until Th3 (eg, Th3 = 2 seconds). Represents no change. On the other hand, when RL ≧ Th3, it indicates that the quality gradually decreases after RL has passed Th3 seconds.

The calculation of DR _S (RL) is not limited to the above formula, for example,

として算出されてもよい。あるいは、Th3を用いずに、

May be calculated as Or without using Th3,

として算出されてもよい。ただし、パラメータv34-39は、v31-33と同様に予め定められた係数である。

May be calculated as However, the parameter v34-39 is a predetermined coefficient similarly to v31-33.

  The total communication degradation amount estimation unit 140 estimates a total communication degradation amount that is a quality degradation amount with respect to video content reproduction stop based on a plurality of single communication degradation amounts estimated in a predetermined period. The amount is provided to the video quality estimation unit 150. In one embodiment, the total communication degradation amount estimation unit 140 considers the forgetting effect based on the playback stop time length of each video content playback stop, the single communication degradation amount, and the playback stop time in a predetermined period. Estimate the amount of degradation. For example, the total communication degradation amount estimation unit 140 may include the time series when the playback stop occurs.

と、単一通信劣化量推定部１３０が算出した単一通信劣化量の系列

And a series of single communication deterioration amounts calculated by the single communication deterioration amount estimation unit 130

とに基づき、所定の期間内に発生した再生停止に対する品質劣化量である総通信劣化量DR_Mを推定する。ここで、パラメータPは、所定の期間内に発生した再生停止の回数である。総通信劣化量DR_Mは、

Preparative the basis to estimate the total communication deterioration amount DR _M is a quality deterioration amount with respect to the reproduction stop that occurred within a predetermined time period. Here, the parameter P is the number of times playback has stopped within a predetermined period. The total communication deterioration amount DR _M is,

により算出できる。ここで、w₂は重み関数であり、例えば、係数v41,v42を用いて、

Can be calculated. Here, w ₂ is a weight function, for example, using coefficients v41 and v42,

などとして算出できる。しかしながら、重み関数w₂は、上記の式に限定されず、例えば、０より大きく１以下の定数v43を用いて、

And so on. However, the weighting function w ₂ is not limited to the above formula, and for example, using a constant v43 that is greater than 0 and equal to or less than 1,

として算出されてもよい。重み関数の積は、何回も前の再生停止や、発生後に時間が経過している再生停止が最近発生した再生停止よりも映像品質に与える影響が小さいこと（忘却効果）を表している。総通信劣化量の算出は、上記の式に限定されず、例えば、現在時刻T_NOWを用いて、

May be calculated as The product of the weighting functions represents that the previous playback stop or the playback stop that has passed after the occurrence has a smaller influence on the video quality than the recently generated playback stop (forgetting effect). The calculation of the total communication deterioration amount is not limited to the above formula, for example, using the current time T _NOW ,

として算出してもよい。w₃は、w₂と同様の重み関数であり、忘却効果を表している。

May be calculated as w ₃ is a weight function similar to w ₂ and represents the forgetting effect.

The video quality estimation unit 150 estimates the video quality of the video content based on the long-time video encoding quality and the total communication degradation amount. In one embodiment, the video quality estimation unit 150 includes the long-time video coding quality V_MOSC _L calculated by the long-time video coding quality estimation unit 120 and the total communication degradation amount DR _M calculated by the total communication degradation amount estimation unit 140. Based on the above, the video quality V_MOS is estimated. The video quality V_MOS is, for example,

により算出できる。ここで、パラメータA₄は、予め定められた係数であり、サービス提供者が決めてよい。

Can be calculated. Here, the parameter A ₄ is a predetermined coefficient and may be determined by the service provider.

However, the calculation of the video quality V_MOS is not limited to the above formula, for example, by using the coefficients A _5,

として算出されてもよい。

May be calculated as

  FIG. 4 is a flowchart showing processing in the video quality estimation apparatus according to an embodiment of the present invention. As shown in FIG. 4, the process is started in response to input of video resolution, video frame rate, video bit rate, video feature amount, playback stop time length, and playback stop time.

  In step S101, the short-time video encoding quality estimation unit 110 calculates the short-time video encoding quality from the input video resolution, video frame rate, video bit rate, and video feature amount according to any of the calculation methods described above. The calculated short-time video coding quality is provided to the long-time video coding quality estimation unit 120.

  In step S102, the long-time video coding quality estimation unit 120 estimates the long-time video coding quality based on the received short-time video coding quality according to any of the calculation methods described above.

  In step S103, the single communication deterioration amount estimation unit 130 calculates a single communication deterioration amount for each reproduction stop based on the input reproduction stop time length according to any of the calculation methods described above, and calculates the calculated single The communication deterioration amount is provided to the total communication deterioration amount estimation unit 140.

  In step S104, the total communication deterioration amount estimation unit 140 calculates the total communication deterioration amount based on the received single communication deterioration amount, the input reproduction stop time length, and the reproduction stop time according to any of the above-described calculation methods. And provided to the video quality estimation unit 150.

  As shown in the figure, steps S101 and S102 and steps S103 and S104 may be executed in parallel.

  In step S105, the video quality estimation unit 150 calculates video quality based on the received long-time video encoding quality and the total communication degradation amount according to any of the above-described calculation methods, and outputs the calculated video quality evaluation value. To do. Thereafter, the video quality estimation apparatus 100 ends the process.

  According to this embodiment, the amount of quality degradation due to playback stop is calculated in consideration of the forgetting effect, and the total video quality estimation for long-time video viewing is performed together with the long-time video coding quality. Can do. Thereby, it is possible to estimate the video quality with higher accuracy than the calculation based on a simple average value of the reproduction stop time.

[Modification of video quality estimation device]
FIG. 5 is a block diagram showing a configuration of a video quality estimation apparatus according to another embodiment of the present invention. The video quality estimation apparatus according to the present embodiment estimates video quality of video content based on video resolution, video frame rate, and video bit rate without using video feature quantities.

  As shown in FIG. 5, the video quality estimation apparatus 100 ′ includes a short-time video coding quality estimation unit 110 ′, a long-time video coding quality estimation unit 120, a single communication degradation amount estimation unit 130, A communication degradation amount estimation unit 140 and a video quality estimation unit 150 are included. The video quality estimation apparatus 100 ′ according to the present embodiment is different from the video quality estimation apparatus 100 shown in FIG. 1 in that the video feature quantity (average bit amount of I frame in the embodiment of FIG. 1) is not input. The method of calculating the short time video coding quality in the temporal video coding quality estimation unit 110 ′ is different from that of the short time video coding quality estimation unit 110 shown in FIG. Therefore, here, the processing of the short-time video coding quality estimation unit 110 ′ will be described, and the long-time video coding quality estimation unit 120, the single communication degradation amount estimation unit 130, the total communication degradation amount estimation unit 140, and the video quality Since the estimation unit 150 is the same as that of the video quality estimation apparatus 100, description thereof will be omitted.

