JP5981803B2 - Image quality evaluation apparatus, image quality evaluation method, and image quality evaluation program - Google Patents

Description

  The present invention relates to an image quality evaluation apparatus, an image quality evaluation method, and an image quality evaluation program for evaluating the subjective image quality of an image.

  To estimate the subjective image quality of moving images, (1) substitute with objective image quality obtained from the difference with the original image such as PSNR (Peak Signal-to-Noise Ratio), SSIM (Structural SIMilarity), (2) Subjective evaluation Some are estimated by experiment, and (3) the subjective image quality is calculated from the characteristics of the moving image. Among these, (1) has the problem that the quality of the objective image quality does not always correspond to the quality of the subjective image quality. In addition, when the frame rate and the resolution are changed from the original moving image, the objective image between the moving images having different frame rates and resolutions. There is a problem that it is difficult to compare image quality. (2) takes time to evaluate and cannot be used for estimation of subjective evaluation of individual moving images. In response to these problems, a subjective image quality estimation method (3) in which subjective image quality is calculated from the characteristics of moving images has been proposed (for example, see Non-Patent Documents 1 to 5).

M. Chien, R. Wang, C. Chiu, and P. Chang., "Quality Driven Frame Rate Optimization for Rate constrained Video Encoding," IEEE Trans on. Broadcast. Vol. 58, pp. 200-208, June. 2012 . R. Feghali, F. Speranza, D. Wang, ando A. Vincent., "Video Quality Metric for Bit Rate Control Via Joint Adjustment of Quantization and Frame Rate," IEEE Trans. Broadcast., Vol. 53, no. 1, Mar. 2007, pp. 441-46. Y.-F. Ou, Z. Ma, T. Liu, and Y. Wang., "Perceptual Quality Assessment of Video Considering Both Frame Rate and Quantization Artifacts," IEEE Trans. Circuits Syst. Video Tech., Vol. 21, no. 3, Mar. 2011, pp. 286-98. CS Kim SH Jin, DJ Seo, and YM Ro., "Measuring Video Quality on Full Scalability of H. 264 / AVC Scalable Video Coding," IEICE Trans. Commun., Vol. E91-B, no. 5, May 2008, pp. 1269-78. H. Sohn, H. Yoo, WD Neve, CS Kim, and YM Ro., "Full Reference Video Quality Metric for Fully Scalable and Mobile SVC Content," IEEE Trans. Broadcast., Vol. 56, no. 3, Sept. 2010, pp. 269-80. Y. Yamasaki, and T. Yoshida., "MOS-Based Bit Allocation in SNR-Temporal Scalable Coding," Proc. PCS, Dec. 2010. pp. 242-245

  However, in the methods described in Non-Patent Documents 1 to 5, in addition to the resolution and frame rate of moving images, evaluation when substituting with objective image quality obtained from differences from the original image such as texture complexity, amount of motion, PSNR, etc. Since the subjective evaluation is calculated based on the value, there is a problem that the amount of calculation increases. In order to solve such a problem, Non-Patent Document 6 proposes a technique for reducing the amount of calculation by fixing a bit rate and resolution and estimating a frame rate that optimizes subjective image quality.

  However, there is a strong demand for optimizing the subjective image quality by maintaining the frame rate and changing other parameters in the coding of moving images such as sports videos where movement is important. Therefore, when the bit rate and the frame rate are kept constant, it is necessary to be able to determine the moving image coding parameters for optimizing the subjective image quality at high speed.

  The present invention has been made in view of such circumstances, and an image quality evaluation apparatus and image quality capable of performing subjective evaluation at a high speed when the encoding resolution is changed with respect to a combination of a bit rate and a frame rate. An object is to provide an evaluation method and an image quality evaluation program.

