CN101572586B - Codec method, device and system - Google Patents

Abstract

The embodiment of the invention discloses an encoding and decoding method, device and system, which relate to communication technology and can be encoded to form a code stream with an embedded structure to obtain a higher-quality decoded signal. The technical solution provided by the embodiment of the present invention includes an encoding method, the method comprising: acquiring at least two subbands of the signal; acquiring the amplitude envelope quantization value of each subband; using the amplitude envelope quantization value of each subband Obtain the hierarchical ordering of the subbands and the encoding information of the subbands; write the encoding information of the subbands into the code stream according to the hierarchical ordering of the subbands.

Description

Codec method, device and system

technical field

The present invention relates to communication technology, in particular to a codec method, device and system.

Background technique

With the rapid development of multimedia technology, speech and audio coding technology has been more and more widely used. In the existing speech and audio coding technology, transform coding technology is mainly used, and the speech and audio signals are first transformed in the frequency domain, for example, Fast Fourier Transform (Fast Fourier Transform, FFT), Modified Discrete Cosine Transform (Modified Discrete Cosine Transform) , MDCT), Modulated Lapped Transform (MLT), etc., and then quantize and encode the obtained frequency domain coefficients, and write the encoding results into the code stream for transmission. Usually, the Huffman coding method based on MLT transform is adopted.

In the process of realizing the present invention, the inventor found that there are at least the following problems in the prior art: the speech and audio encoding method provided in the prior art is a fixed-rate encoding method, which is not convenient for network transmission; The impact of packet loss is relatively large, and high-quality decoded signals cannot be guaranteed.

Contents of the invention

On the one hand, the embodiment of the present invention provides an encoding method, which can encode and form a code stream with an embedded structure, and obtain a higher-quality decoded signal after decoding the code stream.

The technical scheme that the embodiment of the present invention adopts is as follows: a kind of coding method, this method comprises:

acquiring at least two subbands of the signal;

Acquiring the amplitude envelope quantization value of each sub-band;

Obtaining the hierarchical ordering of the subbands and the coding information of the subbands by using the amplitude envelope quantization values of the subbands;

Writing the encoding information of each sub-band into a code stream according to the hierarchical ordering of each sub-band.

On the other hand, an embodiment of the present invention provides an encoding device capable of encoding to form a code stream with an embedded structure, and obtaining a higher-quality decoded signal after decoding the code stream.

The technical scheme adopted in the embodiment of the present invention is as follows: an encoding device, which comprises:

a subband acquisition unit, configured to acquire at least two subbands of the signal;

an amplitude envelope quantization value acquisition unit, configured to acquire the amplitude envelope quantization value of the subband acquired by the subband acquisition unit;

a hierarchical ordering and coding information acquisition unit, configured to obtain the hierarchical ordering of the subbands and the coding information of the subbands by using the amplitude envelope quantization values of the subbands;

The code stream writing unit is configured to write the encoding information of each sub-band obtained by the encoding information obtaining unit into the code stream according to the hierarchical order determined by the hierarchical order determining unit.

In another aspect, an embodiment of the present invention provides a decoding method, which can obtain a higher-quality decoded signal.

The technical scheme that the embodiment of the present invention adopts is as follows: a kind of decoding method, this method comprises:

Obtain decoding information from the code stream;

determining a hierarchical ordering of subbands in a signal according to the decoding information, the signal including at least two subbands;

Decoding the decoded information according to the hierarchical ordering to obtain a decoded signal.

In another aspect, an embodiment of the present invention provides a decoding device capable of obtaining a higher-quality decoded signal.

The technical scheme adopted in the embodiment of the present invention is as follows: a decoding device, which comprises:

a decoding information obtaining unit, configured to obtain decoding information from the code stream;

A hierarchical ordering acquisition unit, configured to determine the hierarchical ordering of each subband in the signal according to the decoding information acquired by the decoding information acquisition unit, and the signal includes at least two subbands;

The decoded signal acquiring unit is configured to decode the decoded information acquired by the decoded information acquiring unit according to the layer order acquired by the layer order acquire unit, and acquire a decoded signal.

In another aspect, an embodiment of the present invention provides a codec system, the system includes an encoding device and a decoding device, and the encoding device includes:

A subband acquisition module, configured to acquire at least two subbands of the signal;

An amplitude envelope quantization value acquisition module, configured to acquire the amplitude envelope quantization value of the subband acquired by the subband acquisition module;

A hierarchical ordering and coding information acquisition module, configured to obtain the hierarchical ordering of each sub-band and the encoding information of each sub-band by using the amplitude envelope quantization value acquired by the amplitude envelope quantization value acquisition module;

The code stream writing module is configured to write the coding information of each sub-band acquired by the coding information obtaining module into the code stream according to the hierarchical order determined by the hierarchical order determining module.

The decoding device includes:

A decoding information acquisition module is used to obtain decoding information from the code stream;

A hierarchical ordering acquisition module, configured to determine the hierarchical ordering of each subband in the signal according to the decoding information acquired by the decoding information acquisition module, the signal including at least two subbands;

The decoding signal obtaining module is configured to decode the decoding information obtained by the decoding information obtaining module according to the hierarchical ordering obtained by the hierarchical ordering obtaining module, and obtain a decoded signal.

The technical solution adopted in the embodiment of the present invention first determines the hierarchical ordering for each different subband in the signal, and the hierarchical ordering sorts each subband according to the importance of the information it carries, and when writing the code stream According to the hierarchical ordering, the coding information of the corresponding sub-bands is written in order according to the importance from large to small, and the coding information of the sub-bands carrying more important information is written first, so that the code stream structure written is based on the carried information. The difference in importance is divided into different levels, forming a code stream with an embedded structure, which can transmit code streams with different levels according to network conditions and user requirements for service quality. During decoding, according to the arrangement order of each sub-band, the decoding information of each sub-band is decoded in sequence according to the magnitude of the importance, and the decoding information with higher importance is decoded first, so that even if the Part of the decoding information of low importance is missing in the code stream, and the technical solution provided by the embodiment of the present invention can also obtain a higher-quality decoding signal.

Description of drawings

FIG. 1 is a flowchart of an encoding method provided by Embodiment 1 of the present invention;

FIG. 2 is a flowchart of an encoding method provided by Embodiment 2 of the present invention;

FIG. 3 is a schematic diagram of a code stream structure provided by an embodiment of the present invention;

Fig. 4 is a schematic diagram of the position of the truncation point when the code stream is truncated according to the embodiment of the present invention;

FIG. 5 is a schematic diagram of code stream truncation provided by an embodiment of the present invention;

FIG. 6 is a flowchart of a decoding method provided by an embodiment of the present invention;

FIG. 7 is a schematic diagram of a codec system provided by an embodiment of the present invention;

FIG. 8 is a structural diagram of an encoding device provided in Embodiment 1 of the present invention;

FIG. 9 is a structural diagram of an encoding device provided in Embodiment 2 of the present invention;

FIG. 10 is a structural diagram of a decoding device provided by Embodiment 1 of the present invention;

FIG. 11 is a structural diagram of a decoding device provided by Embodiment 2 of the present invention.