The short time video coding quality estimation unit 110 ′ calculates the short time video coding quality V_MOSC _S of the video content based on the video resolution Rs, the video frame rate Fr, and the video bit rate Br, and the calculated short time video code The encoded quality MOSC _S is output to the long-time video encoding quality estimation unit 120. Similar to the short-time video coding quality estimation unit 110 in FIG. 1, the short-time video coding quality estimation unit 110 ′ relates to the bit amount per pixel based on the video resolution Rs, the video frame rate Fr, and the video bit rate Br. The parameter NBr is calculated, and DC is calculated based on the calculated NBr, for example,

により算出する。パラメータA₆, A₇, v51, v52は、予め定められた係数であり、主観品質実験等によりサービス事業者が決めてよい。

Calculated by Parameters A ₆ , A ₇ , v 51, v 52 are predetermined coefficients, and may be determined by the service provider through subjective quality experiments or the like.

The short-time video coding quality V_MOSC _S uses the coding degradation amount DC and the threshold values Th4 and Th5,

により算出されてもよい。ここで、閾値Th4, Th5及び係数v53-57は、定数であり、主観品質実験等によりサービス事業者が決めてよい。ただし、上記の短時間映像符号化品質V_MOSC_Sの算出において、閾値Th4, Th5を用いることなく

May be calculated. Here, the threshold values Th4 and Th5 and the coefficient v53-57 are constants, and may be determined by the service provider through subjective quality experiments or the like. However, without calculating the short-time video coding quality V_MOSC _S , the thresholds Th4 and Th5 are not used.

の何れかにより算出してもよい。

You may calculate by either.

  Moreover, the calculation of short-time video encoding quality is not limited to the above method, and for example, it can be obtained as follows using the above-mentioned NBr.

ここで、パラメータA9,v61-69は、予め定められた係数であり、主観品質実験等によりサービス事業者が決めてよい。

Here, the parameters A9, v61-69 are predetermined coefficients, and may be determined by the service provider through subjective quality experiments or the like.

  FIG. 6 is a flowchart showing processing in the video quality estimation apparatus according to an embodiment of the present invention. As shown in FIG. 6, the processing is started in response to input of video resolution, video frame rate, video bit rate, playback stop time length, and playback stop time.

  In step S101 ′, the short-time video coding quality estimation unit 110 ′ calculates the short-time video coding quality from the input video resolution, video frame rate, and video bit rate according to any of the calculation methods described above. The calculated short-time video coding quality is provided to the long-time video coding quality estimation unit 120.

  In step S102, the long-time video coding quality estimation unit 120 estimates the long-time video coding quality based on the received short-time video coding quality according to any of the calculation methods described above.

  In step S103, the single communication deterioration amount estimation unit 130 calculates a single communication deterioration amount for each reproduction stop based on the input reproduction stop time length according to any of the calculation methods described above, and calculates the calculated single The communication deterioration amount is provided to the total communication deterioration amount estimation unit 140.

  In step S104, the total communication deterioration amount estimation unit 140 calculates the total communication deterioration amount based on the received single communication deterioration amount, the input reproduction stop time length, and the reproduction stop time according to any of the above-described calculation methods. And provided to the video quality estimation unit 150.

  As shown in the figure, steps S101 ′ and S102 and steps S103 and S104 may be executed in parallel.

  In step S105, the video quality estimation unit 150 calculates video quality based on the received long-time video encoding quality and the total communication degradation amount according to any of the above-described calculation methods, and outputs the calculated video quality evaluation value. To do. Thereafter, the video quality estimation apparatus 100 ′ ends the process.

  According to the present embodiment, the amount of quality deterioration with respect to the reproduction stop is calculated in consideration of the forgetting effect with fewer input parameters than the video quality estimation apparatus 100 of FIG. It is possible to estimate the overall video quality for the video viewing. This makes it possible to estimate the quality with high accuracy by using fewer input parameters as compared with calculation based on a simple average value or the like of the reproduction stop time.

[Acoustic quality estimation device]
With reference to FIGS. 7-10, the acoustic quality estimation apparatus by 2nd Example of this invention is demonstrated.

  FIG. 7 is a block diagram showing a configuration of an acoustic quality estimation apparatus according to an embodiment of the present invention. As illustrated in FIG. 7, the acoustic quality estimation apparatus 200 includes a short-time acoustic coding quality estimation unit 210, a long-time acoustic coding quality estimation unit 220, a single communication degradation amount estimation unit 230, and total communication degradation. A quantity estimation unit 240 and an acoustic quality estimation unit 250 are included.

The short-time acoustic coding quality estimation unit 210 estimates the short-time acoustic coding quality of the acoustic content. In one embodiment, the short-term acoustic coding quality estimation unit 210 estimates the short-term acoustic coding quality of the acoustic content based on the acoustic bit rate of the acoustic content, for example, the short-term acoustic coding based on the acoustic bit rate A_Br. The quality A_MOSC _S is calculated, and the calculated short-time acoustic coding quality A_MOSC _S is output to the long-time acoustic coding quality estimation unit 220.

  Here, the short-time acoustic encoding quality is an acoustic quality in consideration of deterioration due to encoding of acoustic data for a short time (for example, a predetermined period of about several seconds to several tens of seconds). Here, it is assumed that the short-time acoustic coding quality estimation unit 210 estimates the acoustic quality of the acoustic content for T seconds. The acoustic bit rate may be estimated and acquired from the acoustic data packet in the network or at the user terminal. The specific calculation method of the short-time acoustic coding quality in the short-time acoustic coding quality estimation unit 210 is calculated using, for example, the method defined in ITU-T recommendation P.1201.1 (Non-patent Document 3). can do.

  The long-term acoustic coding quality estimation unit 220 estimates the long-term acoustic coding quality based on the short-term acoustic coding quality output by the short-time acoustic coding quality estimation unit 210. The long-term acoustic coding quality is, for example, the acoustic coding quality of acoustic content of several minutes to several tens of minutes. Here, the long-term acoustic coding quality estimation unit 220

秒間の音響品質を推定するものとする（Nは１以上の整数）。長時間音響符号化品質A_MOSC_Lは、短時間音響符号化品質推定部２１０が所定の期間内に出力した各T秒間の音響品質値

The sound quality per second is estimated (N is an integer of 1 or more). The long-time acoustic coding quality A_MOSC _L is the acoustic quality value for each T seconds output by the short-time acoustic coding quality estimation unit 210 within a predetermined period.