The present invention is an image quality evaluation apparatus that performs subjective evaluation of image quality when the encoding resolution is changed with respect to a combination of a bit rate and a frame rate. Encoding parameter setting means for setting the resolution required for processing while changing, the resolution set by the encoding parameter setting means while the bit rate and the frame rate are kept constant, and encoding Encoded data input means for inputting encoded data encoded by the encoder that has input the target image, motion vector extracting means for extracting a plurality of motion vectors from the input encoded data, and the extracted Calculate the ratio between the standard deviation of the magnitude of the motion vector and the average of the magnitude of the motion vector, and calculate the ratio Characterized in that it comprises a quality evaluation value output means for outputting the value as an evaluation value of the image quality of the encoded data.

The present invention includes: the resolution set, and the image quality evaluation value storage means for storing in association with said evaluation value output from said image quality evaluation value output means upon setting the resolution, the image quality evaluation value storage unit Output means for outputting the resolution associated with the highest evaluation value among the stored evaluation values is further provided.

The present invention relates to an image quality evaluation method performed by an image quality evaluation apparatus that performs subjective evaluation of image quality when the resolution to be encoded is changed with respect to a combination of a bit rate and a frame rate. The encoding parameter setting step for setting the resolution necessary for the encoding process, the bit rate, and the frame rate are kept constant while the encoding parameter setting step sets the resolution. An encoded data input step for inputting encoded data encoded by the encoder to which the resolution and the image to be encoded are input, and a motion vector extracting step for extracting a plurality of motion vectors from the input encoded data And the standard deviation of the size of the extracted motion vector and the size of the motion vector Calculating a ratio of the average, and having an image quality evaluation value output step of outputting the calculated value of the said ratio as the evaluation value of the image quality of the encoded data.

The present invention relates to the image quality evaluation value storage step for storing the set resolution and the evaluation value output from the image quality evaluation value output step when the resolution is set in a storage device, and the storage device An output step of outputting the resolution related to the highest evaluation value among the stored evaluation values is further included.

  The present invention is an image quality evaluation program for causing a computer to function as the image quality evaluation apparatus.

  According to the present invention, since the ratio M / V between the standard deviation V and the average M of the magnitudes of all extracted motion vectors is output as the evaluation value of the image quality of the encoded data, the evaluation value can be obtained only with a simple calculation. Can be calculated and the calculation load can be reduced. In addition, it is possible to determine an encoding parameter that optimizes subjective image quality using the image quality evaluation value. Further, by calculating the image quality evaluation value by changing only the resolution while keeping the bit rate and the frame rate constant, it is possible to determine the resolution that optimizes the subjective image quality.

It is a block diagram which shows the structure of one Embodiment of this invention. It is a flowchart which shows operation | movement of the encoding parameter control part 2 shown in FIG. It is a flowchart which shows operation | movement of the image quality evaluation part 3 shown in FIG. It is explanatory drawing which shows the table structure of the image quality evaluation value memory | storage part 4 shown in FIG. It is a figure which shows the relationship between the reciprocal number (average / standard deviation) of a coefficient of variation (Coefficient of Variation), and an average opinion score (Mean Opinion Score).

  Hereinafter, an image quality evaluation apparatus according to an embodiment of the present invention will be described with reference to the drawings. The encoded data has the property that the subjective image quality is higher as the coefficient of variation of the motion vector, which is the ratio between the standard deviation and average of the motion vector, is smaller. FIG. 5 is a diagram showing the relationship between the reciprocal (average / standard deviation) of the coefficient of variation in the four sample videos (1) to (4) and the mean opinion score. As shown in FIG. 5, it can be seen that the subjective image quality is higher as the coefficient of variation of the motion vector, which is the ratio between the standard deviation and the average of the motion vector for each frame rate, is smaller. This embodiment is based on the property that the subjective image quality is higher as the coefficient of variation of the motion vector, which is the ratio of the standard deviation and average of the motion vector, is smaller under the fixed frame rate. Based on the subjective image quality.