Detailed ways

In order to illustrate the technical solutions of the embodiments of the present invention more clearly, the embodiments of the present invention will be described in detail below in conjunction with the accompanying drawings. The following descriptions are only some embodiments of the present invention. For those of ordinary skill in the art, On the premise of not paying creative efforts, other implementation modes of the present invention can also be obtained according to these embodiments.

In order to solve the problem that the encoding method based on MLT transformation in the prior art is not convenient for network transmission, the decoding is greatly affected by network packet loss, and the quality of the acquired voice and audio is not good, the embodiment of the present invention provides an encoding Decoding method, device and system. The following describes the embodiments of the present invention in detail.

The embodiment of the present invention provides an encoding method, as shown in Figure 1, the method includes the following steps:

Step 11, acquiring at least two subbands of the signal;

Step 12, obtaining the amplitude envelope quantization value of each sub-band;

Step 13, using the quantized values of the amplitude envelopes of the sub-bands to obtain the hierarchical ordering of the sub-bands and the coding information of the sub-bands;

Step 14: Write the encoding information of each sub-band into a code stream according to the hierarchical ordering of each sub-band.

The technical solution adopted in the embodiment of the present invention first determines the hierarchical ordering for each different subband in the signal, and the hierarchical ordering sorts each subband according to the importance of the information it carries, and when writing the code stream According to the hierarchical ordering, the coding information of the corresponding sub-bands is written in order according to the importance from large to small, and the coding information of the sub-bands carrying more important information is written first, so that the code stream structure written is based on the carried information. The difference in importance is divided into different levels, forming a code stream with an embedded structure, which can transmit code streams with different levels according to network conditions and user requirements for service quality. During decoding, according to the arrangement order of each sub-band, the decoding information of each sub-band is decoded in sequence according to the magnitude of the importance, and the decoding information with higher importance is decoded first, so that even if the Part of the decoding information of low importance is missing in the code stream, and the technical solution provided by the embodiment of the present invention can also obtain a higher-quality decoding signal.

As shown in FIG. 2 , the encoding method provided by the embodiment of the present invention will be described in detail below.

The encoding method provided by the embodiment of the present invention is based on MLT transform. The signal described in the embodiment of the present invention includes at least one of a speech signal and an audio signal, and the signal can be expressed as x(n), where n represents a sampling point in the time domain label.

Step 201. Acquire MLT coefficients in a signal.

At the encoding end, firstly, add a sine window and a transformation factor to the speech and audio signals, and perform MLT transformation to obtain the MLT coefficient mlt(m). The transformation formula used is:

$\mathrm{<span\; class="notranslate">\; mlt</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; m</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; =</span>\underset{\mathrm{<span\; class="notranslate">\; no</span>}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 0</span>}}{\overset{\mathrm{<span\; class="notranslate">\; 2</span>}\mathrm{<span\; class="notranslate">\; N</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 1</span>}}{\mathrm{<span\; class="notranslate">\; \&Sigma;</span>}}}\sqrt{\frac{\mathrm{<span\; class="notranslate">\; 2</span>}}{\mathrm{<span\; class="notranslate">\; N</span>}}}\mathrm{<span\; class="notranslate">\; sin</span>}<span\; class="notranslate">\; (</span>\frac{\mathrm{<span\; class="notranslate">\; \&pi;</span>}}{\mathrm{<span\; class="notranslate">\; 2</span>}\mathrm{<span\; class="notranslate">\; N</span>}}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; no</span>}<span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; 0.5</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; )</span>\mathrm{<span\; class="notranslate">\; cos</span>}<span\; class="notranslate">\; (</span>\frac{\mathrm{<span\; class="notranslate">\; \&pi;</span>}}{\mathrm{<span\; class="notranslate">\; N</span>}}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; no</span>}<span\; class="notranslate">\; -</span>\frac{\mathrm{<span\; class="notranslate">\; N</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 1</span>}}{\mathrm{<span\; class="notranslate">\; 2</span>}}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; m</span>}<span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; 0.5</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; )</span>\mathrm{<span\; class="notranslate">\; x</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; no</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; )</span>$

in, $\sin \left(\frac{\mathrm{\&pi;}}{2N}(\mathrm{no}+0.5)\right)$ is the added sine window, $\cos \left(\frac{\mathrm{\&pi;}}{N}(\mathrm{no}-\frac{N-1}{2})(m+0.5)\right)$ is the added transformation factor, m represents the label of the sampling point in the frequency domain, and N represents the length of one frame of signal, that is, the total number of sampling points in one frame of signal. Because the previous frame information will be used when performing MLT transformation on the current frame signal, 2N sampling points are used in the formula (1), and the sampling points include the sampling points in the previous frame signal of the currently processed frame signal.

In the embodiment of the present invention, 320 sampling points are taken in one frame signal, that is, N=320, and 320 MLT coefficients can be obtained.

Step 202. Divide the MLT coefficients into at least two groups, and each group of MLT coefficients is a sub-band.

In the embodiment of the present invention, every 20 MLT coefficients are divided into a group, and as a subband, the embodiment of the present invention includes 16 subbands.

Step 203. Acquire amplitude envelope values of each subband according to the MLT coefficients in each subband.

In the embodiment of the present invention, the amplitude envelope value of the subband is the root mean square of the MLT coefficient contained in the subband, which can be obtained according to the following formula (2):

$\mathrm{<span\; class="notranslate">\; rms</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; r</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; =</span>\sqrt{\frac{\mathrm{<span\; class="notranslate">\; 1</span>}}{\mathrm{<span\; class="notranslate">\; 20</span>}}\underset{\mathrm{<span\; class="notranslate">\; no</span>}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 0</span>}}{\overset{\mathrm{<span\; class="notranslate">\; 19</span>}}{\mathrm{<span\; class="notranslate">\; \&Sigma;</span>}}}\mathrm{<span\; class="notranslate">\; mlt</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 20</span>}\mathrm{<span\; class="notranslate">\; r</span>}<span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; no</span>}<span\; class="notranslate">\; )</span>\mathrm{<span\; class="notranslate">\; mlt</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 20</span>}\mathrm{<span\; class="notranslate">\; r</span>}<span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; no</span>}<span\; class="notranslate">\; )</span>}<span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 2</span>}<span\; class="notranslate">\; )</span>$

Among them, r is the label of the subband, and rms(r) is the amplitude envelope value of the rth subband.

Step 204. Acquire quantized amplitude envelope values of the sub-bands according to the amplitude envelope values.

A quantization interval is determined according to the acquired maximum and minimum amplitude envelope values of each sub-band, and the quantization value of the amplitude envelope of each sub-band is acquired by using a scalar quantization method.

Step 205, using the quantized value of the amplitude envelope of each sub-band to determine the hierarchical order of each sub-band.

The hierarchical sorting provided by the embodiment of the present invention is an energy sorting, and the subband corresponding to the larger amplitude envelope quantization value has greater energy, so the embodiment of the present invention is based on the amplitude envelope quantization value of each subband The sub-bands are sorted in descending order, and the sub-bands are hierarchically sorted according to the order in which the corresponding amplitude envelope quantization values of the sub-bands are arranged in descending order.

The coding information of each sub-band includes: the amplitude envelope index value of each sub-band, the bit classification control information and the MLT coefficient quantization index value of each sub-band.