のうち、直近のN個を単純に平均して、

Of these, simply average the last N,

により求めてもよい。

You may ask for.

FIG. 8 is a conceptual diagram related to estimation of short-term acoustic coding quality and long-term acoustic coding quality according to an embodiment of the present invention. As shown in FIG. 8, A_MOSC _S (1) is the recently output short-term acoustic coding quality, and A_MOSC _S (i) is obtained from A_MOSC _S (1) when there is no playback stop.

秒前に出力された短時間音響符号化品質である。

This is the short-time acoustic coding quality output in seconds.

  However, the long-term acoustic coding quality is not limited to the above formula, and the long-term acoustic coding quality estimation unit 220 takes into account the forgetting effect based on the short-term acoustic coding quality. Long-term acoustic coding quality may be estimated. For example, long-term audio coding quality takes into account the forgetting effect and quality evaluation characteristics for quality change (perceiving the change as degradation when the quality changes)

により求められてもよい。第１項は、聴いた直後のものはよく覚えているので影響は大きいが、時間が経過すると忘れていく忘却効果を示しており、第２項は、品質の変化が発生すると"変化"の絶対値が劣化として感じられ、それは聴いた直後であるほど強いことを表している。パラメータA₄,A₅は、予め定められた係数であり、サービス事業者が決めてよい。w₁(i)は重み関数であり、例えば、係数v30を用いて、

May be required. The first term shows the forgetting effect that is forgotten over time as it is well remembered immediately after listening, but the second term shows the “change” when quality changes. The absolute value is perceived as degradation, indicating that it is stronger immediately after listening. The parameters A ₄ and A ₅ are predetermined coefficients and may be determined by the service provider. w ₁ (i) is a weight function, for example, using the coefficient v30,

により計算できる。

Can be calculated by

The single communication degradation amount estimation unit 230 estimates a single communication degradation amount that is an acoustic quality degradation amount with respect to one stop of reproduction of the acoustic content. In one embodiment, the single communication deterioration amount estimation unit 230 is based on the reproduction stop time length RL that is the time length during which reproduction is stopped, and the single communication deterioration amount DR _S ( RL).

FIG. 9 is a conceptual diagram related to the stop of playback of the sound quality estimation apparatus according to an embodiment of the present invention. As shown in FIG. 9, the playback stop time is set to ST1, ST2,... From the side closer to the current time T _NOW , and the playback stop time length is also set to RL1, RL2,.

Single communication degradation amount DR _S (RL), for example, using a threshold Th3,

により算出できる。ここで、パラメータv31-33は、予め定められた係数であり、サービス事業者が決めてよい。上記のRL<Th3であるときは、再生停止を知覚してからある秒数（例えば、１秒程度）までは急激に品質が劣化し、その後Th3（例えばTh3=2秒）まではあまり品質が変化しないことを表している。他方、RL≧Th3であるときは、RLがTh3秒経過した後、緩やかに品質が低下していくことを表している。

Can be calculated. Here, the parameter v31-33 is a predetermined coefficient and may be determined by the service provider. When the above RL <Th3, the quality deteriorates rapidly until a certain number of seconds (for example, about 1 second) after perceiving the stop of playback, and then the quality is too low until Th3 (for example, Th3 = 2 seconds). Indicates that there is no change. On the other hand, when RL ≧ Th3, it indicates that the quality gradually decreases after RL has passed Th3 seconds.

However, the calculation of DR _S (RL) is not limited to the above formula, for example,

により算出されてもよい。

May be calculated.

  Also, without using the threshold Th3,

により算出されてもよい。ただし、パラメータv34-39は、v31-33と同様に予め定められた係数である。

May be calculated. However, the parameter v34-39 is a predetermined coefficient similarly to v31-33.

  The total communication degradation amount estimation unit 240 estimates a total communication degradation amount that is a quality degradation amount with respect to the stop of reproduction of the audio content, based on each single communication degradation amount estimated in a predetermined period. In one embodiment, the total communication degradation amount estimation unit 240 is a sequence of times when playback of audio content has been stopped.

と、単一通信劣化量推定部２３０が算出した単一通信劣化量の系列

And a sequence of single communication deterioration amounts calculated by the single communication deterioration amount estimation unit 230

とに基づき、所定の期間内に発生した再生停止に対する音響品質劣化量である総通信劣化量DR_Mを推定する。ここで、パラメータPは、所定の期間内に発生した再生停止の回数である。

Preparative the basis to estimate the total communication deterioration amount DR _M is an acoustic quality deterioration amount with respect to the reproduction stop that occurred within a predetermined time period. Here, the parameter P is the number of times playback has stopped within a predetermined period.

In one embodiment, the total communication degradation amount estimation unit 240 considers the forgetting effect based on the playback stop time length of each playback stop, the single communication degradation amount, and the playback stop time of the audio content in a predetermined period. The amount of deterioration may be estimated. For example, the total communication deterioration amount DR _M is

により算出されてもよい。ここで、w₂は重み関数であり、例えば、係数v41,v42を用いて、

May be calculated. Here, w ₂ is a weight function, for example, using coefficients v41 and v42,

などにより算出できる。

It can be calculated by

  However, the weight function is not limited to the above formula, and for example, using a constant v43 that is greater than 0 and equal to or less than 1,

として算出されてもよい。重み関数の積は、何回も前の再生停止や、発生後に時間が経過している再生停止は、最近発生した再生停止よりも品質に与える影響が小さいこと（忘却効果）を表している。

May be calculated as The product of the weighting functions indicates that the playback stop many times before and the playback stop that has passed after the occurrence have less influence on the quality than the recently generated playback stop (forgetting effect).

Further, the calculation of the total communication deterioration amount is not limited to the above formula, for example, using the current time T _NOW ,

により算出されてもよい。w₃は、w₂と同様の重み関数であり、忘却効果を表している。

May be calculated. w ₃ is a weight function similar to w ₂ and represents the forgetting effect.

The sound quality estimation unit 250 estimates the sound quality of the sound content based on the long-time sound coding quality and the total communication deterioration amount. In one embodiment, the acoustic quality estimation unit 250 includes the long-term acoustic encoding quality A_MOSC _L calculated by the long-term acoustic encoding quality estimation unit 220 and the total communication degradation amount DR _M calculated by the total communication degradation amount estimation unit 240. Based on the above, the acoustic quality A_MOS is estimated. The acoustic quality A_MOS is, for example,

により算出できる。ここで、パラメータA₄は、予め定められた係数であり、サービス提供者が決めてよい。

Can be calculated. Here, the parameter A ₄ is a predetermined coefficient and may be determined by the service provider.

However, the calculation of the acoustic quality A_MOS is not limited to the above formula, for example, using the coefficient A ₅ ,

により算出されてもよい。

May be calculated.