  The encoding technology for moving images is H.264. H.264, MPEG-4, MPEG-2, HEVC, and the like. The moving image encoder is implemented based on these techniques, and can specify a bit rate, a resolution, and a frame rate as encoding parameters. The bit rate is determined by the compression rate of the moving image, and is actually determined by an internal parameter for each encoding method, for example, the value of a quantization coefficient. Therefore, the bit rate designation method depends on the encoder, and there is a method in which the bit rate can be directly designated. On the other hand, in some cases, it is necessary to directly designate an internal parameter of encoding such as a quantization coefficient. The bit rate specification includes fixed bit rate encoding that keeps the average bit rate of a moving image for a certain period of time within a certain range. Also, the internal parameters of the encoding are set, and accordingly, the bit rate depends on the nature of the moving image. There is variable bit rate coding in which the rate changes midway in moving picture coding.

  In this embodiment, the case of a fixed bit rate will be described. FIG. 1 is a block diagram showing the configuration of the embodiment. In this figure, reference numeral 1 denotes an encoder that inputs moving image data and encoding parameters, performs encoding processing, and outputs encoded data. Reference numeral 2 denotes an encoding parameter control unit that inputs uncompressed moving image data and encoding parameters from the outside, and outputs them to the encoder 1. The encoding parameters input by the encoding parameter control unit 2 are a bit rate, a frame rate, and a resolution table (table data in which a plurality of resolutions are defined). Reference numeral 3 denotes an image quality evaluation unit that receives encoded data output from the encoder 1 and calculates and outputs the reciprocal of the coefficient of variation of the motion vector as an evaluation value of the subjective image quality. Code 4 stores the coding parameters (bit rate, frame rate, resolution) output to the encoder 1 and the reciprocal (evaluation value) of the variation coefficient of the motion vector output from the image quality evaluation unit 3 in association with each other. It is an image quality evaluation value storage unit. Reference numeral 5 denotes an output unit that outputs an encoding parameter related to the highest evaluation value among the evaluation values stored in the image quality evaluation value storage unit 4.

  Next, an operation in which the encoding parameter control unit 2 shown in FIG. 1 outputs the uncompressed moving image data and the encoding parameter while changing the encoding parameter such as the resolution will be described with reference to FIG. FIG. 2 is a flowchart showing the operation of the encoding parameter control unit 2 shown in FIG.

  First, the encoding parameter control unit 2 inputs the uncompressed moving image data I, the bit rate b, the frame rate f, and the resolution table T from the outside (step S1). In the resolution table T, a plurality of resolution values are defined. In the resolution table, for example, resolutions are defined in the order of QVGA (320 pixels × 240 pixels), VGA (640 pixels × 480 pixels), 1280 pixels × 720 pixels, 1920 pixels × 1080 pixels, and the like.

  Next, the encoding parameter control unit 2 takes out the first resolution r of the resolution table T (step S2). Then, the encoding parameter control unit 2 outputs the uncompressed moving image data I, the bit rate b, the frame rate f, and the resolution r to the encoder 1 (step S3).

  In response to this, the encoder 1 performs an encoding process using the input encoding parameters (bit rate b, frame rate f, resolution r), and outputs encoded data to the image quality evaluation unit 3. The image quality evaluation unit 3 evaluates the image quality of the encoded data. The image quality evaluation processing operation will be described later.

  Next, when one image quality evaluation process is completed in the image quality evaluation unit 3, the encoding parameter control unit 2 determines whether or not an element remains in the resolution table T (step S4). Exit. On the other hand, if the element remains in the resolution table T, the encoding parameter control unit 2 extracts the next resolution r of the resolution table T (step S5), returns to step S3, and stores the moving image data I and bits before compression. The operation of outputting the rate b, the frame rate f, and the resolution r to the encoder 1 is repeated. With this processing operation, the resolution r defined in the resolution table T is sequentially output to the encoder 1 while keeping the bit rate and the frame rate constant, and the encoding parameter is set for each encoding process. Can be changed.

