JP4796019B2 - Video quality estimation apparatus, method, and program - Google Patents

Description

  The present invention relates to a video communication technique, and more particularly to a video quality estimation technique for estimating video quality related to subjective video quality perceived by a viewer when video media encoded into a plurality of frames is received and reproduced by a terminal.

  With the increase in speed and bandwidth of Internet access lines, the spread of video communication services for transferring video media including video and audio between terminals or between servers and terminals via the Internet is expected.

  In a best effort network such as the Internet used for video communication services, communication quality is not always guaranteed. For this reason, when streaming content with temporal continuity such as video media is transferred over the Internet, if the bandwidth of the communication line is narrow or the communication line is congested, it is received and played back via the communication line. It is easily perceived as the quality perceived by the viewer with respect to the video media, that is, the deterioration of subjective video quality. Specifically, when quality degradation is added to video media, it is perceived as blurring, blurring, mosaic distortion, and jerky feeling.

  In this way, in video communication services that transfer video media, quality degradation is easily perceived, and in order to provide video communication services with good quality, quality design of the network prior to service provision and quality after service start Management is important. Therefore, there is a need for a video quality evaluation technique that can appropriately express the video quality enjoyed by the viewer and that is simple and efficient.

  Conventionally, as a technique for estimating the quality of video media, a video quality objective evaluation method using a video signal as an input has been proposed as a recommendation (see, for example, Non-Patent Document 1). Further, based on VoIP (Voice over IP), a technique for estimating an average conversation quality of an IP telephone from a quality parameter related to voice quality such as packet loss and delay time of voice packets has been proposed as a recommendation ( For example, refer nonpatent literature 2). Also, based on the encoding bit rate indicating the number of encoded bits per unit time for the video media, the frame rate indicating the number of frames per unit time, and the packet loss rate, the subjective feeling that the viewer feels using a predetermined estimation model A technique for estimating video quality has been proposed as a recommendation (see, for example, Non-Patent Document 3).

ITU-T Recommendation J.I. 144 ITU-T Recommendation G. 107 ITU-T Recommendation G. 1070

However, in such a conventional technique, when estimating the video quality, since the change in the video quality due to the compression encoding complexity of the video is not taken into consideration, there is a problem that the individual video quality cannot be estimated. .
In this type of video communication service, in order to improve the transfer efficiency of video media, the video media is compressed into multiple frames using the autocorrelation between pixels or images and the visual characteristics of humans. A compression-encoding communication method is used in which the data is transferred after being converted into a data.

  In a general compression encoding process, since the compression encoding efficiency changes depending on the complexity of the video to be encoded, the subjective video quality also changes. For example, in DCT (Discrete Cosine Transform) conversion processing widely used for video compression coding, the original image signal is decomposed into a two-dimensional spatial frequency distribution in the horizontal and vertical directions on the image, and the image quality is improved. The image is efficiently compressed by removing high-frequency components that do not affect much.

  Therefore, such compression encoding processing differs in compression efficiency according to the complexity of the compression encoding processing, that is, the compression encoding complexity. An image with a small change in pixel value between adjacent pixels, such as a natural image, that is, an image with a low compression encoding complexity can be efficiently compressed because there is a tendency to reduce the high-frequency AC component, but the pixel value change between adjacent pixels is large. An image, that is, an image having a high compression encoding complexity, cannot be efficiently compressed because a high-frequency AC component tends to increase.

At this time, for example, when the encoding bit rate of the video media is low, high compression efficiency is required, so it is necessary to remove even a relatively low frequency AC component. For this reason, when the amount of AC component to be removed is large, the roughness of the video is affected, and as a result, the video quality is degraded.
Even if the encoding bit rate is the same, it is removed during compression encoding according to the compression encoding complexity of the video to be processed, that is, according to the amount of high-frequency AC components included in the original image. The amount changes.

  FIG. 10 is an explanatory diagram showing changes in video quality depending on the compression encoding complexity. The horizontal axis indicates the encoding bit rate of the video media, and the vertical axis indicates the subjective video quality value. FIG. 10 shows characteristics 9A to 9C indicating changes in subjective video quality values according to the encoding bit rate for three videos A to C. Among these videos A to C, the video C has the lowest compression encoding complexity, and the video A has the highest compression encoding complexity.

Therefore, video C has a low compression encoding complexity and a low amount of high-frequency AC components, so even when the encoding bit rate is low, the amount of high-frequency AC components to be removed is small, and the subjective video quality value is not significantly reduced. Not noticeable.
On the other hand, video A has a high compression encoding complexity and a high amount of high-frequency AC components, so that when the encoding bit rate is low, the amount of high-frequency AC components to be removed increases and the subjective video quality value The decline of the is remarkable.

As described above, the subjective video quality value has a property of changing according to the compression encoding complexity of the video, but the conventional technology does not consider the change in the video quality due to the compression encoding complexity.
The present invention is for solving such problems, and a video quality estimation apparatus, method, and program capable of estimating subjective video quality in consideration of changes in video quality due to compression coding complexity The purpose is to provide.