  FIG. 10 is a flowchart showing processing in the sound quality estimation apparatus according to the embodiment of the present invention. As shown in FIG. 10, the process is started in response to the input of the acoustic bit rate, the playback stop time length, and the playback stop time.

  In step S201, the short-term acoustic coding quality estimation unit 210 calculates the short-term acoustic coding quality from the input acoustic bit rate according to any of the calculation methods described above, and the calculated short-term acoustic coding quality. Is provided to the long-term acoustic coding quality estimation unit 220.

  In step S202, the long-term acoustic coding quality estimation unit 220 estimates the long-term acoustic coding quality based on the short-term acoustic coding quality received according to any of the calculation methods described above, and the estimated long-term acoustic code. The quality is provided to the acoustic quality estimation unit 250.

  In step S203, the single communication deterioration amount estimation unit 230 calculates a single communication deterioration amount for each reproduction stop based on the input reproduction stop time length according to any of the calculation methods described above, and calculates the calculated single unit. The communication deterioration amount is provided to the total communication deterioration amount estimation unit 240.

  In step S204, the total communication deterioration amount estimation unit 240 calculates the total communication deterioration amount based on the received single communication deterioration amount, the input reproduction stop time length, and the reproduction stop time according to any of the calculation methods described above. And provided to the acoustic quality estimation unit 250.

  As shown in the figure, steps S201 and S202 and steps S203 and S204 may be executed in parallel.

  In step S205, the acoustic quality estimation unit 250 calculates the acoustic quality based on the received long-time acoustic encoding quality and the total communication degradation amount according to any of the calculation methods described above, and outputs the calculated acoustic quality evaluation value. To do. Thereafter, the sound quality estimation apparatus 200 ends the process.

  According to the present embodiment, the amount of quality degradation due to playback stop is calculated in consideration of the forgetting effect, and the overall acoustic quality estimation for the long-time acoustic listening is performed together with the long-term acoustic coding quality. Can do. Thereby, it is possible to estimate the sound quality with higher accuracy than the calculation based on a simple average value or the like of the reproduction stop time.

[AV quality estimation device]
An AV quality estimation apparatus according to the third embodiment of the present invention will be described with reference to FIGS.

  FIG. 11 is a block diagram showing a configuration of an AV quality estimation apparatus according to an embodiment of the present invention. As shown in FIG. 11, the AV quality estimation apparatus 300 includes a short-time AV coding quality estimation unit 310, a long-time AV coding quality estimation unit 320, a single communication degradation amount estimation unit 330, and total communication degradation. A quantity estimating unit 340 and an AV quality estimating unit 350; The short time AV coding quality estimation unit 310 includes a short time audio coding quality estimation unit 311, a short time video coding quality estimation unit 312, and a short time AV coding quality calculation unit 313.

The short-time AV coding quality estimation unit 310 estimates the short-time AV coding quality of the AV content. In one embodiment, the short-time AV coding quality estimation unit 310 performs the short-time AV coding quality AV_MOSC _S based on the audio bit rate A_Br, the video resolution Rs, the video frame rate Fr, the video bit rate Br, and the video feature quantity Ir. And the short-term AV coding quality AV_MOSC _S is output to the long-term AV coding quality estimation unit 320.

  Here, the short-time AV coding quality is AV quality considering degradation due to coding of audio data and video data in a short time (for example, a predetermined period of about several seconds to several tens of seconds). Here, it is assumed that short-time AV encoding quality estimation section 310 estimates the AV quality of AV content for T seconds.

  The short-time acoustic coding quality estimation unit 311 is the same as the short-term acoustic coding quality estimation unit 210 of the acoustic quality estimation apparatus 200, and the short-term acoustic coding quality of the audio data of the AV content based on the acoustic bit rate. Is estimated. Further details are omitted here to avoid duplication of explanation.

  The short-time video encoding quality estimation unit 312 is the same as the short-time video encoding quality estimation unit 110 of the video quality estimation apparatus 100, and based on the video resolution, video frame rate, video bit rate, and video feature amount, Estimate short-time video coding quality of video data of AV content. Alternatively, in another embodiment, the short-time video coding quality estimation unit 312 may be the same as the short-time video coding quality estimation unit 110 ′ of the video quality estimation device 100 ′, and the video resolution, video Based on the frame rate and the video bit rate, the short-time video encoding quality of the video data of the AV content may be estimated.

  The short time AV coding quality calculation unit 313 estimates the short time AV coding quality of the AV content based on the estimated short time audio coding quality and short time video coding quality. The short-term AV encoding quality can be calculated by, for example, a method defined in ITU-T recommendation P.1201.1 (Non-patent document 3) or P.1201.2 (Non-patent document 4).

  The long-time AV coding quality estimation unit 320 estimates the long-time AV coding quality of the AV content based on the short-time AV coding quality. In one embodiment, the long-time AV encoding quality estimation unit 320 estimates the long-term AV encoding quality of the AV content based on the short-time AV encoding quality output by the short-time AV encoding quality estimation unit 310.

  Here, the long-time AV encoding quality is, for example, the AV encoding quality of AV content in a predetermined period of several minutes to several tens of minutes. Here, the long-term AV encoding quality estimation unit 320

秒間のAV品質を推定するものとする（Nは１以上の整数）。例えば、長時間AV符号化品質AV_MOSC_Lは、短時間AV符号化品質推定部３１０が所定の期間内に出力した各T秒間のAV品質値

Assume that the AV quality per second is estimated (N is an integer of 1 or more). For example, the long-term AV coding quality AV_MOSC _L is the AV quality value for each T seconds output by the short-time AV coding quality estimation unit 310 within a predetermined period.

のうち、直近のN個を単純に平均して、

Of these, simply average the last N,

により求められてもよい。

May be required.

FIG. 12 is a conceptual diagram regarding short-time AV coding quality and long-time AV coding quality estimation according to an embodiment of the present invention. As shown in FIG. 12, AV_MOSC _S (1) is the short-term AV encoding quality output recently, and AV_MOSC _S (i) is calculated from AV_MOSC _S (1) when there is no playback stop.