  In the processing operation shown in FIG. 2, an example in which only the resolution is changed while the bit rate b and the frame rate f are kept constant is shown, but not only the resolution but also the bit rate and the frame rate may be changed. . Furthermore, the target encoding parameters are not limited to the bit rate, frame rate, and resolution, and other encoding parameters may be used.

  Next, referring to FIG. 3, the image quality evaluation unit 3 shown in FIG. 1 receives the encoded data output from the encoder 1, calculates the reciprocal of the coefficient of variation of the motion vector as the subjective image quality evaluation value, and outputs it. The operation | movement which performs is demonstrated. FIG. 3 is a flowchart showing the operation of the image quality evaluation unit 3 shown in FIG.

  First, the image quality evaluation unit 3 receives encoded data output from the encoder 1 (step S11). Subsequently, the image quality evaluation unit 3 extracts all motion vectors from the input encoded data (step S12). Then, the image quality evaluation unit 3 obtains the reciprocal C of the coefficient of variation of the motion vector (step S13). The reciprocal C of the coefficient of variation of the motion vector is obtained from the ratio (C = M / V) between the standard deviation V of the motion vector magnitude and the average M of the motion vector magnitude. The image quality evaluation unit 3 stores the obtained reciprocal C of the coefficient of variation of the motion vector in the image quality evaluation value storage unit 4 as the subjective image quality evaluation value (step S14). At this time, the image quality evaluation unit 3 correlates the bit rate b, the frame rate f, and the resolution r output to the encoder 1 with the reciprocal C (evaluation value) of the obtained variation coefficient of the motion vector. Stored in the evaluation value storage unit 4. As a result, the image quality evaluation value (reciprocal C of the coefficient of variation of the motion vector) when the encoding process is performed with the bit rate b, the frame rate f, and the resolution r set for the encoder 1 is associated and stored. become.

  The image quality evaluation unit 3 repeatedly evaluates the image quality of the encoded data output from the encoder 1 every time a new encoding parameter (bit rate b, frame rate f, resolution r) is set. As a result, the image quality evaluation value storage unit 4 stores a table of evaluation values for combinations of encoding parameters. FIG. 4 is a diagram illustrating the structure of a table in which evaluation values are associated with combinations of bit rates, frame rates, and resolutions. The image quality evaluation value storage unit 4 stores the table shown in FIG. By referring to this table, an evaluation value for each combination of encoding parameters can be obtained. Therefore, when determining the encoding parameters, it is possible to easily determine the optimized encoding parameters. Then, the output unit 5 outputs an encoding parameter related to the highest evaluation value among the evaluation values stored in the image quality evaluation value storage unit 4.

  Note that the motion vectors extracted from the encoded data are not necessarily all, and a plurality of motion vectors that can obtain the standard deviation and average of the magnitudes of the motion vectors may be extracted.

  As described above, the ratio M / V of the standard deviation V and the average M of all extracted motion vectors is output as the evaluation value of the image quality of the encoded data. The subjective image quality evaluation value can be calculated directly from the encoded data of the data, and the calculation load can be reduced. Therefore, the quality evaluation of the encoded moving image data can be speeded up and highly reliable. In addition, when a moving image is encoded, subjective evaluation of image quality can be performed at high speed when the encoding resolution is changed with respect to a predetermined bit rate and frame rate combination. In addition, since the encoding parameter and the evaluation value are stored in association with each other when the encoding parameter is set, a moving image for optimizing the subjective image quality by referring to the relationship between the encoding parameter and the evaluation value It is also possible to easily determine the encoding parameter of the above.

