KR20080112000A - Encoding and decoding using the resemblance of a tonality - Google Patents

Abstract

An encoding and decoding method using resemblance of tonality to increase data processing speed are provided to increase an audio signal compression rate by decoding an audio signal in case of a frame in which a tonality is similar with a past frame. An encoding and decoding method using resemblance of tonality comprises the following steps: a step for extracting past frame information and difference information about a difference of a current frame and a past frame; a step for referring decoding information of the past frame corresponding to an extracted past frame information; and a step for compensating referred decoding information according to the difference information which is extracted.

Description

Encoding and Decoding using the Resemblance of a Tonality}

1 is a block diagram showing a configuration of an encoding apparatus according to an embodiment of the present invention;

2 is a diagram illustrating a form of a bitstream generated by an encoding apparatus according to an embodiment of the present invention.

FIG. 3 is a block diagram illustrating a configuration of a tone characteristic extracting unit of the encoding apparatus of FIG. 1;

4 is a matrix showing the center frequency of the bandpass filter constituting the filter of FIG.

5 is a block diagram showing the configuration of a decoding apparatus according to an embodiment of the present invention.

The present invention relates to a method and apparatus for encoding / decoding a signal. More particularly, the present invention relates to a method and apparatus for encoding / decoding which can reduce the amount of bitstream data by compressing a signal by using similarities of repeated tones.

In general, in encoding an audio signal, a signal is compressed in a psychoacoustic aspect. After analyzing the audio signal for each frequency, the method compares the energy level of each frequency and omits a signal that does not exist within the audible frequency or a signal that cannot be recognized due to a masking effect.

However, even though most music has a repetitive process, the existing compression method performs encoding using only the information of the corresponding frame. Therefore, even in the case of substantially repeated music, each is represented independently.

Therefore, there is a limit to the compression of the signal.

An object of the present invention is to provide an encoding / decoding method and apparatus capable of increasing the compression ratio of a signal by using encoding / decoding information of a past audio signal in the case of a repeated audio signal.

The decoding method according to the present invention for solving the technical problem as described above, in the audio signal decoding method, when the current frame included in the input bitstream refers to the past frame decoding information, the past frame information to be referred to, Extracting difference information about a difference between a current frame and the referenced past frame, referring to decoding information of the past frame corresponding to the extracted past frame information, and extracting the referenced decoding information from the extracted difference Compensating according to the information.

The encoding method according to the present invention for solving the technical problem as described above, in the audio signal encoding method, calculating the difference information between the current frame and the past frame included in the input signal, and the value of the difference information is set If it is less than the threshold value, characterized in that it comprises the step of generating a bitstream using the past frame information and the difference information.

According to an aspect of the present invention, there is provided a decoding apparatus including a storage unit for storing decoding information for each frame, and a corresponding past frame when a current frame included in an input bitstream refers to decoding information of a past frame. And a bit information unpacking unit for extracting the difference information of the difference information, and a difference information compensation unit for generating an audio signal by compensating the decoding information of the past frame according to the extracted difference information.

According to an aspect of the present invention, there is provided a coding apparatus including: a storage unit storing timbre characteristics of a past input frame, a timbre characteristic of a current frame included in an input signal, and a past frame timbre characteristic of the storage unit And difference information, and if the calculated difference information is less than or equal to a threshold value, a tone feature extraction unit for outputting corresponding past frame information and difference information, and a bit packing unit for generating a bitstream using the past frame information and the difference information. It is characterized by including.

The encoding / decoding method according to the present invention may preferably be implemented as a computer-readable recording medium having recorded thereon a program for execution on a computer.

The encoding / decoding apparatus according to the present invention may preferably be implemented as a multimedia broadcasting apparatus.

Hereinafter, a method and apparatus for encoding / decoding a signal according to the present invention will be described in detail with reference to the accompanying drawings.

1 is a block diagram showing an embodiment of the configuration of an audio signal encoding apparatus according to the present invention. The encoding apparatus shown in FIG. 1 includes an encoder 10, a tone characteristic extractor 20, a tone database 30, and a bit. It comprises a packing unit 40.

