JP2016165054A - Encoding device, decoding device, and program - Google Patents

Abstract

PROBLEM TO BE SOLVED: To improve quality of a decoded signal by transmitting a residual component that cannot completely reproduce an input signal, from an encoding device to a decoding device when performing high-efficiency compression transmission.SOLUTION: An encoding device 1 comprises: a preprocessing part 11 which performs conversion processing on an input signal to generate a preprocessing signal; an encoding part 31 which encodes the preprocessing signal and generates an encoded signal; a local decoding part 32 which decodes the encoded signal and generates a locally decoded signal; a post-processing simulation part 12 which performs conversion processing on the locally decoded signal by using a plurality of parameters and generates a plurality of post-processing simulation signals; an auxiliary information encoding part 14 which encodes an optimal decision signal indicating a parameter that is used for generating the post-processing simulation signal of which the difference from the input signal is minimum, and generates an auxiliary information encoded signal; and a residual encoding part 16 which encodes a residual signal that is a differential between the input signal and the post-processing simulation signal of which the difference from the input signal is minimum, and generates a residual encoded signal. The encoded signal, the auxiliary information encoded signal, and the residual encoded signal are transmitted.SELECTED DRAWING: Figure 1

Description

  The present invention relates to an encoding device, a decoding device, and a program thereof.

  As a signal encoding method, there is a method of predicting a waveform and encoding a prediction residual. In particular, when encoding a video signal, when encoding a certain frame, the video is divided into blocks, and prediction processing is performed from another frame already encoded for each encoding target block or an image in the frame. And performing transform processing such as discrete cosine transform on the error between the prediction processing result and the encoding target block, quantizing the transform coefficient resulting from the transform processing, and further entropy coding the quantization result Thus, a method for realizing data compression is used. Such an encoding method is called hybrid encoding or MC-DCT (Motion Compensation / Discrete Cosine Transform) encoding. For example, MPEG-1, MPEG-2, MPEG-4, MPEG-4 AVC / H. H.264, and MPEG-H HEVC / H.264. Each video encoding method of H.265 belongs to MC-DCT encoding.

  Patent Document 1 discloses a technique for improving coding efficiency and subjective image quality by providing pre-processing and post-processing in an existing video coding system including MC-DCT coding. This method uses resolution reduction processing for preprocessing to reduce the compression rate of existing video coding methods, while applying super-resolution technology as postprocessing to achieve resolution restoration with less blur. ing. At this time, resolution restoration processing can be adjusted, an optimum adjustment value (parameter) that can best reproduce the input video is obtained in advance by the encoding device, and the optimum adjustment value is transmitted to the decoding device as auxiliary information. Thus, highly efficient compressed transmission is possible.

Japanese Patent No. 5419795

  However, in the method of Patent Document 1, although the encoding efficiency can be improved by the optimal adjustment of the resolution restoration, the residual component that cannot completely reproduce the input signal (original image) even if the optimal adjustment is performed is not a signal at all. There is no transmission. For this reason, when the restoration result cannot be driven to the required image quality, the image quality deterioration has become apparent in the unit of adjustment (for example, block unit) and may be detected as block distortion.

  The object of the present invention made in view of such circumstances is that the high-efficiency compressed transmission as in the method of Patent Document 1 cannot be reproduced even if optimal adjustment is performed. An object of the present invention is to provide an encoding device, a decoding device, and a program capable of transmitting a difference component from an encoding device to a decoding device and improving the quality of an output signal of the decoding device.

  In order to solve the above problems, an encoding apparatus according to the present invention generates a preprocess signal by performing a conversion process on an input signal and generates a preprocess signal by encoding the preprocess signal. A local decoding unit that decodes the encoded signal to generate a local decoded signal, performs a conversion process on the local decoded signal using a plurality of parameters, and generates a plurality of post-processing simulation signals. An auxiliary information code for generating an auxiliary information encoded signal by encoding an optimum determination signal indicating a parameter used for generating the post-processing simulation signal that minimizes a difference between the post-processing simulation unit to be generated and the input signal A residual encoding unit that encodes a residual signal that is a difference between the input signal and the post-processing simulation signal that minimizes the difference between the input signal and a residual encoded signal; The encoded signal, the auxiliary Broadcast coded signal, and wherein the transmitting the residual coding signal.