Step 206. Obtain the amplitude envelope index value of each sub-band according to the amplitude envelope quantization value of each sub-band.

In the embodiment of the present invention, Huffman coding is performed on the amplitude envelope quantization value of each sub-band to obtain the amplitude envelope quantization index value of each sub-band.

Step 207. Acquire the bit classification control information according to the coding rate and the amplitude envelope quantization value of each subband.

The total number of bits occupied by each frame of speech and audio signals is determined according to the encoding rate.

The bit classification control information in each frame occupies 4 bits, and the bit classification control information indicates the classification method of each subband, and the classification method is a method of bit allocation and vector division for each subband method. The method for obtaining the bit classification control information will be described in detail below.

The bit classification control information determines a classification method for each subband, and the classification method determines the number of bits occupied by each subband in the code stream and the vector division method of the MLT coefficients in each subband. The different classification situations are that the number of bits allocated to each sub-band and the division of the vectors in each sub-band are different.

Initially classify each sub-band, as shown in Table 1, each sub-band has 8 classifications to choose from (class 0 to class 7), and each class corresponds to the corresponding number of allocated bits in the table.

The bit allocation is carried out according to the quantization value of the amplitude envelope, and the importance of the subband with a larger quantization value of the amplitude envelope is relatively greater, so the classification method with a smaller label is selected to allocate more bits for the subband; otherwise, the magnitude envelope The sub-band with a smaller quantization value is less important, and the classification situation with a larger label is selected, and the number of bits allocated to the sub-band is less.

Table 1

Classification The number of bits required to apply this classification code 0 52 1 47 2 43 3 37 4 29 5 twenty two 6 16 7 0

According to the difference between the total number of bits occupied by each frame of speech and audio signals determined by the encoding rate and the bit classification control information (occupying 4 bits), the number of remaining encoding bits is obtained. The total number of bits allocated for each subband should be close to but not greater than the number of remaining coded bits. The method for determining the category to which each subband belongs is introduced below.

First determine an offset range, the embodiment of the present invention takes [-32, 31], select an initial offset in the offset range, and preliminarily determine the category to which a subband belongs by the following formula (3):

category(r)=MAX{0, MIN{7, (offset-rms′(r))/2}} (3)

Wherein, offset is the offset, rms'(r) is the amplitude envelope quantization value of subband r, and category(r) indicates the category (category 0 to category 7) to which subband r belongs.

Under the selected initial offset, the total number of bits allocated for all subbands is shown in equation (4):

$\mathrm{<span\; class="notranslate">\; expected</span>}<span\; class="notranslate">\; \_</span>\mathrm{<span\; class="notranslate">\; number</span>}<span\; class="notranslate">\; \_</span>\mathrm{<span\; class="notranslate">\; or</span>}<span\; class="notranslate">\; \_</span>\mathrm{<span\; class="notranslate">\; code</span>}<span\; class="notranslate">\; \_</span>\mathrm{<span\; class="notranslate">\; bits</span>}<span\; class="notranslate">\; =</span>\underset{\mathrm{<span\; class="notranslate">\; r</span>}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 0</span>}}{\overset{\mathrm{<span\; class="notranslate">\; number</span>}<span\; class="notranslate">\; \_</span>\mathrm{<span\; class="notranslate">\; of</span>}<span\; class="notranslate">\; \_</span>\mathrm{<span\; class="notranslate">\; region</span>}}{\mathrm{<span\; class="notranslate">\; \&Sigma;</span>}}}\mathrm{<span\; class="notranslate">\; expected</span>}<span\; class="notranslate">\; \_</span>\mathrm{<span\; class="notranslate">\; bits</span>}<span\; class="notranslate">\; \_</span>\mathrm{<span\; class="notranslate">\; table</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; category</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; r</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; )</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 4</span>}<span\; class="notranslate">\; )</span>$

Among them, expected_bits_table(category(r)) indicates the number of bits required when subband r belongs to category category(r), number_of_region is the total number of subbands, and expected_number_of_code_bits indicates the total number of bits allocated for all subbands.

Adjust the offset value in the range of [-32, 31], under different offsets, repeatedly calculate the total number of bits allocated for all subbands, and select the one closest to but not exceeding the number of remaining coded bits The classification method corresponding to the total number of bits of all subbands is used as the initial classification method.

Adjust the obtained initial classification method to find the most suitable classification method. When adjusting the classification, there are 16 kinds of adjustment classification methods. According to the number of remaining coding bits and the coding rate, those skilled in the art can obtain an optimal classification method through the methods disclosed in the prior art, thereby , determining the number of bits allocated for each subband and the vector division method for the MLT coefficients in each subband. The label corresponding to the optimal classification method, that is, the bit classification control information, occupies 4 bits in the code stream.

Step 208: Obtain quantization index values of MLT coefficients in each subband according to the MLT coefficients and amplitude envelope quantization values in each subband. The MLT coefficient quantization index value includes an MLT coefficient amplitude index value and an MLT coefficient sign index value, and this step specifically includes:

Step S1. Dividing the MLT coefficients in each subband by the corresponding amplitude envelope quantization value of each subband to obtain the normalized value of the MLT coefficient in each subband;

Step S2, quantizing the normalized value of the MLT coefficient in each subband, and obtaining the quantized value of the normalized value of the MLT coefficient in each subband;

Step S3, divide the normalized quantization values of the MLT coefficients in each subband into at least two vectors according to the bit classification control information, the vectors include magnitude and sign;

Step S4, obtaining the corresponding MLT coefficient magnitude index value of each vector in each subband according to the magnitude of each vector;

Step S5 , according to the signs of the vectors, obtain the MLT coefficient sign index value corresponding to each vector in each subband, and the MLT coefficient sign index value indicates the sign of its corresponding MLT coefficient magnitude index value.

Step 209, write the amplitude envelope index value of each sub-band into the code stream.

Step 210, write the bit classification control information into a code stream.

Step 211, according to the hierarchical ordering of the subbands and the bit classification control information, sequentially quantize the MLT coefficient index values of the vectors in the subbands, that is, the MLT coefficient amplitude index value and the MLT coefficient Symbol index value, written into the code stream.

After the code stream is written, the obtained code stream structure is shown in Figure 3, which includes the amplitude envelope index value of each sub-band, bit classification control information and the quantization of the MLT coefficients from the first bit to the last bit of the code stream. The index value has three parts.

The encoding method provided by the embodiment of the present invention ensures that the important information in the speech and audio signals is stored in the front part of the code stream by sorting the levels of the sub-bands, that is, the lower layer of the code stream, and the secondary important information is stored in the The rear part of the code stream, that is, the upper layer of the code stream.

The coding method provided by the embodiment of the present invention generates a code stream with a higher coding rate, which is relative to the code rate required for transmission in an actual network, and the higher coding rate is not less than The required code rate for transmission. When the code rate required in network transmission is lower than the encoding code rate, the code stream is truncated to obtain a code stream at the required code rate.

The encoding method provided by the embodiment of the present invention also includes:

According to the encoding code rate and the currently required code rate, truncate the MLT coefficient quantization index value in the code stream to obtain a truncation point. Acquire the required code stream according to the truncation point, the required code stream includes the amplitude envelope index value of each subband, bit classification control information and the MLT of each subband before the truncation point according to the hierarchical ordering Coefficient quantization index value.