If an audio signal is given, it is divided into frames and encoded in a psychoacoustic aspect. For example, the sampled data is sampled according to a sound quality or a transmission rate to implement an input audio signal, which is an analog signal, and then quantized. Quantized signals are very sensitive to noise during transmission, so they are coded for transmission and processing.

The encoder 10 of FIG. 1 performs sampling, quantization, and encoding of an input audio signal, and compresses the audio signal through a variable length encoding or an entropy encoding scheme. In the case of MPEG audio, the signal is divided into frequency components and each subband is encoded. For each subband, a scale factor and a quantization step are adjusted in the quantization step, and psychoacoustic coding is performed so that the quantization noise is less than or equal to the minimum audible limit during the adjustment process.

At this time, the timbre characteristic extraction unit 20 extracts timbre characteristics of the input audio signal and compares how similar the timbre characteristics are extracted to the timbre characteristics of the previously input audio signal.

The tone database 30 is a database in which information on tone characteristics of the audio signal extracted by the tone feature extraction unit 20 is stored. Accordingly, when the tone feature extractor 20 extracts the tone feature of the input audio signal, the tone feature extractor 20 stores the extracted tone feature information in the tone database 30. In addition, the timbre feature extraction unit 20 determines how similar to the timbre feature of the past input audio signal stored in the timbre database 30.

In general, the comparison of the tone characteristics may be performed in units of frames, and the term “frame” used below is a concept including an audio signal unit that can be reproduced for a predetermined time. This is because the repetition of tone characteristics can be found in the chorus. In other words, the audio is generally composed of a plurality of choruses that are reproduced for a predetermined time at regular time intervals.

In the present invention, timbre (Timbre, Tonality) means the characteristics of the pitch of the sound source constituting each frame, the relative magnitude or vibration method between each sound. In general, the tone refers to the sensory characteristics generated by the difference in the components of the sound. In the present invention, characteristics that may affect the feeling of the sound in psychoacoustics are extracted in units of frames and the similarities are compared. Even if the tone has the same pitch, the tone is affected by the difference between the structure of the harmonic component and the energy ratio of each frequency component. For example, even if you play a note of the standard scale A4 = 440 Hz, the tone is different depending on the structure of harmonic components such as 880 Hz, 1320 Hz.

Therefore, the tone feature extractor 20 extracts the tone feature by performing frequency analysis on the input audio signal in units of frames. Then, the extracted tone characteristics are compared with the tone characteristics of past frames stored in the tone database 30 to calculate difference information.

When the difference information extracted by the tone feature extractor 20 is less than the set threshold, the tone of the currently input audio signal is determined to be similar to the tone of the past frame. In this case, the bit packing unit 40 does not generate the bitstream according to the audio data encoded by the encoder 10, but generates the bitstream according to the data ① transmitted from the timbre feature extraction unit 20. .

When the detected difference information is smaller than the threshold value, the timbre characteristic extraction unit 20 transmits past frame information and difference information referred to the bit packing unit 40. This will be described later in FIG. 3.

An embodiment of a bitstream generated by the bit packing unit 40 is illustrated in FIG. 2.

If the tone of the audio signal currently input by the tone feature extractor 20 is not determined to be similar to the tone of the past frame, the bit packing unit 40 generates a bit stream of the form as shown in (a). . This is a case where the past frame information and the difference information are not output to the bit packing unit 40 in the tone feature extractor 20.

That is, the number "0" is included in the header A in the meaning that the audio signal does not need to refer to the past frame information. The bitstream B following the header A contains the data encoded by the encoder 10. When the data included in the header A is '0', the decoding device decodes the subsequent bitstreams and outputs the subsequent bitstreams as audio signals.

If it is determined that the timbre of the audio signal currently input by the timbre characteristic extractor 20 is similar to the timbre of the past frame, the bit packing unit 40 generates a bitstream having a form as shown in (b). This is the case in which the tone characteristic extractor 20 outputs the past frame information and the difference information to the bit packing unit 40.

That is, the number "1" is included in the header C in the meaning that it indicates that the audio signal refers to the past frame information. The subsequent bitstream includes past frame information D to be referred to, and past frame and difference information E to be referred to.