  In order to achieve such an object, the video quality estimation apparatus according to the present invention reproduces video communication for transmitting video media compressed and encoded into a plurality of frames to an arbitrary terminal via a communication network. A video quality estimation apparatus for calculating an estimated value of subjective video quality perceived by a viewer from the video media, the quality parameter relating to the video quality included in a packet related to the video media and an average of the video media A storage unit that stores a video quality estimation model indicating a correspondence relationship with a subjective video quality estimation value, and an average compression encoding complexity indicating an average complexity related to compression encoding processing of video media, and an estimation target A quality parameter acquisition unit that acquires quality parameters from packets related to video media, and a quality parameter based on the video quality estimation model of the storage unit. A video quality estimator for estimating an average subjective video quality value corresponding to a meter, and a compression for calculating an evaluation compression encoding complexity indicating a complexity in a quality evaluation section related to a compression encoding process of video media based on the quality parameter Based on the difference compression coding complexity consisting of the difference between the coding complexity calculation unit and the average compression coding complexity of the storage unit and the evaluation compression coding complexity calculated by the compression coding complexity calculation unit, A video quality correction unit that corrects the average subjective video quality value estimated by the video quality estimation unit.

  At this time, the quality parameter acquisition unit acquires the quantization value of each transform coefficient corresponding to the pixel block provided in an arbitrary frame and the quantization step width corresponding to these transform coefficients from the packet, and performs compression encoding. By calculating and averaging the product of the quantized value of the transform coefficient and the quantization step width corresponding to the transform coefficient for each transform coefficient of the pixel block provided in an arbitrary frame in the complexity calculation unit Calculating the compression encoding complexity of the pixel block, and averaging the compression encoding complexity of the pixel block for each pixel block of each frame in the desired section of the video media, thereby calculating the compression encoding complexity in the desired section May be calculated.

  The storage unit also includes a differential compression coding complexity indicating a difference between an average compression coding complexity of the video media and a compression coding complexity in the quality evaluation section of the video media, and the differential compression coding complexity. Stores a video quality correction model indicating the correspondence with the correction rate corresponding to the degree, and the compression encoding complexity calculation unit calculates the evaluation compression encoding complexity in the quality estimation period of the video media, and corrects the video quality Corresponding to the differential compression coding complexity consisting of the difference between the average compression coding complexity of the storage unit and the evaluation compression coding complexity calculated by the compression coding complexity calculation unit based on the video quality correction model. The correction rate to be calculated may be calculated, and the average subjective video quality value estimated by the video quality estimation unit may be corrected based on the correction rate.

  In addition, as a video quality correction model, a function equation in which the correction rate monotonously increases as the differential compression coding complexity increases and the correction rate becomes 1 when the differential compression coding complexity is 0 may be used. Good.

  In addition, the video quality estimation method according to the present invention relates to video communication that includes an arithmetic processing unit and a storage unit, and transmits video media compressed and encoded into a plurality of frames to an arbitrary terminal via a communication network. A video quality estimation method used in a video quality estimation device for calculating an estimated value of subjective video quality perceived by a viewer from the video media reproduced in step 1, wherein the storage unit calculates an average for compression encoding processing of the video media. Storing an average compression coding complexity indicating general complexity, and an evaluation compression coding complexity indicating complexity in a quality evaluation section related to compression coding processing of video media based on a quality parameter by a calculation processing unit The compression coding complexity calculation step for calculating the degree of calculation and the arithmetic processing unit calculate the average compression coding complexity and the compression coding complexity calculation step of the storage unit. Based on the differential compression encoding complexity made from the difference between the evaluation compression encoding complexity that is, and a video quality correction step of correcting the average subjective image quality value of the video media.

  At this time, the quality parameter acquisition step includes a step of acquiring a quantization value of each transform coefficient corresponding to a pixel block provided in an arbitrary frame and a quantization step width corresponding to these transform coefficients from the packet, In the compression coding complexity calculation step, for each transform coefficient of the pixel block provided in an arbitrary frame, the product of the quantization value of the transform coefficient and the quantization step width corresponding to the transform coefficient is calculated and averaged Calculating the compression encoding complexity of the pixel block, and averaging the compression encoding complexity of the pixel block in each pixel block of each frame in the desired section of the video media. A step of calculating a compression encoding complexity may be provided.

  The storage unit also includes a differential compression coding complexity indicating a difference between an average compression coding complexity of the video media and a compression coding complexity in the quality evaluation section of the video media, and the differential compression coding complexity. A step of storing a video quality correction model indicating a correspondence relationship with a correction rate corresponding to the degree, and calculating an evaluation compression encoding complexity in the quality estimation period of the video media in the compression encoding complexity calculating step And the video quality correction step includes a difference between the average compression coding complexity of the storage unit and the evaluation compression coding complexity calculated in the compression coding complexity calculation step based on the video quality correction model. Calculating a correction factor corresponding to the differential compression coding complexity, and correcting the average subjective video quality value estimated in the video quality estimation step based on the correction factor And a step may be provided that.

  In addition, as a video quality correction model, a function equation in which the correction rate monotonously increases as the differential compression coding complexity increases and the correction rate becomes 1 when the differential compression coding complexity is 0 may be used. Good.