秒前に出力された短時間AV符号化品質である。長時間AV符号化品質推定部３２０は、上記の式に限定されず、短時間AV符号化品質に基づき、忘却効果を考慮してAVコンテンツの長時間AV符号化品質を推定してもよい。例えば、長時間AV符号化品質は、忘却効果や品質変化に対する品質評価特性（品質が変化した時にその変化を劣化として知覚する）を考慮して、

This is the short-time AV encoding quality output a second ago. The long-term AV encoding quality estimation unit 320 is not limited to the above formula, and may estimate the long-term AV encoding quality of the AV content in consideration of the forgetting effect based on the short-term AV encoding quality. For example, the long-time AV coding quality takes into account the forgetting effect and quality evaluation characteristics for quality change (perceiving the change as deterioration when the quality changes)

により求められてもよい。第１項は、視聴直後のものはよく覚えているので影響は大きいが、時間が経過すると忘れていく忘却効果を示しており、第２項は、品質の変化が発生すると"変化"の絶対値が劣化として感じられ、それは視聴直後であるほど強いことを表している。パラメータA₄,A₅は、予め定められた係数であり、サービス事業者が決めてよい。w₁(i)は重み関数であり、例えば、係数v30を用いて、

May be required. The first term shows the forgetting effect that is forgotten for a long time since it is well remembered immediately after viewing, but the second term is the absolute “change” when a change in quality occurs The value is perceived as degraded, indicating that it is stronger immediately after viewing. The parameters A ₄ and A ₅ are predetermined coefficients and may be determined by the service provider. w ₁ (i) is a weight function, for example, using the coefficient v30,

により計算できる。

Can be calculated by

The single communication deterioration amount estimation unit 330 estimates a single communication deterioration amount that is an AV quality deterioration amount with respect to one stop of playback of AV content. In one embodiment, the single communication deterioration amount estimation unit 330 is based on the reproduction stop time length RL that is the time length when the video is stopped, and the single communication deterioration amount DR _S ( RL).

FIG. 13 is a conceptual diagram regarding playback stop of an AV quality estimation apparatus according to an embodiment of the present invention. As shown in FIG. 13, the playback stop time is set to ST1, ST2,... From the side closer to the current time T _NOW , and the playback stop time length is also set to RL1, RL2,. Single communication degradation amount DR _S (RL) is
For example, using the threshold Th3,

により算出できる。ここで、パラメータv31-33は、予め定められた係数であり、サービス事業者が決めてよい。上記のRL<Th3であるときは、再生停止を知覚してからある秒数（例えば、１秒程度）までは急激に品質が劣化し、その後Th3（例えば、Th3=2秒）まではあまり品質が変化しないことを表している。他方、RL≧Th3であるときは、RLがTh3秒経過した後、緩やかに品質が低下していくことを表している。DR_S (RL)の算出は、上記の式に限定されず、例えば、

Can be calculated. Here, the parameter v31-33 is a predetermined coefficient and may be determined by the service provider. When RL <Th3, the quality deteriorates rapidly until a certain number of seconds (eg, about 1 second) after the playback stop is perceived, and then the quality is too low until Th3 (eg, Th3 = 2 seconds). Represents no change. On the other hand, when RL ≧ Th3, it indicates that the quality gradually decreases after RL has passed Th3 seconds. The calculation of DR_S (RL) is not limited to the above formula, for example,

により算出されてもよく、Th3を用いることなく、

Without using Th3,

を用いて算出されてもよい。ただし、パラメータv34-39は、v31-33と同様に予め定められた係数である。

It may be calculated using However, the parameter v34-39 is a predetermined coefficient similarly to v31-33.

  The total communication degradation amount estimation unit 340 estimates a total communication degradation amount, which is a quality degradation amount with respect to the stop of playback of AV content, based on the single communication degradation amount estimated in a predetermined period. In one embodiment, the total communication degradation amount estimation unit 340 includes a sequence of times at which playback stops.

と、単一通信劣化量推定部３３０が算出した単一通信劣化量の系列

And a series of single communication deterioration amount calculated by the single communication deterioration amount estimation unit 330

とに基づき、所定の期間内に発生した再生停止に対するAV品質劣化量である総通信劣化量DR_Mを推定する。ここで、パラメータPは、所定の期間内に発生した再生停止の回数である。

Preparative the basis to estimate the total communication deterioration amount DR _M is an AV quality deterioration amount with respect to the reproduction stop that occurred within a predetermined time period. Here, the parameter P is the number of times playback has stopped within a predetermined period.

In one embodiment, the total communication degradation amount estimation unit 340 considers the forgetting effect based on the playback stop time length of each AV content playback stop, the single communication degradation amount, and the playback stop time in a predetermined period. The amount of deterioration may be estimated. For example, the total communication deterioration amount DR _M is

により算出できる。ここで、w₂は重み関数であり、例えば、係数v41,v42を用いて、

Can be calculated. Here, w ₂ is a weight function, for example, using coefficients v41 and v42,

などとして算出できる。しかしながら、重み関数は上記の式に限定されず、例えば、０より大きく１以下の定数v43を用いて、

And so on. However, the weight function is not limited to the above formula, and for example, using a constant v43 that is greater than 0 and less than or equal to 1,

として算出されてもよい。重み関数の積は、何回も前の再生停止や、発生後に時間が経過している再生停止は、最近発生した再生停止よりも品質に与える影響が小さいこと（忘却効果）を表している。

May be calculated as The product of the weighting functions indicates that the playback stop many times before and the playback stop that has passed after the occurrence have less influence on the quality than the recently generated playback stop (forgetting effect).

Further, the calculation of the total communication deterioration amount is not limited to the above formula, for example, using the current time T _NOW ,

として算出されてもよい。w₃は、w₂と同様の重み関数であり、忘却効果を表している。

May be calculated as w ₃ is a weight function similar to w ₂ and represents the forgetting effect.

The AV quality estimation unit 350 estimates the AV quality of the AV content based on the long-time AV encoding quality and the total communication deterioration amount. In one embodiment, the AV quality estimation unit 350 includes the long-time AV coding quality AV_MOSC _L calculated by the long-time AV coding quality estimation unit 320 and the total communication degradation amount DR _M calculated by the total communication degradation amount estimation unit 340. Based on the above, the AV quality AV_MOS is estimated. AV quality AV_MOS is, for example,

により算出できる。ここで、パラメータA₄は、予め定められた係数であり、サービス提供者が決めてよい。しかしながら、映像品質AV_MOSの算出は、上記の式に限定されず、例えば、係数A₅を用いて、

Can be calculated. Here, the parameter A ₄ is a predetermined coefficient and may be determined by the service provider. However, the calculation of the video quality AV_MOS is not limited to the above formula, for example, using the coefficient A ₅ ,

により算出されてもよい。

May be calculated.

  FIG. 14 is a flowchart showing processing in the AV quality estimation apparatus according to an embodiment of the present invention. As shown in FIG. 14, the processing is started in response to input of an audio bit rate, video resolution, video frame rate, video bit rate, video feature amount, playback stop time length, and playback stop time. Note that when the short-time video coding quality estimation unit 311 is the same as the short-time video coding quality estimation unit 110 ′, it is not necessary to input a video feature amount.