There are three positions of the truncation point, as shown in Figure 4,

In the first case (as shown at A in FIG. 4 ), the truncation point is located at the junction of the MLT coefficient sign index value and the MLT coefficient amplitude index value.

In the second case (as shown at B in FIG. 4 ), the truncation point is located in the MLT coefficient sign index value.

In the third case (as shown at C in FIG. 4 ), the truncation point is located in the magnitude index value of the MLT coefficient.

In the embodiment of the present invention, the encoding end encodes the MLT coefficients in the 7 kHz bandwidth range at a code rate of 16 kbit/s (bits per second), and each frame has 320 bits of encoding information. As shown in Figure 5, if 8kbit/s transmission is required, the code stream of 320-bit information should be truncated, and the information of the first 160 bits should be retained, that is, the required code stream obtained after truncation is used for network communication, and after the truncation point is discarded 160bits information. There are three parts of information in the code stream: the amplitude envelope index value, the bit classification control information and the MLT coefficient quantization index value. The number of bits occupied by the first two parts is relatively small, far less than 160 bits, so the truncated The required code stream contains the basic information of the original voice and audio signal, which can guarantee the communication quality.

The embodiment of the present invention also provides a decoding method capable of obtaining a higher-quality decoded signal. A decoding method, as shown in Figure 6, the method comprises the following steps:

Step 601, obtaining decoding information from the code stream;

Step 602. Determine the hierarchical ordering of each subband in the signal according to the decoding information, and the signal includes at least two subbands;

Step 603: Decode the decoded information according to the hierarchical ordering, and obtain a decoded signal.

The following describes the embodiments of the present invention in detail.

Step 601, acquire decoding information from code stream.

The signal in the embodiment of the present invention includes at least one of a voice signal and an audio signal.

The decoding information includes: the amplitude envelope index value of each subband, the bit classification control information and the MLT coefficient quantization index value in each subband.

Step 602, the step of determining the hierarchical order of each subband in the signal according to the decoding information includes:

Step T1, obtaining the amplitude envelope quantization value of each sub-band according to the amplitude envelope index value of each sub-band;

In the embodiment of the present invention, since the Huffman coding method is used to obtain the quantized index value of the amplitude envelope during encoding, when decoding in the embodiment of the present invention, the Huffman code is decoded to recover Output the amplitude envelope quantization value of each subband.

Step T2. Determine the hierarchical order of each sub-band according to the amplitude envelope quantization value of each sub-band.

During decoding, the hierarchical ordering of the subbands is also to sort the subbands in descending order of the magnitude envelope quantization values of the subbands. Since the hierarchical order of the subbands obtained during decoding is the same as the hierarchical order of the subbands during encoding, it is ensured that accurate speech and audio signals can be obtained after decoding.

Step 603: Decode the decoded information according to the hierarchical ordering, and obtain a decoded signal. This step specifically includes:

The MLT coefficient quantization index value includes an MLT coefficient magnitude index value and an MLT coefficient sign index value.

Step U1. Obtain the classification mode of each sub-band according to the coding rate of the code stream, the bit classification control information and the quantized value of the amplitude envelope of each sub-band;

The bit classification control information indicates the label of the optimal classification method selected during encoding, according to the optimal classification method corresponding to the label, the coding rate of the code stream and the amplitude envelope of each sub-band The quantization value is used to obtain the classification method of the sub-bands, that is, to obtain the number of bits occupied by the sub-bands and the division method of the vectors in the sub-bands. The encoding rate of the code stream is the encoding rate used when generating the original embedded structure code stream during encoding.

Step U2, according to the hierarchical ordering of the subbands and the classification method of the subbands, using the MLT coefficient amplitude index value and the MLT coefficient sign index value in the subbands to obtain the subbands in the subbands MLT coefficient;

According to the hierarchical ordering of the sub-bands, the sub-band written into the code stream first, that is, the sub-band carrying more important information is first decoded, and then the sub-bands carrying less important information are sequentially decoded. The number of bits occupied by each subband and the division method of vectors in each subband are obtained from the classification method of each subband, and all MLT coefficient amplitude index values and MLT coefficient sign index values in each subband are obtained . The MLT coefficients in each subband are obtained by decoding using the amplitude and sign index values.

Step U3. Perform inverse modulation concatenation transformation on the MLT coefficients of each sub-band to obtain decoded speech and audio signals.

When network packet loss or truncation occurs in the code stream, resulting in incomplete quantization index values of the MLT coefficients, the decoding method provided by the embodiment of the present invention uses noise filling to obtain the sub-values corresponding to the missing quantization index values of the MLT coefficients The MLT coefficient of the band can be divided into two cases:

In the first case, when all the MLT coefficient sign index values are missing (as shown at A and C places in Figure 4), the amplitude envelope quantization value of the subband corresponding to the missing MLT coefficient sign index value, and The product of predetermined attenuation factors is used as the amplitude value of the MLT coefficient of the subband.

In the second case, when the MLT coefficient sign index value is partially missing (as at B in Figure 4), the amplitude envelope quantization value of the subband corresponding to the missing part of the MLT coefficient sign index value is compared with the predetermined The product of the attenuation factor is used as the missing part of the symbol index value of the MLT coefficient, that is, the MLT coefficient value of the right part of the symbol index where B is located. For the left part of the symbol index where B is located, the obtained symbol index is still used in combination with its corresponding amplitude index to decode its corresponding MLT coefficient. The above decoding method ensures that the effective information in the code stream is decoded to the maximum extent.

In the decoding method adopted in the embodiment of the present invention, when decoding the code stream, by obtaining the hierarchical ordering of the speech and audio signal sub-bands in the code stream, the sub-bands are obtained according to the information carried by the sub-bands. Arrangement order of importance from large to small, during decoding, according to the arrangement order of each sub-band, and in turn according to the magnitude of the importance, decode the decoding information of each sub-band, first decode the information with higher importance Therefore, even if some decoding information of low importance is missing in the code stream, a higher-quality decoding signal can be finally obtained.

As shown in Figure 7, an embodiment of the present invention also provides a codec system, the system includes an encoding device and a decoding device, and the encoding device includes:

A subband acquisition module 701, configured to acquire at least two subbands of a signal;

An amplitude envelope quantization value acquisition module 702, configured to acquire the amplitude envelope quantization value of the subband acquired by the subband acquisition module;

A hierarchical ordering and encoding information acquisition module 703, configured to acquire the hierarchical ordering of each subband and the encoding information of each subband by using the amplitude envelope quantization value acquired by the amplitude envelope quantization value acquisition module;

The code stream writing module 704 is configured to write the coding information of each sub-band acquired by the coding information obtaining module into the code stream according to the hierarchical order determined by the hierarchical order determining module.

The decoding device includes:

Decoding information acquisition module 705, configured to acquire decoding information from the code stream;

A hierarchical ordering acquisition module 706, configured to determine the hierarchical ordering of each subband in the signal according to the decoding information acquired by the decoding information acquisition module, and the signal includes at least two subbands;

The decoded signal obtaining module 707 is configured to decode the decoded information obtained by the decoding information obtaining module according to the hierarchical order obtained by the hierarchical order obtaining module, and obtain a decoded signal.