  The program according to the present invention includes an arithmetic processing unit and a storage unit, and is reproduced by a terminal for video communication in which video media compressed and encoded into a plurality of frames is transmitted to an arbitrary terminal via a communication network. In addition, a computer of a video quality estimation apparatus that calculates an estimated value of subjective video quality perceived by a viewer from the video media is related to the video quality obtained from a packet related to the video media to be estimated by an arithmetic processing unit. Based on the quality parameter, the average compression coding complexity calculating step for calculating the average compression coding complexity indicating the average complexity of the compression coding processing of the video media based on the quality parameter, and the arithmetic processing unit Compressed code for calculating the evaluation compression coding complexity indicating the complexity in the quality evaluation section for video media compression coding processing The average subjective video quality value of the video media is corrected based on the differential compression coding complexity that is the difference between the average compression coding complexity and the evaluation compression coding complexity by the complexity calculation step and the arithmetic processing unit. And a video quality correction step.

  According to the present invention, it is possible to estimate more accurate subjective video quality in consideration of the complexity of compression encoding processing of video media, that is, the characteristics of subjective video quality that changes according to the compression encoding complexity. It becomes possible. Further, the correction of the average subjective video quality value by the compression encoding complexity can be performed based on the quality parameter acquired from the packet related to the target video media. For this reason, the subjective video quality value considering the characteristic of the subjective video quality that changes according to the compression coding complexity can be extremely reduced without requiring processing or configuration for decoding the compression-encoded video media. Easy to estimate.

  As a result, it is possible to easily and accurately determine whether or not a certain level of quality is maintained for a user who uses the video distribution service. For this reason, it becomes possible to very easily grasp and manage the design of quality parameters used in the video distribution service and the actual quality of the video distribution service being provided.

Next, embodiments of the present invention will be described with reference to the drawings.
[Video quality estimation device]
First, a video quality estimation apparatus according to an embodiment of the present invention will be described with reference to FIG. FIG. 1 is a block diagram showing a configuration of a video quality estimation apparatus according to an embodiment of the present invention.

  The video quality estimation apparatus 10 is composed of an information processing apparatus such as a computer that performs arithmetic processing on input information, and a video communication medium encoded into a plurality of frames is transmitted from the communication terminal 2A to the communication terminal 2B terminal. 3 is an apparatus for calculating an estimated value of subjective video quality perceived by a viewer from the video media reproduced by the communication terminal 2B, using the video media and the estimation condition regarding the communication network as input. is there.

  The video quality estimation apparatus 10 includes a communication interface unit (hereinafter referred to as a communication I / F unit) 11, an operation input unit 12, a screen display unit 13, a storage unit 14, and an arithmetic processing unit 15 as main functional units. Is provided. The arithmetic processing unit 15 includes a quality parameter acquisition unit 15A, a video quality estimation unit 15B, a compression encoding complexity calculation unit 15C, and a video quality correction unit 15D as main processing units.

  In the present embodiment, the storage unit 14 uses the video quality estimation model indicating the correspondence between the quality parameter related to the video quality included in the packet related to the video media and the average subjective video quality estimation value of the video media, And an average compression encoding complexity indicating an average complexity relating to the compression encoding processing of the video media, and the quality parameter acquisition unit 15A acquires the quality parameters from the packets related to the video media to be estimated. The video quality estimation unit 15B estimates an average subjective video quality value corresponding to the quality parameter based on the video quality estimation model in the storage unit 14, and the compression coding complexity calculation unit 15C calculates the video based on the quality parameter. Calculate the evaluation compression coding complexity indicating the complexity in the quality evaluation section for media compression coding processing. Based on the difference compression coding complexity formed by the difference between the average compression coding complexity of the storage unit 14 and the evaluation compression coding complexity calculated by the compression coding complexity calculation unit 15C by the video quality correction unit 15D. The average subjective video quality value estimated by the video quality estimation unit 15B is corrected.

[Configuration of video quality estimation device]
Next, the configuration of the video quality estimation apparatus according to the embodiment of the present invention will be described in detail with reference to FIG.
The communication I / F unit of the video quality estimation apparatus 10 is composed of a dedicated communication circuit, and has a function of transmitting and receiving packets via the packet communication network 3.
The operation input unit 12 includes an operation input device such as a keyboard and a mouse, and has a function of detecting an operator operation and outputting the operation to the arithmetic processing unit 15.
The screen display unit 13 includes a screen display device such as an LCD or a PDP, and has a function of displaying display data such as an operation menu and an estimated video quality value output from the arithmetic processing unit 15 on the screen.

The storage unit 14 includes a storage device such as a hard disk and a memory, and has a function of storing various processing information used for processing operations in the arithmetic processing unit 15 and a program 14P. The program 14P is read in advance from an external device or a recording medium (not shown) via a data input / output function such as the communication I / F unit 11 provided in the video quality estimation apparatus 10, and is stored in the storage unit 14. ing.
Main processing information stored in the storage unit 14 includes a quality parameter 14A, a video quality estimation model 14B, an average subjective video quality value 14C, a video quality correction model 14D, an average compression encoding complexity 14E, an evaluation compression encoding complexity. 14F and a corrected subjective video quality value 14G.