3 is a diagram showing the structure of the tone characteristics extracting unit 20 according to an embodiment of the present invention. Filter unit 21 for performing frequency analysis of the input audio signal, and band information comparison analysis unit for comparing and analyzing the tone characteristics of the input audio signal calculated as a result of the frequency analysis with the tone characteristics of the past frame stored in the tone database 30 It consists of 22.

In one embodiment of the present invention, the tone characteristics of each frame may be represented by energy of each frequency band. In this case, the current audio signal, the past frame, and the difference information are information on energy label differences for each frequency band. However, this is only one embodiment for expressing the timbre, and it will be said that the use of other information for the timbre is included in the scope of the present invention.

As shown in FIG. 3, the filter unit 21 is composed of a plurality of band pass filters. In order to analyze the tone characteristics of the input audio signal, the filter unit 21 should include a bandpass filter having the center frequency as the frequency of the scale consisting of at least one octave. In the present exemplary embodiment, a band pass filter (# 1 to # 9 octave BPF) having a center frequency of a musical scale composed of nine octaves is included.

In other words, one octave scale consists of 12 bands. Accordingly, the filter unit 21 according to the present embodiment band-pass filters the input audio signal for a total of 9 octaves by 12 bands. 12 bands * 9 octaves, bandpass filtering for a total of 108 bands. This means that the input audio signal is filtered for each octave and for each pitch within each octave (12 bands because of 12 bands).

The band information comparison analyzer 22 calculates energy level information for each band according to the frequency analysis result of the filter unit 21 and stores the energy level information in the tone database 30 (②). The energy level information can be calculated using the amplitude of each frequency or the like. In addition, it compares with the energy level information for each frequency of the past frame stored in the tone database 30.

The band information comparison analyzer 22 calculates difference information between energy levels of the current audio signal and the previous frame. When the calculated difference information is less than or equal to the threshold, it can be seen that the current audio signal has a substantially similar energy level for each frequency.

Accordingly, the band information comparison and analysis unit 22 transmits the past frame information, which is substantially similar in frequency to the current input audio signal, and the difference information between the past frames to the bit packing unit 40.

The bit packing unit 40 generates and outputs a bit stream including the '1' in the header and the difference information and the header information transmitted from the band information comparison analyzer 22 (FIG. 2B).

The threshold compared with the calculated difference information is determined in consideration of psychoacoustic factors. In other words, based on the energy extracted for each band, Equal Loudness Contour (a curve with the same sensational magnitude that a person feels about sound in acoustics is experimentally obtained according to frequency and reference sound level), masking effect, etc. Is determined.

The threshold will ultimately be determined depending on the economics of transmitting the encoded data of the audio signal and transmitting the difference information. If the data amount of the encoding data transmission of the audio signal is not large, the encoding data (output data of the encoder 20) of the input audio signal will be included in the bitstream.

However, if there is no problem in decoding the audio signal by transmitting only the difference information and the amount of data of the difference information is smaller than the encoded data, a bitstream composed of the difference information and the referenced frame information is generated.

4 illustrates a center frequency (Hz) of a band pass filter constituting the filter unit 21 according to an embodiment of the present invention in a matrix form.

The y-direction of the matrix lists the bands, or pitches, of one octave according to frequency magnitude. As it is 12 bands, it can be seen that there are 12 pitches (A ~ G #). To the right is a higher note with a higher frequency.

The x direction of the matrix is an octave bandpass filter that filters the higher frequency band toward the right, which lists the octave numbers 1 to 9 covered by the plurality of bandpass filters included in the filter unit 21. .

In particular, the number of band pass filters constituting the filter unit 21 of the present invention can be reduced according to the sound quality to be implemented, and is based on a standard scale with A4 = 440 Hz. In general, almost all of the sound sources are made by the standard scale based on 440 Hz, and even slightly differently tuned sound sources near 440 Hz will not be overwhelmed by passing through a bandpiece filter of 440 Hz center frequency. It can be said that it is a characteristic of invention.

Meanwhile, the decoding apparatus according to the present invention may restore an audio signal by performing an inverse process of the above-described operation of the encoding apparatus.