In the embodiment of the present invention, the signal includes at least one of a speech signal and an audio signal.

The layer sorting and coding information acquisition module includes a coding sorting module, configured to sort the subbands according to the descending order of the amplitude envelope quantization values of the subbands.

The hierarchical sorting acquisition module includes a decoding sorting module, configured to sort the subbands in descending order of the amplitude envelope quantization values of the subbands.

The embodiment of the present invention also provides an encoding device, as shown in Figure 8, the device includes:

A subband acquiring unit 81, configured to acquire at least two subbands of the signal;

An amplitude envelope quantization value acquisition unit 82, configured to acquire the amplitude envelope quantization value of the subband acquired by the subband acquisition unit;

A hierarchical ordering and encoding information acquisition unit 83, configured to acquire the hierarchical ordering of each subband and the encoding information of each subband by using the amplitude envelope quantization value acquired by the amplitude envelope quantization value acquisition module;

The code stream writing unit 84 is configured to write the encoding information of each sub-band obtained by the encoding information obtaining unit into the code stream according to the hierarchical ordering and the hierarchical order obtained by the encoding information obtaining unit.

The embodiments of the present invention are described in detail below.

In the embodiment of the present invention, the signal includes at least one of a speech signal and an audio signal.

As shown in Figure 9, the subband acquisition unit 81 includes:

MLT coefficient acquisition module 811, configured to acquire the MLT coefficient in the signal;

The subband division module 812 is configured to divide the MLT coefficients acquired by the MLT coefficient acquisition module into at least two groups, and each group of MLT coefficients is a subband.

The amplitude envelope quantization value acquisition unit 82 is configured to acquire the amplitude envelope quantization value of the sub-band acquired by the sub-band acquisition unit.

The hierarchical sorting provided by the embodiment of the present invention is an energy sorting, and the subband corresponding to the larger amplitude envelope quantization value has greater energy, so the embodiment of the present invention is based on the amplitude envelope quantization value of each subband The sub-bands are sorted in order from large to small, and the hierarchical order of the sub-bands is the order in which the quantized values of the amplitude envelopes are arranged from large to small, that is, the order of the sub-bands from the lower layer to the higher layer corresponds to the amplitude Largest to smallest ordering of envelope quantization values. The hierarchical ordering and encoding information acquisition unit 83 includes:

The hierarchy sorting determination module 831 is configured to sort the subbands in descending order of the amplitude envelope quantization values of the subbands.

The encoding device provided by the embodiment of the present invention, through the hierarchical order determination unit, ensures that the important information in the voice and audio signals is stored in the front part of the code stream, that is, the lower layer of the code stream, and the secondary important information is stored in the code stream The rear part of the code stream is the upper layer.

In the embodiment of the present invention, the encoding information of each subband includes: the amplitude envelope index value of each subband, the bit classification control information and the MLT coefficient quantization index value in each subband, and the hierarchical ordering and encoding information Acquisition unit 83 includes:

An amplitude envelope index value acquisition module 832, configured to acquire the amplitude envelope index value of each sub-band according to the amplitude envelope quantization value of each sub-band;

A bit classification control information acquisition module 833, configured to obtain the bit classification control information according to the coding rate and the quantized value of the amplitude envelope of each subband;

The MLT coefficient quantization index value acquisition module 834 is configured to acquire the MLT coefficient in each subband, the amplitude envelope quantization value, and the bit classification control information obtained by the bit classification control information acquisition module. MLT coefficient quantization index value.

In this embodiment of the present invention, the MLT coefficient quantization index value includes an MLT coefficient magnitude index value and an MLT coefficient sign index value, and the MLT coefficient quantization index value acquisition module 933 further includes:

A normalized value acquisition module, configured to divide the MLT coefficients in the subbands by the corresponding amplitude envelope quantization values of the subbands to obtain the normalized values of the MLT coefficients in the subbands;

A normalized quantized value obtaining module, configured to obtain the quantized value of the normalized value of the MLT coefficient obtained by the normalized value obtaining module;

A vector division module, configured to divide the normalized quantized value obtained by the normalized quantized value obtaining module into at least two vectors according to the bit classification control information, and the vector includes magnitude and sign;

The MLT coefficient magnitude index value acquisition module is used to obtain the MLT coefficient magnitude index value corresponding to each vector in each subband according to the magnitude of each vector;

The MLT coefficient sign index value acquiring module is configured to acquire the MLT coefficient sign index value corresponding to each vector in each subband according to the sign of each vector.

The code stream writing unit 84 includes:

The amplitude envelope index value writing module 841 is used to write the amplitude envelope index value of each sub-band into the code stream;

A bit classification control information writing module 842, configured to write the bit classification control information into a code stream;

The MLT coefficient quantization index value writing module 843 is used to sequentially write the MLT coefficient amplitude index value and the MLT coefficient of each vector in each subband according to the hierarchical ordering of each subband and the bit classification control information The symbol index value is written to the codestream.

The coding device provided by the embodiment of the present invention generates a code stream with a higher coding rate, which is relative to the code rate required for transmission in an actual network, and the higher coding rate is not less than The required code rate for transmission. When the code rate required in network transmission is lower than the encoding code rate, the code stream is truncated to obtain a code stream at the required code rate. When the code rate required in actual transmission is less than the code rate, the encoding device also includes:

The truncation point acquisition unit is used to truncate the MLT coefficient quantization index value in the code stream according to the encoding code rate and the current required code rate to obtain a truncation point; there are three positions of the truncation point, as shown in FIG. 4,

In the first case (as shown at A in FIG. 4 ), the truncation point is located at the junction of the MLT coefficient sign index value and the MLT coefficient amplitude index value.

In the second case (as shown at B in FIG. 4 ), the truncation point is located in the MLT coefficient sign index value.

In the third case (as shown at C in FIG. 4 ), the truncation point is located in the magnitude index value of the MLT coefficient.

The required code stream acquisition unit is used to obtain the required code stream according to the truncation point obtained by the truncation point acquisition unit, and the required code stream includes the amplitude envelope index value of each sub-band, bit classification control information and according to The hierarchy sorts the MLT coefficient quantization index values of the subbands before the truncation point.

The encoding device provided by the embodiment of the present invention firstly determines the hierarchical ordering for each different subband in the speech and audio signal, and the said hierarchical ordering sorts each subband according to the importance of the information it carries. When the code stream is sorted according to the levels, the coded information of the corresponding subbands is written in order according to the importance from large to small, and the coded information of the subbands carrying more important information is written first, so that the code stream structure is written according to The importance of carrying information is divided into different levels, forming a code stream with an embedded structure, which can transmit code streams with different levels according to network conditions and user requirements for service quality, and is suitable for current networks Communication; at the same time, when the code rate required in network transmission is lower than the encoding code rate, the code stream is truncated to obtain the code stream at the required code rate. Since the code stream has an embedded structure, the truncation The required code stream obtained later contains the basic information of the original voice and audio signal, which can guarantee the communication quality.