  The quality parameter 14A is a parameter that influences the video quality of the target video media to be estimated, and is used for video quality estimation and compression coding complexity calculation. Specific examples of the quality parameter 14A include parameters such as an encoding method, a video format, an encoding bit rate, frame information, and packet loss information.

  When the target video medium uses a compression coding method such as MPEG2, parameters such as a frame rate, an intra flash rate, a GOP, a frame type, and residual information may be used as the frame information. This residual information is information on the residual signal after motion compensation. Specifically, the residual information includes quantization values of transform coefficients (spatial compression transform coefficients) such as DCT coefficients, integer transform coefficients, and wavelet coefficients, A parameter such as a step width may be used. As the packet loss information, parameters such as a packet loss rate, a packet loss pattern, and a packet delay time may be used.

  These quality parameters 14A include a header portion of a packet related to the target video medium, encoded data distributed in these packets, for example, in the case of MPEG2, a transport stream (TS: Transport Stream), an elementary It is stored in encoded data such as a stream (ES: Elementary Stream), and is acquired from these packets by the quality parameter acquisition unit 15A and stored in the storage unit 14. In addition, these parameters are stored in association with time position information on the video media from which the parameters are acquired, for example, identification information indicating a frame or an image block, as necessary.

The video quality estimation model 14B is an estimation model indicating a correspondence relationship between a video quality estimation parameter of the quality parameters 14A and an average subjective video quality value of the video media, that is, an average subjective video quality value. Specifically, in addition to the function expressions indicating these correspondence relationships, they may be configured in a known format such as a table format or a database.
The average subjective video quality value 14C is an average subjective video quality value related to the target video media.

The average compression encoding complexity 14E is a compression encoding complexity indicating an average complexity related to a compression encoding process calculated and stored in advance.
The evaluation compression encoding complexity 14F is a compression encoding complexity indicating the complexity in the quality evaluation section related to the compression encoding processing of the target video media, calculated from the parameters for calculating the compression encoding complexity of the quality parameters 14A. is there.

The video quality correction model 14D corresponds to the correspondence between the differential compression coding complexity formed by the difference between the average compression coding complexity of the video media and the evaluation compression coding complexity and the correction rate for the average subjective video quality value of the video media. It is a correction model that indicates the relationship, and specifically, it may be configured in a known format such as a table format or a database in addition to the function expression indicating the correspondence relationship.
The corrected subjective video quality value 14G is obtained from the average subjective video according to the correction rate corresponding to the differential compression coding complexity of the average compression coding complexity 14E and the evaluation compression coding complexity 14F obtained from the video quality correction model 14D. This is a subjective video quality value obtained by correcting the quality value 14C.

  The arithmetic processing unit 15 has a microprocessor such as a CPU and its peripheral circuits, reads the program 14P of the storage unit 14 and executes it by the microprocessor, thereby causing the hardware and the program to cooperate with each other, thereby causing the quality parameter It has a function of realizing various processing units such as an acquisition unit 15A, a video quality estimation unit 15B, a compression encoding complexity calculation unit 15C, and a video quality correction unit 15D.

  The quality parameter acquisition unit 15A has a function of acquiring various quality parameters 14A used for estimation of video quality and calculation of compression encoding complexity from various packets related to the target video media. The quality parameter 14A is stored in a header portion of a packet related to the target video media or encoded data transmitted in a distributed manner to these packets. The quality parameter acquisition unit 15A acquires the quality parameter 14A from these packets and stores it in the storage unit 14. For example, in the case of MPEG2, encoded data such as a transport stream or an elementary stream is analyzed by the quality parameter acquisition unit 15A, and a necessary quality parameter 14A is extracted.

The video quality estimation unit 15B has a function of estimating an average subjective video quality value 14C corresponding to a video quality estimation parameter among the quality parameters 14A based on the video quality estimation model 14B of the storage unit 14. As specific examples of the video quality estimation model 14B and the video quality estimation parameters used in the video quality estimation unit 15B, publicly known ones described in Non-Patent Document 3 described above may be used.
The compression coding complexity calculation unit 15C is an evaluation compression that indicates the complexity in the quality evaluation section related to the compression coding processing of the video media, based on the compression coding complexity calculation parameter among the quality parameters 14A of the storage unit 14. And a function of calculating the encoding complexity 14F.

  Based on the video quality correction model 14D stored in the storage unit 14, the video quality correction unit 15D corrects the difference compression coding complexity that is the difference between the average compression coding complexity 14E and the evaluation compression coding complexity 14F. And a function for calculating the corrected subjective video quality value 14G by correcting the average subjective video quality value 14C estimated by the video quality estimation unit 15B based on the obtained correction rate. is doing.

[Operation of video quality estimation device]
Next, the operation of the video quality estimation apparatus according to the embodiment of the present invention will be described with reference to FIGS. FIG. 2 is a flowchart showing the subjective video quality estimation operation of the video quality estimation apparatus according to the embodiment of the present invention. FIG. 3 is a flowchart showing the subjective video quality estimation process of the video quality estimation apparatus according to the embodiment of the present invention.