  In step S301, the short-time video encoding quality estimation unit 311 calculates the short-time video encoding quality from the input video resolution, video frame rate, video bit rate, and video feature amount according to any of the calculation methods described above. The short-time video coding quality is calculated and provided to the short-time AV coding quality calculation unit 313.

  In step S302, the short-term acoustic coding quality estimation unit 312 calculates the short-term acoustic coding quality from the input acoustic bit rate according to any of the calculation methods described above, and the calculated short-term acoustic coding quality. Is provided to the short-time AV encoding quality calculation unit 313.

  In step S303, the short time AV coding quality calculation unit 313 performs a short time according to any of the calculation methods described above from the short time video coding quality and the short time sound coding quality estimated in steps S301 and S302. The AV coding quality is calculated, and the calculated short-time AV coding quality is provided to the long-time AV coding quality estimation unit 320.

  In step S304, the long-time AV coding quality estimation unit 320 estimates and estimates the long-time AV coding quality from the short-time AV coding quality calculated in step S303 according to any of the calculation methods described above. The long-term AV encoding quality is provided to the AV quality estimation unit 350.

  In step S305, the single communication deterioration amount estimation unit 330 calculates a single communication deterioration amount for each reproduction stop based on the input reproduction stop time length according to any of the calculation methods described above, and calculates the calculated single communication deterioration amount. The communication deterioration amount is provided to the total communication deterioration amount estimation unit 340.

  In step S306, the total communication deterioration amount estimation unit 340 calculates the total communication deterioration amount based on the received single communication deterioration amount, the input reproduction stop time length, and the reproduction stop time according to any of the calculation methods described above. Then, the calculated total communication deterioration amount is provided to the AV quality estimation unit 350.

  As illustrated, steps S301 to S304 and steps S305 and S306 may be executed in parallel.

  In step S307, the AV quality estimation unit 350 calculates the AV quality based on the received long-term AV encoded quality and the total communication degradation amount according to any of the above-described calculation methods, and outputs the calculated AV quality evaluation value. To do. Thereafter, the AV quality estimation apparatus 300 ends the process.

  According to the present embodiment, the amount of quality degradation due to playback stop is calculated in consideration of the forgetting effect, and comprehensive AV quality estimation for long-time AV viewing is performed together with long-time AV coding quality. Can do. This makes it possible to estimate the AV quality with higher accuracy than the calculation based on a simple average value of the reproduction stop time.

[Modification of AV quality estimation device]
FIG. 15 is a block diagram showing a configuration of an AV quality estimation apparatus according to another embodiment of the present invention. The AV quality estimation apparatus according to the present embodiment does not estimate the short-time AV coding quality, and the long-time video coding quality and the long-time acoustic coding quality from the short-time video coding quality and the short-time acoustic coding quality, respectively. And the long-term AV coding quality is estimated based on the estimated long-time video coding quality and long-time acoustic coding quality.

  As shown in FIG. 15, the AV quality estimation apparatus 400 includes a short-time AV coding quality estimation unit 410, a long-time AV coding quality estimation unit 420, a single communication degradation amount estimation unit 430, and total communication degradation. A quantity estimation unit 440 and an AV quality estimation unit 450. The short time AV coding quality estimation unit 410 includes a short time audio coding quality estimation unit 411 and a short time video coding quality estimation unit 412, and the long time AV coding quality estimation unit 420 includes a long time AV coding quality estimation unit 412. The audio coding quality estimation unit 421, the long-time video coding quality estimation unit 422, and the long-time AV coding quality calculation unit 423 are included. That is, the short time AV coding quality estimation unit 410 is different from the short time AV coding quality estimation unit 310 of FIG. 11 in that it does not include the short time AV coding quality calculation unit 313. The quality estimation unit 420 includes a long-time audio coding quality estimation unit 421, a long-time video coding quality estimation unit 422, and a long-time AV coding quality calculation unit 423. Different from the quality estimation unit 320.

The short time AV coding quality estimation unit 410 estimates the short time audio coding quality A_MOSC _S and the short time video coding quality V_MOSC _S in the same manner as the short time AV coding quality estimation unit 310 of FIG. The short time AV coding quality estimation unit 410 does not calculate the short time AV coding quality A_MOSC _S from the estimated short time sound coding quality A_MOSC _S and the short time video coding quality V_MOSC _S. The short-time video coding quality V_MOSC _S is output to the long-time AV coding quality estimation unit 420.

  The long-time audio coding quality estimation unit 420 is similar to the procedure for estimating the long-time audio coding quality from the short-time sound coding quality in the sound quality estimation apparatus 200 of FIG. Is used to estimate the long-term audio encoding quality from the short-time audio encoding quality, and the video quality estimation apparatus 100 in FIG. 1 or the video quality estimation apparatus 100 ′ in FIG. Similar to the procedure for estimating the coding quality, the long-time video coding quality estimation unit 422 is used to estimate the long-time video coding quality from the short-time video coding quality.

The long-time AV coding quality estimation unit 423 calculates the long-time AV coding quality based on the received long-time audio coding quality and long-time video coding quality. For a specific calculation method of long-term AV coding quality, for example, AV quality is calculated from the sound quality and video quality in ITU-T Recommendations P.1201.1 (Non-Patent Document 3) and P.1201.2 (Non-Patent Document 4). A method similar to the method of performing may be used. According to the method described in ITU-T recommendation P.1201.1, the long-term AV encoding quality AV_MOSC _L is

により算出される。ここで、パラメータv71-74は、予め定められた係数であり、サービス提供者が決めてよい。

Is calculated by Here, the parameter v71-74 is a predetermined coefficient and may be determined by the service provider.

  The single communication degradation amount estimation unit 430, the total communication degradation amount estimation unit 440, and the AV quality estimation unit 450 are the same as those of the AV quality estimation apparatus 300, and thus description thereof is omitted.

  FIG. 16 is a flowchart showing processing in the AV quality estimation apparatus according to another embodiment of the present invention.

  As shown in FIG. 16, the process is started in response to input of an audio bit rate, video resolution, video frame rate, video bit rate, video feature amount, playback stop time length, and playback stop time. Note that when the short-time video coding quality estimation unit 412 is the same as the short-time video coding quality estimation unit 110 ′, it is not necessary to input video feature quantities.

  In step S401, the short-time video coding quality estimation unit 412 calculates the short-time video coding quality from the input video resolution, video frame rate, video bit rate, and video feature amount according to any of the calculation methods described above. The calculated short time video coding quality is provided to the long time AV coding quality estimation unit 420.

  In step S402, the long-time video coding quality estimation unit 422 estimates and estimates the long-time video coding quality from the short-time video coding quality calculated in step S401 according to any of the calculation methods described above. The long-time video coding quality is provided to the long-time AV coding quality calculation unit 423.