The embodiment of the present invention also provides a decoding device, as shown in Figure 10, the device includes:

Decoding information acquiring unit 101, configured to acquire decoding information from code stream;

A hierarchy order acquisition unit 102, configured to determine the hierarchy order of each subband in the speech and audio signal according to the decoding information acquired by the decoding information acquisition unit, the speech and audio signal including at least two subbands;

The decoded signal acquiring unit 103 is configured to decode the decoded information acquired by the decoding information acquiring unit according to the hierarchical order acquired by the hierarchical order acquiring unit, and acquire decoded speech and audio signals.

The following describes the embodiments of the present invention in detail.

In the embodiment of the present invention, the decoding information includes: the amplitude envelope index value of each subband, the bit classification control information and the MLT coefficient quantization index value in each subband;

As shown in Figure 12, the hierarchical ordering acquisition unit 102 includes:

An amplitude envelope quantization value acquisition module 1021, configured to acquire the amplitude envelope quantization value of each sub-band according to the amplitude envelope index value of each sub-band;

The hierarchical order obtaining module 1022 is configured to determine the hierarchical order of each sub-band according to the amplitude envelope quantization value obtained by the amplitude envelope quantized value obtaining module.

The MLT coefficient quantization index value includes an MLT coefficient magnitude index value and an MLT coefficient sign index value;

The decoded signal acquisition unit 103 includes:

A classification acquisition module 1031, configured to acquire the classification mode of each sub-band according to the coding rate of the code stream, the bit classification control information, and the quantized value of the amplitude envelope of each sub-band;

The MLT coefficient acquisition module 1032 is configured to acquire the MLT coefficient amplitude index value and the MLT coefficient sign index value in each sub-band according to the hierarchical ordering of the sub-bands and the classification method of the sub-bands to obtain the MLT coefficients in each subband;

The decoded signal acquisition module 1033 is configured to perform inverse modulation concatenated transformation on the MLT coefficients of each subband to acquire decoded speech and audio signals.

When network packet loss or truncation occurs in the code stream, causing the MLT coefficient quantization index value to be incomplete, the decoding device provided by the embodiment of the present invention uses noise padding to obtain the subfields corresponding to the missing MLT coefficient quantization index value. The MLT coefficient of band, described MLT coefficient obtaining module comprises:

The MLT coefficient acquisition module of the missing index value is used to obtain the MLT coefficients of the sub-bands corresponding to the missing MLT coefficient quantization index values by using noise filling when the quantization index value of the MLT coefficient is incomplete, which can be divided into Two situations:

In the first case, when all the MLT coefficient sign index values are missing (as shown at A and C places in Figure 4), the MLT coefficient acquisition module of the missing index value corresponding to the missing MLT coefficient sign index value The product of the quantized value of the amplitude envelope of the subband and the predetermined attenuation factor is used as the amplitude value of the MLT coefficient of the subband.

In the second case, when the MLT coefficient sign index value is partially missing (as at B in Figure 4), the MLT coefficient acquisition module of the missing index value will use the subband corresponding to the missing part of the MLT coefficient sign index value The product of the amplitude envelope quantization value of , and a predetermined attenuation factor is used as the missing part of the sign index value of the MLT coefficient, that is, the MLT coefficient value of the right part of the sign index where B is located. For the left part of the symbol index where B is located, the obtained symbol index is still used in combination with its corresponding amplitude index to decode its corresponding MLT coefficient. The above decoding method ensures that the effective information in the code stream is decoded to the maximum extent.

When the decoding device provided by the embodiment of the present invention decodes the code stream, by obtaining the hierarchical ordering of the subbands of the voice and audio signals in the code stream, the subbands are obtained according to the importance of the information they carry. Arrangement order from large to small, when decoding, according to the arrangement order of each sub-band, the decoding information of each sub-band is decoded according to the magnitude of the importance, thus ensuring accurate speech and audio by decoding At the same time, the decoded information with higher importance is decoded first, so that even if some decoded information with lower importance is missing in the code stream, a higher-quality decoded signal is finally obtained.

Those of ordinary skill in the art can understand that all or part of the steps in the above embodiments can be implemented through program instructions and related hardware. The software corresponding to the embodiments may be stored in a computer-readable medium.

Of course, there are many other embodiments of the present invention, and those skilled in the art can make various corresponding changes and modifications according to the embodiments of the present invention without departing from the spirit and essence of the embodiments of the present invention. , but these corresponding changes and modifications should all belong to the protection scope of the appended claims of the embodiments of the present invention.

Claims (23)