[Subjective image quality estimation processing]
With reference to FIG. 2 and FIG. 3, the subjective video quality estimation process in the video quality estimation apparatus 10 will be described.
The arithmetic processing unit 15 of the video quality estimation apparatus 10 starts the subjective video quality estimation process shown in FIG. 3 in response to an operator operation instructing the start of subjective video quality estimation detected by the operation input unit 12.
First, the arithmetic processing unit 15 controls the communication I / F unit 11 to capture various packets related to the target video media from the communication path between the communication terminals 2A and 2B via the packet communication network 3 ( In step 100), the quality parameter acquisition unit 15A acquires various quality parameters 14A used for video quality estimation and compression encoding complexity calculation, and stores them in the storage unit 14 (step 101).

Next, the arithmetic processing unit 15 uses the video quality estimation unit 15B to calculate the average subjective video quality value 14C of the target video media using the quality parameter 14A and the video quality estimation model 14B of the storage unit 14.
At this time, the video quality estimation unit 15B reads the video quality estimation model 14B from the storage unit 14 (step 102), and based on the video quality estimation model 14B, the video quality estimation parameter among the quality parameters 14A of the storage unit 14 Is calculated as an average subjective video quality value ν (14C) and stored in the storage unit 14 (step 103).

Subsequently, the arithmetic processing unit 15 executes a compression coding complexity calculation process described later by the compression coding complexity calculation unit 15C to calculate the evaluation compression coding complexity γ _est (14F) (step 104). ).

Next, the arithmetic processing unit 15 performs a video quality correction process (to be described later) by the video quality correction unit 15D, thereby comprising a difference between the average compression coding complexity γ _avg and the evaluation compression coding complexity γ _est. The corrected subjective video quality value V (14G) is calculated by correcting the average subjective video quality value ν of the storage unit 14 with the correction rate R corresponding to the differential compression coding complexity Δγ (step 105).
Thereafter, the arithmetic processing unit 15 stores the corrected subjective video quality value V as a subjective video quality value in the storage unit 14 or displays it on the screen display unit 13 as an estimated evaluation result (step 106), and a series of subjective video quality values. The quality estimation process ends.

[Compression coding complexity calculation processing]
Next, the compression coding complexity calculation process in the video quality estimation apparatus 10 will be described with reference to FIGS. FIG. 4 is a flowchart showing compression coding complexity calculation processing of the video quality estimation apparatus according to the embodiment of the present invention. FIG. 5 is an explanatory diagram showing a pixel distribution in a block in a video frame. FIG. 6 is an explanatory diagram showing a spatial frequency distribution in a block in a video frame. FIG. 7 is an explanatory diagram showing the quantization step width.

The compression encoding complexity calculation unit 15C of the arithmetic processing unit 15 executes the compression encoding complexity calculation process of FIG. 4 in step 105 of FIG. 3 described above. At this time, when the evaluation compression coding complexity γ _est is calculated in step 105, the quality evaluation section is used as the target section. These quality evaluation sections are assumed to be designated in advance by the operator and stored in the storage unit 14.

  First, the compression encoding complexity calculation unit 15C first processes an unprocessed frame (f) in which the compression encoding complexity is not calculated from F frames constituting the target section (where f is an integer of 1 to F). (Step 110), and an unprocessed image block (m, n) for which compression coding complexity has not been calculated from M × N image blocks constituting the frame (f) (where m is 1) (An integer of ~ M, n is an integer of 1 to N) is selected (step 111).

  As shown in FIG. 5, when a video is compression-encoded, an image of each frame is usually divided into M × N image blocks, and a compression-encoding process is performed for each image block. The pixel value of the image block is composed of I × J pixels arranged in a matrix in the horizontal direction x and the vertical direction y. These pixels have a luminance signal value (Y) as a pixel value, and in the case of a color image, a color difference signal value (Cr, Cb) is added.

In the compression encoding process, the pixel distribution of such an image block is converted into a spatial frequency distribution as shown in FIG. 6 by orthogonal transform processing such as DCT. This spatial frequency distribution shows the frequency of occurrence of pixels corresponding to each basic pattern in a matrix of I × J basic patterns having spatial frequencies preset in the horizontal direction x and the vertical direction y. This occurrence frequency becomes a conversion coefficient. Here, the conversion coefficient of the basic pattern (i, j) of the image block (m, n) of the frame (f) is expressed as _cfmnij .
In the compression encoding process, the transform coefficient c _fmnij is quantized to be converted into a quantized value q _fmnij having a small amount of information. At this time, as shown in FIG. 7, the width of the transform coefficient converted to the same quantized value q _fmnij as the transform coefficient c _fmnij is called a quantization step width h _fmnij .

The compression coding complexity calculation unit 15C _obtains the quantization value q _fmnij corresponding to the I × J transform coefficients c _fmnij constituting the selected image block (m, n) from the quality parameter 14A of the storage unit 14. At the same time (step 112), the quantization step width h _fmnij corresponding to these transform coefficients c _mnij is also acquired from the quality parameter 14A of the storage unit 14 (step 113). Then, based on the following equation (1), the corresponding conversion coefficient c _fmnij is calculated by the product of the quantized value _qfmnij and the quantization step width h _fmnij, and these are averaged for I × J basic patterns. Thus, the compression encoding complexity γ _fmn for the selected image block (m, n) is calculated. At this time, the compression encoding complexity γ _fmn may be calculated using the square of the product as shown in Equation (2).