  In step S403, the short-term acoustic coding quality estimation unit 411 calculates the short-term acoustic coding quality from the input acoustic bit rate according to any of the calculation methods described above, and the calculated short-term acoustic coding quality. Is provided to the long-term AV encoding quality estimation unit 420.

  In step S404, the long-term acoustic coding quality estimation unit 421 estimates and estimates the long-term acoustic coding quality from the short-term acoustic coding quality calculated in step S403 according to any of the calculation methods described above. The long-term audio coding quality is provided to the long-time AV coding quality calculation unit 423.

  In step S405, the long-time AV coding quality calculation unit 423 calculates the long-time AV coding quality from the received long-time video coding quality and long-time audio coding quality according to any of the calculation methods described above. Provided to the AV quality estimation unit 450.

  In step S406, the single communication deterioration amount estimation unit 430 calculates a single communication deterioration amount for each reproduction stop based on the input reproduction stop time length according to any of the calculation methods described above, and calculates the calculated single unit. The communication degradation amount is provided to the total communication degradation amount estimation unit 440.

  In step S407, the total communication deterioration amount estimation unit 440 calculates the total communication deterioration amount based on the received single communication deterioration amount, the input reproduction stop time length, and the reproduction stop time according to any of the above-described calculation methods. Then, the calculated total communication deterioration amount is provided to the AV quality estimation unit 450.

  As illustrated, steps S401 to S405 and steps S406 and S407 may be executed in parallel.

  In step S408, the AV quality estimation unit 450 calculates the AV quality based on the received long-time AV encoded quality and the total communication degradation amount according to any of the calculation methods described above, and outputs the calculated AV quality evaluation value. To do. Thereafter, the AV quality estimation apparatus 400 ends the process.

  According to the present embodiment, the amount of quality degradation due to playback stop is calculated in consideration of the forgetting effect, and comprehensive AV quality estimation for long-time AV viewing is performed together with long-time AV coding quality. Can do. This makes it possible to estimate the AV quality with higher accuracy than the calculation based on a simple average value of the reproduction stop time.

FIG. 17 is a block diagram showing a configuration of an AV quality estimation apparatus according to another embodiment of the present invention. The AV quality estimation apparatus according to the present embodiment estimates the short-time AV encoded quality AV_MOSC _S based on the AV bit rate obtained by combining the video bit rate and the audio bit rate.

  As shown in FIG. 17, the AV quality estimation apparatus 500 includes a short-time AV coding quality estimation unit 510, a long-time AV coding quality estimation unit 520, a single communication degradation amount estimation unit 530, and total communication degradation. A quantity estimation unit 540 and an AV quality estimation unit 550; The AV quality estimation apparatus 500 according to this embodiment differs from the AV quality estimation apparatus 300 shown in FIG. 11 in that the short-time AV coding quality is obtained from the input AV bit rate. The short-term AV coding quality calculation method in the estimation unit 510 is different from the short-time AV coding quality estimation unit 310 shown in FIG. Therefore, the processing of the short-time AV coding quality estimation unit 510 will be described, and the long-time AV coding quality estimation unit 520, the single communication degradation amount estimation unit 530, the total communication degradation amount estimation unit 540, and the video quality estimation unit 550. Is the same as that of the AV quality estimation apparatus 300, and thus the description thereof is omitted.

  Based on the AV bit rate AV_Br, the short-time AV coding quality estimation unit 510 estimates the short-time AV coding quality using, for example, the following formula, and uses the estimated short-time AV coding quality as the long-time AV coding quality estimation. Part 520.

ここで、パラメータv80-83は、予め定められた係数であり、サービス事業者が決めてよい。

Here, the parameter v80-83 is a predetermined coefficient and may be determined by the service provider.

  In addition, the short-term AV encoding quality estimation is not limited to the above formula, for example,

により推定されてもよい。ここで、パラメータv84-87は、予め定められた係数であり、サービス事業者が決めてよい。

May be estimated. Here, the parameter v84-87 is a predetermined coefficient and may be determined by the service provider.

  FIG. 18 is a flowchart showing processing in the AV quality estimation apparatus according to another embodiment of the present invention. As shown in FIG. 18, the process is started in response to the input of the AV bit rate, the playback stop time length, and the playback stop time.

  In step S501, the short time AV coding quality estimation unit 511 calculates the short time AV coding quality from the input AV bit rate according to any of the calculation methods described above, and calculates the calculated short time AV coding quality. Is provided to the long-term AV encoding quality estimation unit 520.

  In step S502, the long-term AV encoding quality estimation unit 520 estimates and estimates the long-term AV encoding quality from the short-term AV encoding quality calculated in step S501 according to any of the above-described calculation methods. The long-term AV encoding quality is provided to the AV quality estimation unit 550.

  In step S503, the single communication deterioration amount estimation unit 530 calculates a single communication deterioration amount for each reproduction stop based on the input reproduction stop time length according to any of the calculation methods described above, and calculates the calculated single unit. The communication degradation amount is provided to the total communication degradation amount estimation unit 540.

  In step S504, the total communication deterioration amount estimation unit 540 calculates the total communication deterioration amount based on the received single communication deterioration amount, the input playback stop time length, and the playback stop time according to any of the above-described calculation methods. Then, the calculated total communication deterioration amount is provided to the AV quality estimation unit 550.

  As illustrated, steps S501 and S502 and steps S503 and S504 may be executed in parallel.

  In step S505, the AV quality estimation unit 550 calculates the AV quality based on the received long-time AV encoded quality and the total communication degradation amount according to any of the calculation methods described above, and outputs the calculated AV quality evaluation value. To do. Thereafter, the AV quality estimation apparatus 500 ends the process.

  According to the present embodiment, the amount of quality degradation due to playback stop is calculated in consideration of the forgetting effect, and comprehensive AV quality estimation for long-time AV viewing is performed together with long-time AV coding quality. Can do. This makes it possible to estimate the AV quality with higher accuracy than the calculation based on a simple average value of the reproduction stop time. Further, since the short-time AV encoding quality is estimated by a relatively simple procedure based on the AV bit rate, high-speed processing is possible.

  The video quality estimation device, the sound quality estimation device, and the AV quality estimation device described above typically include various hardware resources such as an auxiliary storage device, a memory device, a CPU (Central Processing Unit), a communication device, and an interface device. Composed. The auxiliary storage device is composed of a hard disk, a flash memory, and the like, and stores programs and data for realizing various processes in the device. These programs and data may be downloaded from various servers on the network, for example, or may be provided from a storage medium connected to the interface of the device. The memory device is composed of a RAM (Random Access Memory) or the like, and reads and stores the program from the auxiliary storage device when an instruction to start the program is given. The CPU functions as a processor for processing information, and implements the various functions described above in the device by executing a program stored in the memory device. The communication device includes various communication circuits for connecting to an external device such as a server via a network. The interface device realizes an interface function for exchanging data and information with an external device. Note that the video quality estimation device, the audio quality estimation device, and the AV quality estimation device according to the present invention are not limited to the hardware configuration described above, but any other such as including circuits that implement various functions described below. It may have an appropriate hardware configuration.