  Furthermore, in the encoding device according to the present invention, the encoded signal, the auxiliary information encoded signal, and the residual encoded signal are the encoded signal, the auxiliary information encoded signal, and the residual encoded signal. In this order, encoding is performed using different transmission paths or modulation schemes so that error tolerance increases.

  A program according to the present invention causes a computer to function as the encoding device.

  In order to solve the above problem, a decoding apparatus according to the present invention is a decoding apparatus that acquires the encoded signal, the auxiliary information encoded signal, and the residual encoded signal from the encoding apparatus, A decoding unit that decodes the encoded signal to generate a decoded signal; an auxiliary information decoding unit that decodes the auxiliary information encoded signal to generate an auxiliary information decoded signal; and the auxiliary information decoded signal with respect to the decoded signal A post-processing unit that generates a post-processing signal by performing a transformation process using the parameters indicated by: a residual decoding unit that decodes the residual encoded signal to generate a residual decoded signal; and the post-processing signal And an adder that outputs a signal obtained by adding the residual decoded signal to the outside.

  A program according to the present invention causes a computer to function as the decoding device.

  According to the present invention, when performing highly efficient compressed transmission, it is possible to further transmit a residual component that cannot be reproduced even after optimal adjustment from the encoding device to the decoding device. Therefore, the quality of the output signal of the decoding device can be improved.

It is a block diagram which shows the structural example of the encoding apparatus and decoding apparatus which concern on one Embodiment of this invention.

  Hereinafter, an embodiment of the present invention will be described in detail with reference to the drawings. In the embodiment, the signal to be transmitted is assumed to be a video signal. However, in the case of transmitting a signal of another form such as a still image signal or an audio signal, the same configuration and the same operation are performed. Can do.

  FIG. 1 is a block diagram illustrating a configuration example of an encoding device and a decoding device according to an embodiment of the present invention. The existing processing unit 3 that straddles the encoding device 1 and the decoding device 2 encodes and decodes the signal using the existing encoding method. Although illustration is omitted, a recording medium may be interposed between the encoding device 1 and the decoding device 2. In that case, the encoding apparatus 1 records the encoded signal on a recording medium, and the decoding apparatus 2 acquires the encoded signal from the recording medium. That is, the encoding device 1 transmits the encoded signal to the decoding device 2 directly or indirectly via a recording medium.

[Encoding device]
First, details of the encoding apparatus 1 will be described below. As shown in FIG. 1, the encoding apparatus 1 includes a preprocessing unit 11, a postprocessing simulation unit 12, an optimization unit 13, an auxiliary information encoding unit 14, a subtraction unit 15, and a residual encoding unit. 16, an encoding unit 31, and a local decoding unit 32.

  When the signal to be transmitted is a video signal, examples of the existing processing unit 3 include MPEG-1, MPEG-2, MPEG-4, MPEG-4 AVC / H. H.264, MPEG-H HEVC, and other standardized video encoding methods can be used.

  The preprocessing unit 11 performs a conversion process on the input signal H to generate a preprocessed signal L, and outputs the preprocessed signal L to the encoding unit 31. The preprocessing unit 11 reduces the data amount and bit depth of the input signal H, or substitutes the data value of each pixel into a variable or invariable conversion formula to convert it into another data value. For example, the preprocessing unit 11 resamples the input signal H to reduce the number of data samples. At this time, a thinning filter such as a low-pass filter may be applied prior to resampling. The cut-off frequency of the decimation filter may be determined according to the Nyquist frequency at the time of resampling, or may be changed according to a change in the signal waveform to be resampled (for example, a change in image between frames). Also good.

  The case where the input signal H is an image signal having a size of horizontal X pixels and vertical Y pixels will be described. Here, the pixel value at the image coordinates (x, y) of the input signal H is expressed as H (x, y) (the same applies to other images). The preprocessing unit 11 obtains a preprocessing signal L (x, y) having a size of X / 2 pixels in the horizontal direction and Y / 2 pixels in the vertical direction by, for example, conversion shown in Expression (1) or (2). .

  In Expression (2), g is a thinning filter, and G is a region representing the kernel size of the thinning filter g, which is represented by Expression (3), for example.

  The encoding unit 31 encodes the preprocessed signal L generated by the preprocessing unit 11 using, for example, a standardized video encoding method to generate an encoded signal, and outputs the encoded signal to the local decoding unit 32. The encoding unit 31 transmits the encoded signal to the decoding device 2 directly or indirectly via a recording medium.