Thereafter, the compression coding complexity calculation unit 15C determines whether or not the compression coding complexity calculation processing has been performed for all the image blocks in the selected frame (f) (step 115), and there is an unprocessed image block. In that case (step 115: NO), the process returns to step 111 and continues.
On the other hand, when the compression coding complexity calculation processing has been completed for all the image blocks in the selected frame (f) (step 115: YES), the compression coding complexity calculation unit 15C calculates the following equation (3). Based on the above, the compression coding complexity γ _f for the selected frame (f) is calculated by averaging the compression coding complexity γ _fmn for each image block in the selected frame (f) (step 116). .

Thereafter, the compression coding complexity calculation unit 15C determines whether the compression coding complexity calculation processing has been performed for all the frames (f) in the target section (step 117), and if there is an unprocessed frame, (Step 117: NO), the process returns to Step 110 and continues.
On the other hand, when the compression coding complexity calculation processing has been completed for all the frames (f) in the target section (step 117: YES), the compression coding complexity calculation unit 15C is based on the following equation (4). Then, the compression coding complexity γ _f for the target section is calculated by averaging the compression coding complexity γ _f for each frame (f) in the target section (step 118), and a series of compression coding complexity calculations is performed. The process ends.

[Video quality correction processing]
Next, with reference to FIG. 8 and FIG. 9, the video quality correction process of the video quality estimation apparatus according to the embodiment of the present invention will be described. FIG. 8 is a flowchart showing video quality correction processing of the video quality estimation apparatus according to the embodiment of the present invention. FIG. 9 is an explanatory diagram showing a video quality correction model.

The video quality correction unit 15D of the arithmetic processing unit 15 executes the compression coding complexity calculation process of FIG. 8 in step 106 of FIG. 3 described above.
First, the video quality correction unit 15D obtains the average compression coding complexity γ _avg (14E) from the storage unit 14 and the evaluation compression coding complexity γ _est (14F) calculated by the compression coding complexity calculation unit 15C. Reading (step 120), the differential compression coding complexity Δγ is calculated based on the following equation (5) (step 121).

  Next, the video quality correction unit 15D reads the video quality correction model 14D from the storage unit 14 (step 122), and corresponds to the differential compression encoding complexity Δγ based on the equation (6) of the video quality correction model 14D. A correction rate R is calculated (step 123). Here, b in Expression (6) is a coefficient determined by a codec that realizes encoding / decoding processing of video media and a video format.

  The video quality correction model 14D shown in FIG. 9 is an example composed of an exponential function equation. However, the correction rate R increases monotonously with an increase in the differential compression encoding complexity Δγ, and the differential compression encoding complexity Δγ is Any function equation that has a correction rate R of 1 when 0 can be used as the video quality correction model 14D. For example, the correction factor R may be calculated using a linear function expression such as Expression (7) corresponding to the broken line in FIG.

  Subsequently, the video quality correction unit 15D reads the average subjective video quality value ν (14C) calculated by the video quality estimation unit 15B from the storage unit 14 (step 124), and based on the following equation (8), the difference The compression coding complexity Δγ is corrected with the correction rate R, and the corrected subjective video quality value V is calculated (step 125).

  Thereafter, the video quality correction unit 15D stores the calculated corrected subjective video quality value V in the storage unit 14 (step 121), and ends a series of video quality correction processing.

[Effect of video quality estimation device]
Thus, in the present embodiment, the quality parameter acquisition unit 15A acquires the quality parameter 14A from the packet related to the video media to be estimated, and the video quality estimation unit 15B acquires the video quality estimation model in the storage unit 14. Based on 14B, an average subjective video quality value 14C corresponding to the quality parameter is estimated, and the compression coding complexity calculation unit 15C calculates the complexity in the quality evaluation section related to the compression coding processing of the video media based on the quality parameter. The evaluation compression coding complexity 14F is calculated, and the video compression unit 15D calculates the evaluation compression coding complexity 14F calculated by the average compression coding complexity 14E and the compression coding complexity calculation unit 15C of the storage unit 14. Average subjective video estimated by the video quality estimator 15B based on the differential compression coding complexity of the difference between It is corrected the quality value 14C.

  Accordingly, it is possible to estimate the subjective video quality more accurately in consideration of the complexity of the compression encoding process of the video media, that is, the characteristic of the subjective video quality that changes according to the compression encoding complexity. Further, the correction of the average subjective video quality value by the compression encoding complexity can be performed based on the quality parameter acquired from the packet related to the target video media. For this reason, the subjective video quality value considering the characteristic of the subjective video quality that changes according to the compression coding complexity can be extremely reduced without requiring processing or configuration for decoding the compression-encoded video media. Easy to estimate.

  As a result, it is possible to easily and accurately determine whether or not a certain level of quality is maintained for a user who uses the video distribution service. For this reason, it becomes possible to very easily grasp and manage the design of quality parameters used in the video distribution service and the actual quality of the video distribution service being provided.