FIG. 5 is a block diagram schematically illustrating a configuration of a decoding apparatus according to an embodiment of the present invention, wherein the bit unpacking unit 50, the decoder 60, the buffer 70, and the difference information compensator 80 are provided. It is composed.

The bit unpacking unit 50 checks whether the audio signal to be decoded should refer to the decoding information of the past frame through the header of the input bitstream. If '0' is included in the header, decoding is performed through the decoder 60 since subsequent bitstream data is encoding data of an audio signal.

The decoder 60 performs an inverse process of encoding performed by the encoder 10 of the audio signal, and depends on a decoding method corresponding to the encoding method of the encoder 10. In the case of auditory psychometric coding according to variable length coding, entropy coding, or MPEG audio, decoding is performed in the reverse order of the coding. The encoder 10 and the decoder 60 of the present invention can perform encoding and decoding according to various other methods, all of which will belong to the scope of the present invention.

The decoded data is stored in the buffer 70 as the decoded data of the corresponding frame. That is, the buffer 70 stores information about the number of frames input in the past, and decoded data of the corresponding frame.

If '1' is included in the header, the past frame should be referenced when restoring the audio signal. Accordingly, the bit unpacking unit 50 extracts the difference information for compensating the decoding information of the past frame from the input bit stream.

The difference information compensator 80 reads decoding information of a past frame to be referred to according to the past frame information extracted by the bit unpacking unit 50 from the buffer 70. In addition, the decoding information is compensated according to the difference information extracted from the bit unpacking unit 50.

According to the present embodiment, the difference information consists of information on the energy level difference for each frequency. Accordingly, the difference information compensator 80 may add or subtract the energy level for each frequency of the decoded signal of the frame to be referred to based on the information about the energy level difference for each frequency included in the difference information.

The compensated decoding information is stored together with the frame information in the buffer 70 as decoding information of the corresponding frame. Therefore, when a second frame referring to the first frame to be compensated according to the difference information during decoding is input as an audio signal, the difference information of the second frame may be generated based on the decoding information of the first frame. .

In the case of the media in which the sound source is stored, since the buffer 70 is a specific address of the media file in which the sound source is stored, a memory for buffering is not separately required. In this case, the difference information compensator reads decoding information to be directly decoded by directly accessing the media in which the sound source is stored according to the address information to be referred to. Thereafter, the difference information compensator decodes the audio signal by adding or subtracting an energy level corresponding to the difference information to the read decoding information.

Also, in this case, the tone characteristic extractor may not only refer to the past frame but may analyze the entire sound source and set the average of similar frame stages as independent frames. Accordingly, the difference information is difference information based on decoding information of the independent frame. The decoder may decode the entire sound source by compensating the difference information with reference to the independent frame.

Therefore, the amount of data to be stored in the media can be greatly reduced, and the amount of signal processing computation required for audio signal reproduction can be greatly reduced. The playback device with a very small amount of data storage also enables playback of multiple sound sources.

The encoding / decoding method according to the present invention described above may be stored in a computer-readable recording medium that is produced as a program for execution on a computer, and multimedia data having a data structure according to the present invention may also be read by a computer. Can be stored in a recording medium.

The computer readable recording medium includes all kinds of storage devices in which data that can be read by a computer system is stored. Examples of computer-readable recording media include ROM, RAM, CD-ROM, magnetic tape, floppy disk, optical data storage, and the like, and may also be implemented in the form of a carrier wave (for example, transmission over the Internet). Include. The computer readable recording medium can also be distributed over network coupled computer systems so that the computer readable code is stored and executed in a distributed fashion. In addition, functional programs, codes, and code segments for implementing a user tracking method can be easily inferred by programmers in the art to which the present invention belongs. In addition, the bitstream generated by the above-described encoding method may be stored in the computer-readable recording medium or transmitted using a wired / wireless communication network.

The decoding apparatus according to the present invention described above may be provided in a mobile communication terminal, a multimedia player such as a portable multimedia player (PMP), and a playback device such as a personal digital assistant (PDA). In addition, the above-described decoding apparatus may be implemented in hardware and included in a reproduction apparatus or the like, or as described above, may be implemented in a reproduction apparatus in software.