1. An encoding method, characterized in that the method comprises: acquiring at least two subbands of the signal; Acquiring the amplitude envelope quantization value of each sub-band; Using the amplitude envelope quantization value of each sub-band to obtain the hierarchical ordering of each sub-band and the coding information of each sub-band; the coding information of each sub-band includes: the amplitude envelope index value of each sub-band , the bit classification control information and the MLT coefficient quantization index value in each subband; Writing the encoding information of each subband into a code stream according to the hierarchical ordering of each subband; wherein, The step of obtaining the hierarchical order of the subbands and the coding information of the subbands by using the quantized values of the amplitude envelopes of the subbands includes: obtaining the quantized values of the amplitude envelopes of the subbands The amplitude envelope index value of each sub-band; obtain the bit classification control information according to the encoding code rate and the amplitude envelope quantization value of each sub-band; according to the MLT coefficient and amplitude envelope quantization in each sub-band value and the bit classification control information to obtain the MLT coefficient quantization index value in each sub-band; the MLT coefficient quantization index value includes the MLT coefficient magnitude index value and the MLT coefficient sign index value, wherein, The step of obtaining the MLT coefficient quantization index value in each subband according to the MLT coefficient and amplitude envelope quantization value in each subband and the bit classification control information includes: Divide the MLT coefficient by the quantized value of the amplitude envelope of the corresponding sub-bands to obtain the normalized value of the MLT coefficient in the sub-bands; obtain the quantized value of the normalized value of the MLT coefficient in the sub-bands ; According to the bit classification control information, the quantization values of the normalized values of the MLT coefficients in the sub-bands are divided into at least two vectors, the vectors include magnitude and sign; according to the magnitude of each vector, the obtained Describe the corresponding MLT coefficient amplitude index value of each vector in each sub-band; Acquire the corresponding MLT coefficient sign index value of each vector in the described each sub-band according to the sign of each vector; The step of writing the encoding information of the subbands into the code stream according to the hierarchical ordering of the subbands includes: writing the amplitude envelope index values of the subbands into the code stream; classifying the bits Write the control information into the code stream; write the MLT coefficient amplitude index value and the MLT coefficient sign index value of each vector in the each subband sequentially according to the hierarchical ordering of each subband and the bit classification control information stream. 2. The encoding method according to claim 1, wherein the step of obtaining at least two subbands of the signal comprises: Obtaining the MLT coefficients of the modulated concatenated transform of the signal; The MLT coefficients are divided into at least two groups, and each group of MLT coefficients is a sub-band. 3. The coding method according to claim 2, characterized in that, the step of obtaining the hierarchical ordering of each sub-band and the coding information of each sub-band by using the amplitude envelope quantization value of each sub-band include: The subbands are sorted in descending order of the amplitude envelope quantization values of the subbands. 4. The encoding method according to claim 1 or 3, further comprising: According to the encoding code rate and the current required code rate, truncate the MLT coefficient quantization index value in the code stream to obtain a truncation point; According to the truncation point, the required code stream is obtained, and the required code stream includes the amplitude envelope index value of each subband, bit classification control information and the ordering of the subbands before the truncation point according to the hierarchy. MLT coefficient quantization index value. 5. The encoding method according to claim 4, wherein the truncation point is located in the magnitude index value of the MLT coefficient, or in the sign index value of the MLT coefficient, or in the sign index value of the MLT coefficient Junction with MLT coefficient magnitude index value. 6. The encoding method according to claim 1 or 2, wherein the signal comprises at least one of a voice signal and an audio signal. 7. A method for decoding the code stream obtained by the encoding method according to claim 1, characterized in that the method comprises: Obtain decoding information from the code stream; the decoding information includes: the amplitude envelope index value of each sub-band, the bit classification control information and the MLT coefficient quantization index value in each sub-band; determining a hierarchical ordering of subbands in a signal according to the decoding information, the signal including at least two subbands; Decoding the decoded information according to the hierarchical ordering to obtain a decoded signal; wherein, The step of determining the hierarchical ordering of each subband in the signal according to the decoding information includes: obtaining the amplitude envelope quantization value of each subband according to the amplitude envelope index value of each subband; The magnitude envelope quantization value of the band determines the hierarchical ordering of each subband in the signal; The MLT coefficient quantization index value includes an MLT coefficient amplitude index value and an MLT coefficient sign index value; the step of decoding the decoding information according to the hierarchical ordering, and obtaining the decoded signal includes: according to the code stream Acquire the classification mode of the sub-bands according to the coding rate of the sub-bands, the bit classification control information and the quantized value of the amplitude envelope of the sub-bands; according to the hierarchical ordering of the sub-bands and the classification of the sub-bands method, using the MLT coefficient amplitude index value and the MLT coefficient sign index value in each sub-band to obtain the MLT coefficients in each sub-band; performing inverse modulation concatenation transformation on the MLT coefficients in each sub-band to obtain decoded signal. 8. The decoding method according to claim 7, wherein the step of determining the hierarchical ordering of each subband in the signal according to the decoding information comprises: Acquiring the amplitude envelope quantization value of each subband according to the amplitude envelope index value of each subband; The subbands are sorted in descending order of the amplitude envelope quantization values of the subbands. 9. The decoding method according to claim 8, wherein, according to the hierarchical ordering of the subbands and the classification method of the subbands, the MLT coefficient magnitude index value in the subbands is used And MLT coefficient symbol index value, the step of obtaining the MLT coefficient in each subband includes: When the MLT coefficient quantization index value is incomplete, noise filling is used to obtain the MLT coefficients of the sub-bands corresponding to the missing MLT coefficient quantization index value. 10. The decoding method according to claim 9, wherein the step of using noise filling to obtain the missing MLT coefficient quantization index value corresponding to each sub-band MLT coefficient comprises: When all the MLT coefficient sign index values are missing, the product of the amplitude envelope quantization value of the subband corresponding to the missing MLT coefficient sign index value and a predetermined attenuation factor is used as the MLT coefficient of the subband the amplitude value; When the MLT coefficient sign index value is partially missing, the product of the amplitude envelope quantization value of the subband corresponding to the missing part of the MLT coefficient sign index value and a predetermined attenuation factor is used as the MLT coefficient sign index value The magnitude value of the MLT coefficient of the subband corresponding to the missing part. 11. The decoding method according to any one of claims 7 to 10, wherein the signal includes at least one of a speech signal and an audio signal. 12. An encoding device, characterized in that the device comprises: a subband acquisition unit, configured to acquire at least two subbands of the signal; an amplitude envelope quantization value acquisition unit, configured to acquire the amplitude envelope quantization value of the subband acquired by the subband acquisition unit; a hierarchical ordering and coding information acquisition unit, configured to obtain the hierarchical ordering of the subbands and the coding information of the subbands by using the amplitude envelope quantization values of the subbands; A code stream writing unit, configured to write the coded information of each sub-band into a code stream according to the hierarchical order obtained by the hierarchical ordering and coding information acquisition unit; the coded information of each sub-band includes: each sub-band The amplitude envelope index value of the band, the bit classification control information and the MLT coefficient quantization index value in each sub-band, the hierarchical ordering and encoding information acquisition unit includes: The amplitude envelope index value acquisition module is used to acquire the amplitude envelope index value of each sub-band according to the amplitude envelope quantization value of each sub-band; The amplitude envelope quantization value of each sub-band obtains the bit classification control information; the MLT coefficient quantization index value acquisition module is used to control the bit classification according to the MLT coefficient in each sub-band, the amplitude envelope quantization value and the bit classification control The bit classification control information acquired by the information acquisition module acquires the MLT coefficient quantization index value in each subband; wherein, The MLT coefficient quantization index value includes an MLT coefficient magnitude index value and an MLT coefficient sign index value, and the MLT coefficient quantization index value acquisition module includes: a normalization value acquisition module, which is used to convert the MLT coefficients in each subband to Divided by the corresponding quantized value of the amplitude envelope of each sub-band to obtain the normalized value of the MLT coefficient in each sub-band; the normalized quantized value obtaining module is used to obtain the normalized value obtaining module The quantized value of the normalized value of the obtained MLT coefficient; the vector division module is used to divide the quantized value of the normalized value obtained by the normalized quantized value obtaining module into at least two according to the bit classification control information Vector, the vector includes magnitude and sign; MLT coefficient magnitude index value acquisition module, used to obtain the MLT coefficient magnitude index value corresponding to each vector in each subband according to the magnitude of each vector; MLT coefficient sign index value acquisition A module, configured to obtain the MLT coefficient sign index values corresponding to the vectors in the subbands according to the signs of the vectors; The code stream writing unit includes: an amplitude envelope index value writing module for writing the amplitude envelope index values of the sub-bands into the code stream; a bit classification control information writing module for writing the Write the bit classification control information into the code stream; the MLT coefficient quantization index value writing module is used to sequentially write the vectors in the subbands according to the hierarchical ordering of the subbands and the bit classification control information. The MLT coefficient magnitude index value and the MLT coefficient sign index value are written into the code stream. 