  In the present embodiment, the quality parameter acquisition unit 15A determines the quantization value of each transform coefficient corresponding to the pixel block provided in an arbitrary frame from the packet and the quantization step width corresponding to these transform coefficients. The compression encoding complexity calculation unit 15C obtains the product of the quantized value of the transform coefficient and the quantization step width corresponding to the transform coefficient for each transform coefficient of the pixel block provided in an arbitrary frame. By calculating and averaging, the compression encoding complexity of the pixel block is calculated, and the compression encoding complexity of the pixel block is averaged for each pixel block of each frame in a desired section of the video media, thereby obtaining a desired value. Since the compression encoding complexity in the section is calculated, the compression encoding complexity can be calculated with high accuracy.

  In the present embodiment, the compression encoding complexity calculation unit 15C calculates the evaluation compression encoding complexity 14F in the quality estimation period of the video media, and the video quality correction unit 15D converts it to the video quality correction model 14D. On the basis of the correction rate corresponding to the differential compression coding complexity consisting of the difference between the average compression coding complexity 14E of the storage unit 14 and the evaluation compression coding complexity 14F calculated by the compression coding complexity calculation unit 15C. Since the average subjective video quality value 14C is corrected based on this correction factor, the average compression coding complexity can be easily corrected based on the compression coding complexity.

  Further, in the present embodiment, as a video quality correction model, a function equation that increases monotonously with an increase in differential compression encoding complexity and has a correction factor of 1 when the differential compression encoding complexity is 0 is used. For example, when the amount of high-frequency AC components is large and the compression encoding complexity is high, the subjective video quality characteristic that the subjective video quality value is significantly reduced with respect to the average subjective video quality value is accurately expressed. In addition, it can be easily expressed by a video quality correction model.

[Extended embodiment]
In the above embodiment, in the compression coding complexity calculation process of FIG. 4, the selected frame (f) is _obtained by averaging the compression coding complexity γ _fmn for each image block in the selected frame (f). Although the case where the compression encoding complexity γ _f is calculated has been described, there is a case where the packet loss occurs and the compression encoding complexity γ _fmn of an arbitrary image block cannot be calculated. In such a case, as an exception process, the compression coding complexity γ _fmn of the image block is _set to zero, and the compression coding complexity γ _fmn cannot be calculated from the number of image blocks M × N in Expression (3). The average may be calculated by subtracting the number of image blocks.

It is a block diagram which shows the structure of the video quality estimation apparatus concerning one embodiment of this invention. It is a flowchart which shows the subjective video quality estimation operation | movement of the video quality estimation apparatus concerning one embodiment of this invention. It is a flowchart which shows the subjective video quality estimation process of the video quality estimation apparatus concerning one embodiment of this invention. It is a flowchart which shows the compression encoding complexity calculation process of the video quality estimation apparatus concerning one embodiment of this invention. It is explanatory drawing which shows the pixel distribution in the block in a video frame. It is explanatory drawing which shows the spatial frequency distribution in the block in a video frame. It is explanatory drawing which shows a quantization step width. It is a flowchart which shows the video quality correction process of the video quality estimation apparatus concerning one embodiment of this invention. It is explanatory drawing which shows a video quality correction model. It is explanatory drawing which shows the change of the video quality by compression encoding complexity.

Explanation of symbols

  DESCRIPTION OF SYMBOLS 10 ... Video quality estimation apparatus, 11 ... Communication I / F part, 12 ... Operation input part, 13 ... Screen display part, 14 ... Memory | storage part, 14A ... Quality parameter, 14B ... Video quality estimation model, 14C ... Average subjective video quality Value, 14D: Video quality correction model, 14E: Average compression coding complexity, 14F: Evaluation compression coding complexity, 14G: Correction subjective video quality value, 14P: Program, 15: Arithmetic processing unit, 15A: Quality parameter acquisition 15B ... Video subjective quality estimation unit, 15C ... Compression coding complexity calculation unit, 15D ... Video quality correction unit, 2A, 2B ... Communication terminal, 3 ... Packet communication network.

Claims (9)