  As mentioned above, although the Example of this invention was explained in full detail, this invention is not limited to the specific embodiment mentioned above, In the range of the summary of this invention described in the claim, it is various. Deformation / change is possible.

100, 100 'Video quality estimation device 110, 110' Short-time video coding quality estimation unit 120 Long-time video coding quality estimation unit 130 Single communication degradation amount estimation unit 140 Total communication degradation amount estimation unit 150 Video quality estimation unit 200 Acoustic quality estimation device 210 Short-time acoustic coding quality estimation unit 220 Long-time acoustic coding quality estimation unit 230 Single communication degradation amount estimation unit 240 Total communication degradation amount estimation unit 250 Acoustic quality estimation unit 300, 400, 500 AV quality estimation Apparatus 310, 410, 510 Short-time AV coding quality estimation unit 320, 420, 520 Long-time AV coding quality estimation unit 330, 430, 530 Single communication degradation amount estimation unit 340, 440, 540 Total communication degradation amount estimation unit 350, 450, 550 AV quality estimation unit

Claims (11)

A video quality estimation device for estimating video quality of video content,
A short time video coding quality estimation unit for estimating a short time video coding quality of the video content;
A long-time video coding quality estimation unit for estimating a long-time video coding quality of the video content based on the short-time video coding quality;
A single communication deterioration amount estimation unit that estimates a single communication deterioration amount that is a video quality deterioration amount with respect to one stop of playback of the video content;
A total communication deterioration amount estimation unit that estimates a total communication deterioration amount that is a quality deterioration amount with respect to the reproduction stop of the video content based on the single communication deterioration amount in a predetermined period;
A video quality estimation unit that estimates the video quality of the video content based on the long-time video encoding quality and the total communication degradation amount;
A video quality estimation apparatus. The single communication deterioration amount estimation unit estimates the single communication deterioration amount based on a reproduction stop time length of the reproduction stop of the video content,
The total communication degradation amount estimation unit takes into account the forgetting effect based on the playback stop time length of each playback stop of the video content, the single communication degradation amount, and the playback stop time in the predetermined period. The video quality estimation apparatus according to claim 1, wherein the degradation amount is estimated.   The short time video coding quality estimation unit estimates a short time video coding quality of the video content based on a video resolution, a video frame rate, and a video bit rate of the video content. Video quality estimation device.   The video quality estimation apparatus according to claim 3, wherein the short-time video coding quality estimation unit further estimates a short-time video coding quality of the video content based on a video feature amount representing a feature of the video content.   The long-time video coding quality estimation unit estimates the long-time video coding quality of the video content in consideration of a forgetting effect based on the short-time video coding quality. The video quality estimation apparatus according to one item. A quality estimation method in an apparatus for estimating video quality of video content,
Estimating a short time video coding quality of the video content;
Estimating the long-term video coding quality of the video content based on the short-time video coding quality;
Estimating a single communication degradation amount which is a video quality degradation amount for one stop of playback of the video content;
Estimating a total communication deterioration amount, which is a quality deterioration amount with respect to suspension of reproduction of the video content, based on the single communication deterioration amount in a predetermined period;
Estimating the video quality of the video content based on the long-time video encoding quality and the total communication degradation amount;
A quality estimation method comprising: An acoustic quality estimation device for estimating the acoustic quality of acoustic content,
A short-time acoustic coding quality estimation unit that estimates a short-time acoustic coding quality of the acoustic content;
A long-term acoustic coding quality estimation unit that estimates the long-term acoustic coding quality of the acoustic content based on the short-time acoustic coding quality;
A single communication degradation amount estimation unit that estimates a single communication degradation amount that is an acoustic quality degradation amount for one stop of reproduction of the acoustic content;
A total communication deterioration amount estimation unit that estimates a total communication deterioration amount that is a quality deterioration amount with respect to the reproduction stop of the acoustic content, based on the single communication deterioration amount in a predetermined period;
An acoustic quality estimation unit that estimates the acoustic quality of the acoustic content based on the long-time acoustic encoding quality and the total communication degradation amount;
An acoustic quality estimation apparatus having A quality estimation method in an apparatus for estimating the acoustic quality of acoustic content,
Estimating a short-time acoustic coding quality of the acoustic content;
Estimating the long-term acoustic coding quality of the acoustic content based on the short-time acoustic coding quality;
Estimating a single communication degradation amount that is an acoustic quality degradation amount for one stop of reproduction of the acoustic content;
Estimating a total communication deterioration amount, which is a quality deterioration amount with respect to the reproduction stop of the audio content, based on the single communication deterioration amount in a predetermined period;
Estimating the acoustic quality of the acoustic content based on the long-time acoustic encoding quality and the total communication degradation amount;
A quality estimation method comprising: An AV quality estimation device for estimating AV quality of audiovisual (AV) content,
A short-time AV encoding quality estimation unit that estimates short-time AV encoding quality of the AV content;
Based on the short-time AV coding quality, a long-time AV coding quality estimation unit that estimates the long-time AV coding quality of the AV content;
A single communication deterioration amount estimation unit that estimates a single communication deterioration amount that is an AV quality deterioration amount with respect to one stop of playback of the AV content;
A total communication degradation amount estimation unit that estimates a total communication degradation amount, which is a quality degradation amount with respect to suspension of playback of the AV content, based on the single communication degradation amount in a predetermined period;
AV quality estimation unit that estimates the AV quality of the AV content based on the long-time AV encoding quality and the total communication degradation amount;
AV quality estimation apparatus having A quality estimation method in an apparatus for estimating AV quality of AV content,
Estimating a short time AV encoding quality of the AV content;
Estimating the long-term AV coding quality of the AV content based on the short-time AV coding quality;
Estimating a single communication deterioration amount that is an AV quality deterioration amount for one stop of playback of the AV content;
Estimating a total communication deterioration amount, which is a quality deterioration amount with respect to the reproduction stop of the AV content, based on the single communication deterioration amount in a predetermined period;
Estimating the AV quality of the AV content based on the long-time AV encoding quality and the total communication degradation amount;
A quality estimation method comprising:   A program for implementing the quality estimation method according to claim 6, claim 8, or claim 10, wherein the program causes a processor to execute the step of the quality estimation method.

Images