13. The encoding device according to claim 12, wherein the subband acquisition unit comprises: MLT coefficient obtaining module, used for obtaining the MLT coefficient of described signal; A subband division module, configured to divide the MLT coefficients acquired by the MLT coefficient acquisition module into at least two groups, and each group of MLT coefficients is a subband. 14. The encoding device according to claim 12 or 13, wherein the hierarchical ordering and encoding information acquisition unit comprises: The hierarchy sorting determination module is configured to sort the subbands in descending order of the amplitude envelope quantization values of the subbands. 15. The encoding device according to claim 12 or 13, wherein the device further comprises: A truncation point acquisition unit, configured to truncate the MLT coefficient quantization index value in the code stream according to the encoding code rate and the current required code rate, to obtain a truncation point; The required code stream acquisition unit is used to obtain the required code stream according to the truncation point obtained by the truncation point acquisition unit, and the required code stream includes the amplitude envelope index value of each sub-band, bit classification control information and according to The hierarchy sorts the MLT coefficient quantization index values of the subbands before the truncation point. 16. The encoding device according to claim 12 or 13, wherein the signal comprises at least one of a speech signal and an audio signal. 17. A device for decoding the code stream obtained by the encoding device according to claim 12, characterized in that the device comprises: a decoding information obtaining unit, configured to obtain decoding information from the code stream; A hierarchical ordering acquisition unit, configured to determine the hierarchical ordering of each subband in the signal according to the decoding information acquired by the decoding information acquisition unit, and the signal includes at least two subbands; A decoding signal acquisition unit, configured to decode the decoding information acquired by the decoding information acquisition unit according to the hierarchical ordering acquired by the hierarchical ordering acquisition unit, and obtain a decoded signal; wherein, The decoding information includes: the amplitude envelope index value of each subband, the bit classification control information and the MLT coefficient quantization index value in each subband; the hierarchical ordering acquisition unit includes: an amplitude envelope quantization value acquisition module for Acquiring the amplitude envelope quantization value of each sub-band according to the amplitude envelope index value of each sub-band; the hierarchy sorting acquisition module is used to obtain the amplitude envelope quantization value obtained by the amplitude envelope quantization value acquisition module, determining the hierarchical ordering of each subband; The MLT coefficient quantization index value includes an MLT coefficient magnitude index value and an MLT coefficient sign index value; the decoded signal acquisition unit includes: The classification acquisition module is used to obtain the classification mode of each sub-band according to the coding rate of the code stream, the bit classification control information and the amplitude envelope quantization value of each sub-band; the MLT coefficient acquisition module uses According to the hierarchical ordering of the sub-bands and the classification method of the sub-bands, the MLT coefficients in the sub-bands are acquired by using the MLT coefficient amplitude index value and the MLT coefficient sign index value in the sub-bands ; A decoded signal acquisition module, configured to perform inverse modulation concatenation transformation on the MLT coefficients of each sub-band to acquire decoded signals. 18. The decoding device according to claim 17, wherein the hierarchical ordering acquisition module is configured to sort the subbands in descending order of the magnitude envelope quantization values of the subbands. 19. The decoding device according to claim 17, wherein the MLT coefficient acquisition module comprises: The MLT coefficient acquisition module for missing index values is configured to use noise filling to acquire the MLT coefficients of the sub-bands corresponding to the missing MLT coefficient quantization index values when the quantization index values of the MLT coefficients are incomplete. 20. The decoding device according to any one of claims 17 to 19, wherein the signal comprises at least one of a voice signal and an audio signal. 21. A codec system, characterized in that the system includes an encoding device and a decoding device, and the encoding device includes: A subband acquisition module, configured to acquire at least two subbands of the signal; An amplitude envelope quantization value acquisition module, configured to acquire the amplitude envelope quantization value of the subband acquired by the subband acquisition module; A hierarchical ordering and coding information acquisition module, configured to obtain the hierarchical ordering of each sub-band and the encoding information of each sub-band by using the amplitude envelope quantization value acquired by the amplitude envelope quantization value acquisition module; A code stream writing module, configured to write the encoding information of each sub-band acquired by the encoding information acquisition module into the code stream according to the hierarchical ordering obtained by the hierarchical ordering and encoding information acquisition module; wherein, each sub-band The encoding information of the band includes: the amplitude envelope index value of each subband, the bit classification control information and the MLT coefficient quantization index value in each subband, and the hierarchical ordering and encoding information acquisition module includes: The amplitude envelope index value acquisition module is used to acquire the amplitude envelope index value of each sub-band according to the amplitude envelope quantization value of each sub-band; The amplitude envelope quantization value of each sub-band obtains the bit classification control information; the MLT coefficient quantization index value acquisition module is used to control the bit classification according to the MLT coefficient in each sub-band, the amplitude envelope quantization value and the bit classification control The bit classification control information acquired by the information acquisition module acquires the MLT coefficient quantization index value in each subband; Wherein, the MLT coefficient quantization index value includes an MLT coefficient magnitude index value and an MLT coefficient sign index value, and the MLT coefficient quantization index value acquisition module includes: a normalization value acquisition module, which is used to convert the The MLT coefficient is divided by the amplitude envelope quantization value of the corresponding sub-bands to obtain the normalized value of the MLT coefficient in the sub-bands; the normalized quantized value acquisition module is used to obtain the normalized value The quantized value of the normalized value of the MLT coefficient obtained by the obtaining module; the vector division module is used to divide the quantized value of the normalized value obtained by the normalized quantized value obtaining module into at least Two vectors, the vector includes magnitude and sign; MLT coefficient magnitude index value acquisition module, used to obtain the corresponding MLT coefficient magnitude index value of each vector in each subband according to the magnitude of each vector; MLT coefficient sign index A value acquisition module, configured to acquire the MLT coefficient sign index value corresponding to each vector in each subband according to the sign of each vector; Wherein, the code stream writing module includes: The amplitude envelope index value writing module is used to write the amplitude envelope index value of each subband into the code stream; the bit classification control information writing module is used to write the bit classification control information into the code stream; The MLT coefficient quantization index value writing module is used to sequentially write the MLT coefficient magnitude index value and MLT coefficient sign of each vector in each subband according to the hierarchical ordering of each subband and the bit classification control information The index value is written into the code stream; The decoding device includes: A decoding information acquisition module is used to obtain decoding information from the code stream; A hierarchical ordering acquisition module, configured to determine the hierarchical ordering of each subband in the signal according to the decoding information acquired by the decoding information acquisition module, the signal including at least two subbands; A decoding signal acquisition module, configured to decode the decoding information acquired by the decoding information acquisition module according to the hierarchical ordering acquired by the hierarchical ordering acquisition module, and obtain a decoded signal; wherein the decoding information includes: each subband The magnitude envelope index value of the magnitude envelope, the bit classification control information and the MLT coefficient quantization index value in each sub-band; the hierarchical ordering acquisition module includes: an amplitude envelope quantization value acquisition module, for The envelope index value obtains the amplitude envelope quantization value of each sub-band; the hierarchical ordering acquisition module is used to determine the hierarchical ordering of each sub-band according to the amplitude envelope quantization value acquired by the amplitude envelope quantization value acquisition module; The MLT coefficient quantization index value includes an MLT coefficient magnitude index value and an MLT coefficient sign index value; the decoded signal acquisition module includes: a classification acquisition module, which is used to classify the control information according to the code rate of the code stream and the bit classification and the quantized value of the amplitude envelope of each sub-band to obtain the classification method of each sub-band; the MLT coefficient acquisition module is used to use the classification method of each sub-band according to the hierarchical ordering of each sub-band and the classification method of each sub-band The MLT coefficient amplitude index value and the MLT coefficient sign index value in each sub-band are described, and the MLT coefficients in the various sub-bands are obtained; the decoded signal acquisition module is used to reverse-modulate and concatenate the MLT coefficients of the various sub-bands Transform to get the decoded signal. 22. The codec system according to claim 21, wherein the signal comprises at least one of a speech signal and an audio signal. 23. The encoding and decoding system according to claim 21 or 22, wherein the hierarchical ordering and encoding information acquisition module includes an encoding ordering module, which is used to quantize the amplitude envelopes of the sub-bands from large to The small order sorts each subband; The hierarchical sorting acquisition module includes a decoding sorting module, configured to sort the subbands in descending order of the amplitude envelope quantization values of the subbands.

Images