For video communication in which video media compressed and encoded into a plurality of frames is transmitted to an arbitrary terminal via a communication network, an estimated value of subjective video quality perceived by a viewer is calculated from the video media reproduced by the terminal A video quality estimation device,
A video quality estimation model indicating the correspondence between the quality parameter relating to the video quality included in the packet related to the video media and the average subjective video quality estimate of the video media, and an average relating to the compression encoding processing of the video media A storage unit for storing an average compression coding complexity indicating general complexity;
A quality parameter acquisition unit that acquires the quality parameter from a packet related to video media to be estimated;
A video quality estimation unit that estimates an average subjective video quality value corresponding to the quality parameter based on a video quality estimation model of the storage unit;
Based on the quality parameter, a compression encoding complexity calculating unit that calculates an evaluation compression encoding complexity indicating complexity in a quality evaluation section related to compression encoding processing of the video media;
Based on the differential compression coding complexity that is the difference between the average compression coding complexity of the storage unit and the evaluation compression coding complexity calculated by the compression coding complexity calculation unit, the video quality estimation unit A video quality estimation apparatus comprising: a video quality correction unit that corrects the estimated average subjective video quality value. The video quality estimation apparatus according to claim 1,
The quality parameter acquisition unit acquires a quantization value of each transform coefficient corresponding to a pixel block provided in an arbitrary frame and a quantization step width corresponding to these transform coefficients from the packet,
The compression encoding complexity calculation unit calculates, for each transform coefficient of a pixel block provided in an arbitrary frame, a product of a quantization value of the transform coefficient and a quantization step width corresponding to the transform coefficient. By calculating the compression encoding complexity of the pixel block by averaging, and averaging the compression encoding complexity of the pixel block for each pixel block of each frame in the desired section of the video media, A video quality estimation apparatus characterized by calculating a compression encoding complexity. The video quality estimation apparatus according to claim 1,
The storage unit includes a differential compression coding complexity indicating a difference between an average compression coding complexity of the video media and a compression coding complexity in the quality evaluation section of the video media, and the differential compression coding complexity. Store a video quality correction model that shows the correspondence with the correction rate corresponding to
The compression coding complexity calculation unit calculates an evaluation compression coding complexity in a quality estimation period of the video media;
The video quality correction unit includes a difference between an average compression encoding complexity of the storage unit and an evaluation compression encoding complexity calculated by the compression encoding complexity calculation unit based on the video quality correction model. A video quality estimation apparatus that calculates a correction rate corresponding to the compression coding complexity and corrects the average subjective video quality value estimated by the video quality estimation unit based on the correction rate. The video quality estimation apparatus according to claim 3,
The video quality correction model is composed of a functional expression in which the correction rate monotonously increases with an increase in the differential compression coding complexity and the correction rate is 1 when the differential compression coding complexity is 0. A video quality estimation device characterized by the above. For video communication, which has an arithmetic processing unit and a storage unit, and transmits video media compression-encoded into a plurality of frames to an arbitrary terminal via a communication network, a viewer uses the video media reproduced on the terminal. A video quality estimation method used in a video quality estimation apparatus that calculates an estimated value of subjective video quality to be realized,
Storing an average compression encoding complexity indicating an average complexity related to the compression encoding process of the video media by the storage unit;
A compression encoding complexity calculating step for calculating an evaluation compression encoding complexity indicating complexity in a quality evaluation section related to the compression encoding processing of the video media based on the quality parameter by the arithmetic processing unit;
Based on the difference compression coding complexity comprising the difference between the average compression coding complexity of the storage unit and the evaluation compression coding complexity calculated in the compression coding complexity calculation step by the arithmetic processing unit, A video quality estimation method comprising: a video quality correction step of correcting an average subjective video quality value of the video media. The video quality estimation method according to claim 5,
The quality parameter acquisition step includes a step of acquiring a quantization value of each transform coefficient corresponding to a pixel block provided in an arbitrary frame and a quantization step width corresponding to these transform coefficients from the packet,
The compression coding complexity calculation step calculates a product of a quantization value of the transform coefficient and a quantization step width corresponding to the transform coefficient for each transform coefficient of a pixel block provided in an arbitrary frame. Calculating a compression encoding complexity of the pixel block by averaging, and averaging the compression encoding complexity of the pixel block in each pixel block of each frame in a desired section of the video media And a step of calculating a compression encoding complexity in the section. The video quality estimation method according to claim 5,
By the storage unit, a differential compression coding complexity indicating a difference between an average compression coding complexity of the video media and a compression coding complexity in the quality evaluation section of the video media, and the differential compression coding complexity A step of storing a video quality correction model indicating a correspondence relationship with a correction rate corresponding to
The compression encoding complexity calculation step includes a step of calculating an evaluation compression encoding complexity in a quality estimation period of the video media,
The video quality correction step includes a difference consisting of a difference between an average compression coding complexity of the storage unit and an evaluation compression coding complexity calculated in the compression coding complexity calculation step based on the video quality correction model. Video quality comprising: calculating a correction rate corresponding to the compression coding complexity; and correcting the average subjective video quality value estimated in the video quality estimation step based on the correction rate Estimation method. The video quality estimation method according to claim 7,
The video quality correction model is composed of a functional expression in which the correction rate monotonously increases with an increase in the differential compression coding complexity and the correction rate is 1 when the differential compression coding complexity is 0. A method for estimating video quality. For video communication, which has an arithmetic processing unit and a storage unit, and transmits video media compression-encoded into a plurality of frames to an arbitrary terminal via a communication network, a viewer uses the video media reproduced on the terminal. In the computer of the video quality estimation device that calculates the estimated subjective video quality,
Storing an average compression encoding complexity indicating an average complexity related to the compression encoding process of the video media by the storage unit;
A compression encoding complexity calculating step for calculating an evaluation compression encoding complexity indicating complexity in a quality evaluation section related to the compression encoding processing of the video media based on the quality parameter by the arithmetic processing unit;
The arithmetic processing unit corrects an average subjective video quality value of the video media based on a differential compression coding complexity that is a difference between an average compression coding complexity of the storage unit and the evaluation compression coding complexity. A program that executes the video quality correction step.

Images