  The local decoding unit 32 decodes the encoded signal generated by the encoding unit 31, generates a local decoded signal D, and outputs the local decoded signal D to the post-processing simulation unit 12. The local decoded signal D is the same as the decoded signal generated by the decoding unit 33 described later. The local decoding unit 32 may be the same as the decoding unit 33. When the encoding unit 31 is configured to output the locally decoded signal generated internally, the local decoding unit 32 does not need to be separately provided, and the local decoded signal output by the encoding unit 31 is post-processed. Input to the simulation unit 12.

The post-processing simulation unit 12 uses N parameters (N is a natural number of 1 or more) for the local decoded signal D generated by the local decoding unit 32 in order to simulate post-processing in the decoding device 2. By performing signal processing, N post-processing simulation signals E _{1 to} E _N are generated and output to the optimization unit 13. For example, the post-processing simulation unit 12 typically performs a conversion opposite to that of the pre-processing unit 11 or a conversion approximating this. For example, when the preprocessing unit 11 reduces the number of data samples to 1 / M times (for example, M is a real number with M> 1), the post-processing simulation unit 12 increases the number of data samples to M times.

When the local decoded signal D is an image signal, the post-processing simulation unit 12 processes the local decoded signal D using nearest neighbor interpolation (parameter 1), for example, and 2 × 2 times the resolution (number of samples) ) Is the post-processing simulation signal E _1, and the signal obtained by processing the local decoded signal D using bilinear interpolation (parameter 2) and increasing the resolution to 2 × 2 times is the post-processing simulation signal E _2. And a signal that has been processed by using bicubic interpolation (parameter 3) with respect to the local decoded signal D and increased to a resolution of 2 × 2 times is defined as a post-processing simulation signal E ₃ . Thus, post-processing simulation signals E _{1 to} E _N can be generated by different interpolation processes. In addition to or in addition to this, resolution conversion may be performed by super-resolution processing. For example, resolution conversion can be performed by wavelet super-resolution with respect to the local decoded signal D, which is an image signal, or multi-frame super-resolution using the local decoded signal D and frames before and after the local decoded signal D.

The optimization unit 13 compares the _N post-processing simulation signals E _{1 to} E _N generated by the post-processing simulation unit 12 with the input signal H, and performs post-processing simulation that minimizes the difference from the input signal H. An optimum determination signal indicating a parameter used to generate the signal En (1 ≦ n ≦ N) is generated. The optimization unit 13 may determine the optimality for each area obtained by dividing the screen, or may determine the optimality using the entire area of the image as an evaluation target. Furthermore, the optimizing unit 13 may determine an optimality in units of a group of image sequences (a plurality of frames) each including a plurality of time points of the input signal H. In general, the optimization unit 13 defines spatio-temporal partial areas for the post-processing simulation signals E _{1 to} E _N and the input signal H, and the post-processing simulation signals E _{1 to} E _N for each partial area. It is determined which is optimal.

For example, the optimizing unit 13 sets the index of the optimum post-processing simulation signal for each partial region R _p (p is a natural number of 1 or more and P or less, P is the total number of partial regions and is an integer of 1 or more). The optimal determination signal n _opt (p) that is (subscript of E _{1 to} E _N ) is determined using the determination formula of Formula (4) or (5).

Further, the optimization unit 13 _combines the post-processing simulation signals E _{1 to} E _N based on the optimal determination signal n _opt (p), generates an optimization signal F, and outputs it to the subtraction unit 15. The optimization signal F is obtained by, for example, Expression (6).

The auxiliary information encoding unit 14 encodes the optimum determination signal n _opt (p) generated by the optimization unit 13 to generate an auxiliary information encoded signal, and directly to the decoding device 2 or indirectly via a recording medium. Send to. At this time, the auxiliary information encoding unit 14 may apply reversible compression encoding. For lossless compression coding, for example, entropy coding such as Huffman coding or arithmetic coding and prediction between symbols (for example, a difference between symbols) can be used as necessary.

  The subtraction unit 15 calculates a residual signal Q that is a difference between the input signal H and the optimization signal F generated by the optimization unit 13 and outputs the residual signal Q to the residual encoding unit 16. When the input signal H is a video signal or an image signal, for example, the residual signal Q is obtained by Expression (7).