  Note that the program for realizing the function of the processing unit in FIG. 1 is recorded on a computer-readable recording medium, and the program recorded on the recording medium is read into a computer system and executed to evaluate the image quality of the image. Processing may be performed. Here, the “computer system” includes an OS and hardware such as peripheral devices. The “computer system” includes a WWW system having a homepage providing environment (or display environment). The “computer-readable recording medium” refers to a storage device such as a flexible medium, a magneto-optical disk, a portable medium such as a ROM and a CD-ROM, and a hard disk incorporated in a computer system. Further, the “computer-readable recording medium” refers to a volatile memory (RAM) in a computer system that becomes a server or a client when a program is transmitted via a network such as the Internet or a communication line such as a telephone line. In addition, those holding programs for a certain period of time are also included.

  The program may be transmitted from a computer system storing the program in a storage device or the like to another computer system via a transmission medium or by a transmission wave in the transmission medium. Here, the “transmission medium” for transmitting the program refers to a medium having a function of transmitting information, such as a network (communication network) such as the Internet or a communication line (communication line) such as a telephone line. The program may be for realizing a part of the functions described above. Furthermore, what can implement | achieve the function mentioned above in combination with the program already recorded on the computer system, what is called a difference file (difference program) may be sufficient.

  As mentioned above, although embodiment of this invention has been described with reference to drawings, the said embodiment is only the illustration of this invention, and it is clear that this invention is not limited to the said embodiment. is there. Accordingly, additions, omissions, substitutions, and other changes of the components may be made without departing from the technical idea and scope of the present invention.

  When a moving image is encoded, it can be applied to a use in which it is essential to perform subjective evaluation of image quality at a high speed when the encoding resolution is changed with respect to a predetermined bit rate and frame rate combination.

  DESCRIPTION OF SYMBOLS 1 ... Image quality evaluation part, 2 ... Encoder, 3 ... Coding parameter control part, 4 ... Image quality evaluation value memory | storage part, 5 ... Output part

Claims (5)

An image quality evaluation apparatus that performs subjective evaluation of image quality when changing the encoding resolution for a combination of a bit rate and a frame rate,
Encoding parameter setting means for setting an encoder that performs an encoding process while changing the resolution necessary for the encoding process;
While the bit rate and the frame rate are kept constant, the encoded data encoded by the encoder that has input the resolution set by the encoding parameter setting means and the encoding target image is input. Encoded data input means;
Motion vector extracting means for extracting a plurality of motion vectors from the input encoded data;
An image quality evaluation value for calculating a ratio between the standard deviation of the extracted magnitude of the motion vector and the average of the magnitude of the motion vector, and outputting the calculated value of the ratio as an evaluation value of the quality of the encoded data An image quality evaluation apparatus comprising: output means. And set said resolution, and image quality evaluation value storage means for storing in association with said evaluation value output from said image quality evaluation value output means upon setting the resolution,
The image quality evaluation according to claim 1, further comprising: an output unit that outputs the resolution related to the highest evaluation value among the evaluation values stored in the image quality evaluation value storage unit. apparatus. An image quality evaluation method performed by an image quality evaluation apparatus that performs subjective evaluation of image quality when the encoding resolution is changed with respect to a combination of a bit rate and a frame rate,
An encoding parameter setting step for setting an encoder that performs an encoding process while changing the resolution necessary for the encoding process;
While the bit rate and the frame rate are kept constant, the encoded data encoded by the encoder that has input the resolution set by the encoding parameter setting step and the encoding target image is input. An encoded data input step;
A motion vector extraction step of extracting a plurality of motion vectors from the input encoded data;
An image quality evaluation value for calculating a ratio between the standard deviation of the extracted magnitude of the motion vector and the average of the magnitude of the motion vector, and outputting the calculated value of the ratio as an evaluation value of the quality of the encoded data An image quality evaluation method comprising: an output step. An image quality evaluation value storage step that associates the set resolution with the evaluation value output from the image quality evaluation value output step when the resolution is set and stores it in a storage device;
The image quality evaluation method according to claim 3, further comprising: an output step of outputting the resolution related to the highest evaluation value among the evaluation values stored in the storage device. Image quality evaluation program for a computer to function as the image quality evaluation apparatus according to claim 1 or 2.

Images