In addition, although the preferred embodiment of the present invention has been shown and described above, the present invention is not limited to the specific embodiments described above, but the technical field to which the invention belongs without departing from the spirit of the invention claimed in the claims. Of course, various modifications can be made by those skilled in the art, and these modifications should not be individually understood from the technical spirit or the prospect of the present invention.

According to the present invention as described above, in encoding / decoding an audio signal, in the case of a frame having a similar tone to a past frame, the audio signal is decoded by compensating for the difference from the past frame, thereby increasing the audio signal compression rate and increasing the data processing speed. Can be increased.

Claims (23)

In the audio signal decoding method, If the current frame included in the input bitstream refers to the past frame decoding information, extracting the past frame information referred to and difference information about a difference between the current frame and the referenced past frame; Referring to decoding information of the past frame corresponding to the extracted past frame information; And Compensating the referenced decoding information according to the extracted difference information. The method of claim 1, Storing decoding information of the compensated signal as new frame decoding information. The method of claim 1, The difference information is information on the energy level difference for each frequency of the sound source constituting the current frame and the past frame. The method of claim 1, Decoding the input bitstream if the current frame does not refer to decoding information of a past frame. The method of claim 4, wherein Storing decoding information of the decoded bitstream as new frame decoding information. In the audio signal encoding method, Calculating difference information between a current frame and a past frame included in the input signal; And And generating a bitstream by using the past frame information and the difference information when the value of the difference information is equal to or less than a predetermined threshold value. The method of claim 6, The difference information calculating step may include calculating, as difference information, information on a difference between a frequency analysis result of the current frame and a frequency analysis result of the past frame. The method of claim 7, wherein The frequency analysis may include bandpass filtering the input signal using a frequency of a scale consisting of at least one octave as a center frequency; And calculating an energy level of each bandpass filtered band. The method of claim 7, wherein And storing the frequency analysis result of the current frame. The method of claim 6, And the difference information is information on an energy level difference for each frequency of the current frame and the past frame. The method of claim 6, And generating a bitstream by encoding the input signal when the value of the difference information is greater than or equal to a set threshold. A storage unit for storing decoding information for each frame; A bit unpacking unit configured to extract difference information from the past frame when the current frame included in the input bitstream refers to decoding information of the past frame; And And a difference information compensation unit for compensating the decoding information of the past frame according to the extracted difference information to generate an audio signal. The method of claim 12, And the difference information compensation unit stores decoding information of the compensated signal as the new frame decoding information in the storage unit. The method of claim 12, The difference information is information on the energy level difference for each frequency of the sound source constituting the current frame and the past frame. The method of claim 12, And decoding the input bitstream if the bitstream does not refer to decoding information of a past frame. The method of claim 15, And the decoder stores decoding information of the decoded bitstream as the new frame decoding information in the storage unit. A storage unit for storing the timbre characteristics of a frame input in the past; A tone feature extraction unit configured to calculate a tone characteristic of a current frame included in an input signal and a difference information of a past frame tone characteristic in the storage unit, and output corresponding past frame information and difference information when the calculated difference information is less than or equal to a threshold; And And a bit packing unit generating a bit stream using the past frame information and the difference information. The method of claim 17, And the tone feature extracting unit calculates, as difference information, information on a difference between a frequency analysis result of the current frame and a frequency analysis result of the past frame. The method of claim 18, The tone characteristic extractor comprises: a filter unit configured to band pass filter the frequency of the scale consisting of at least one octave of the current frame as a center frequency; And The energy level of each bandpass filtered band is calculated to calculate energy level and difference information for each frequency of a past frame stored in the storage unit, and the bit-packing the past frame information and the difference information whose calculated difference information is equal to or less than a threshold value. And a band information comparison and analysis unit for outputting to a negative unit. The method of claim 18, And an encoder for encoding the input signal. The method of claim 20, And the bit packing unit generates a bitstream according to encoding data of the encoder when past frame information and difference information are not output from the tone feature extraction unit. A computer-readable recording medium having recorded thereon a program for executing the method according to any one of claims 1 to 11. A multimedia broadcasting apparatus comprising the apparatus according to any one of claims 12 to 21.

Images