  The residual encoding unit 16 generates a residual encoded signal by performing lossless or irreversible compression encoding on the residual signal Q generated by the subtracting unit 15, and stores the recording medium directly or in the decoding device 2. Send indirectly through. For example, when the input signal H is a video signal, the residual encoding unit 16 can use an existing video encoding method (for example, MPEG-1, MPEG-2, MPEG-4, MPEG-4 AVC / H.264). MPEG-H HEVC, etc.) and existing still image coding methods (JPEG, etc.) can be used.

[Decoding device]
Next, details of the decoding device 2 will be described below. As illustrated in FIG. 1, the decoding device 2 includes an auxiliary information decoding unit 21, a post-processing unit 22, a residual decoding unit 23, an adding unit 24, and a decoding unit 33. The decoding device 2 directly obtains the encoded signal, the auxiliary information encoded signal, and the residual encoded signal from the encoding device 1 or when a recording medium is interposed between the encoding device 1 and the decoding device 2. Is obtained via a recording medium.

  The decoding unit 33 constitutes the existing processing unit 3 together with the encoding unit 31 and the local decoding unit 32 of the encoding device 1. The decoding unit 33 is paired with the encoding unit 31, acquires the encoded signal generated by the encoding unit 31, and generates a decoded signal D by decoding, for example, according to a standardized video encoding method. Output to the processing unit 22. The decoded signal D generated by the decoding unit 33 matches the local decoded signal D generated by the local decoding unit 32.

The auxiliary information decoding unit 21 acquires the auxiliary information encoded signal encoded by the auxiliary information encoding unit 14, decodes the auxiliary information encoded signal, and determines the optimum determination signal n _opt (p ) And output to the post-processing unit 22.

The post-processing unit 22 is mounted so that any of the N processes executed by the post-processing simulation unit 12 can be executed on the decoded signal generated by the decoding unit 33, and the optimum determination signal n _opt (p) The post-processing signal is generated by performing the conversion process using the parameter specified by the step, and is output to the adding unit 24. When the optimal determination signal n _opt (p) performs optimal determination on a region basis, the processing result is switched for each region. With this operation, the output of the post-processing unit 22 matches the optimization signal F output from the optimization unit 13. For example, the post-processing unit 22 calculates all N conversion processing results by the same implementation as the post-processing simulation unit 12, and according to the input optimum determination signal n _opt (p), these N pieces of conversion processing results. The output may be output while switching for each region so as to select the optimum one from the processing results.

  The residual decoding unit 23 acquires the residual encoded signal encoded by the residual encoding unit 16, decodes the residual encoded signal, generates a residual decoded signal R, and sends the residual decoded signal R to the adding unit 24. Output.

  The adder 24 adds the optimized signal F generated by the post-processor 22 and the residual decoded signal R generated by the residual decoder 23 as shown in Expression (8), and outputs an output signal S. Is generated and output to the outside.

  When the residual encoding unit 16 performs lossy encoding, the residual decoded signal R and the residual signal Q do not necessarily match completely. On the other hand, when the residual encoding unit 16 performs lossless encoding, the residual decoded signal R and the residual signal Q completely match, and as a result, the output signal S and the input signal output from the adding unit 24 H matches perfectly.

  As described above, the encoding device 1 according to the present invention includes the encoded signal encoded by the encoding unit 31, the auxiliary information encoded signal encoded by the auxiliary information encoding unit 14, and the residual code. The residual encoded signal encoded by the encoding unit 16 is transmitted, and the decoding device 2 acquires these signals. For this reason, even when performing highly efficient compressed transmission, the quality of the output signal S of the decoding apparatus 2 can be improved.

  These signals transmitted from the encoding device 1 to the decoding device 2 are preferably transmitted with higher priority in the order of the encoded signal, the auxiliary information encoded signal, and the residual encoded signal. In other words, even when only the encoded signal is transmitted, the signal can be reproduced with a large error by operating the post-processing unit 22 by a predetermined process.

  When only the encoded signal and the auxiliary information encoded signal are transmitted, the post-processing unit 22 can be controlled in an optimal state, and thus the signal is reproduced with higher accuracy than when only the encoded signal is transmitted. Can do.

  Then, when all of the encoded signal, the auxiliary information encoded signal, and the residual encoded signal are transmitted, the signal is reproduced with higher accuracy than when only the encoded signal and the auxiliary information encoded signal are transmitted. be able to.

  From such characteristics, when considering transmission paths and transmission path encoding, it is preferable to increase error tolerance in the order of the encoded signal, the auxiliary information encoded signal, and the residual encoded signal. For example, error resilience can be increased by using a modulation scheme with a small number of modulation levels.

  Note that the encoded signal, the auxiliary information encoded signal, and the residual encoded signal may all be transmitted on the same medium (transmission path), or may be all transmitted on separate media, Two and the other one may be transmitted on different media. For example, in the case of transmitting a video signal, the encoded signal is transmitted with high error resistance on the main transmission path of the broadcast wave, and the auxiliary information encoded signal is transmitted with low error resistance on the sub transmission path of the broadcast wave, Furthermore, a configuration in which the residual encoded signal is transmitted on IP (Internet Protocol) is also possible. That is, the encoded signal, the auxiliary information encoded signal, and the residual encoded signal enable scalable transmission in terms of quality.

  Note that a computer can be suitably used to function as the encoding device 1 or the decoding device 2 described above, and such a computer describes the processing contents for realizing each function of the encoding device 1 or the decoding device 2. This program can be realized by storing the program in a storage unit of the computer, and reading and executing the program by the CPU of the computer. This program can be recorded on a computer-readable recording medium.

  Although the above embodiment has been described as a representative example, it will be apparent to those skilled in the art that many changes and substitutions can be made within the spirit and scope of the invention. Therefore, the present invention should not be construed as being limited by the above-described embodiments, and various modifications and changes can be made without departing from the scope of the claims. For example, although the input signal H has been described as a video signal in the embodiment, the input signal H may be a still image signal or an audio signal. Moreover, it is possible to combine a plurality of constituent blocks described in the embodiment into one, or to divide one constituent block.

DESCRIPTION OF SYMBOLS 1 Encoding apparatus 2 Decoding apparatus 3 Existing processing part 11 Preprocessing part 12 Post-processing simulation part 13 Optimization part 14 Auxiliary information encoding part 15 Subtraction part 16 Residual encoding part 21 Auxiliary information decoding part 22 Post-processing part 23 Residual Difference decoding unit 24 Addition unit 31 Encoding unit 32 Local decoding unit 33 Decoding unit

Claims (5)

A preprocessing unit that performs a conversion process on the input signal to generate a preprocessed signal;
An encoding unit that encodes the preprocessed signal to generate an encoded signal;
A local decoding unit that decodes the encoded signal to generate a local decoded signal;
A post-processing simulation unit that performs conversion processing on the local decoded signal using a plurality of parameters and generates a plurality of post-processing simulation signals;
An auxiliary information encoding unit that generates an auxiliary information encoded signal by encoding an optimum determination signal indicating a parameter used for generating the post-processing simulation signal that minimizes the difference from the input signal;
A residual encoder that encodes a residual signal that is a difference between the input signal and the post-processing simulation signal that minimizes a difference between the input signal and a residual encoded signal;
An encoding apparatus that transmits the encoded signal, the auxiliary information encoded signal, and the residual encoded signal.   The encoded signal, the auxiliary information encoded signal, and the residual encoded signal are respectively increased in error tolerance in the order of the encoded signal, the auxiliary information encoded signal, and the residual encoded signal. The encoding apparatus according to claim 1, wherein encoding is performed using different transmission paths or modulation schemes.   A program for causing a computer to function as the encoding device according to claim 1 or 2. A decoding device that acquires the encoded signal, the auxiliary information encoded signal, and the residual encoded signal from the encoding device according to claim 1 or 2,
A decoding unit for decoding the encoded signal to generate a decoded signal;
An auxiliary information decoding unit that decodes the auxiliary information encoded signal to generate an auxiliary information decoded signal;
A post-processing unit that performs a conversion process on the decoded signal using a parameter indicated by the auxiliary information decoded signal to generate a post-processing signal;
A residual decoding unit that decodes the residual encoded signal to generate a residual decoded signal;
An adder that outputs a signal obtained by adding the residual decoded signal to the post-processing signal;
A decoding apparatus comprising:   The program for functioning a computer as a decoding apparatus of Claim 4.

Images