[go: up one dir, main page]

WO2009121280A1 - 编码方法及装置、解码方法及装置 - Google Patents

编码方法及装置、解码方法及装置 Download PDF

Info

Publication number
WO2009121280A1
WO2009121280A1 PCT/CN2009/071022 CN2009071022W WO2009121280A1 WO 2009121280 A1 WO2009121280 A1 WO 2009121280A1 CN 2009071022 W CN2009071022 W CN 2009071022W WO 2009121280 A1 WO2009121280 A1 WO 2009121280A1
Authority
WO
WIPO (PCT)
Prior art keywords
time domain
envelope information
frame
domain envelope
current frame
Prior art date
Application number
PCT/CN2009/071022
Other languages
English (en)
French (fr)
Inventor
凯瑟·本特
张德明
张琦
Original Assignee
华为技术有限公司
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by 华为技术有限公司 filed Critical 华为技术有限公司
Publication of WO2009121280A1 publication Critical patent/WO2009121280A1/zh

Links

Classifications

    • GPHYSICS
    • G10MUSICAL INSTRUMENTS; ACOUSTICS
    • G10LSPEECH ANALYSIS TECHNIQUES OR SPEECH SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING TECHNIQUES; SPEECH OR AUDIO CODING OR DECODING
    • G10L19/00Speech or audio signals analysis-synthesis techniques for redundancy reduction, e.g. in vocoders; Coding or decoding of speech or audio signals, using source filter models or psychoacoustic analysis
    • G10L19/04Speech or audio signals analysis-synthesis techniques for redundancy reduction, e.g. in vocoders; Coding or decoding of speech or audio signals, using source filter models or psychoacoustic analysis using predictive techniques

Definitions

  • the present invention relates to data coding techniques, and in particular, to a coding method and apparatus, a decoding method and apparatus. Background technique
  • a loop is established between the two session terminals, through which the transmitting end transmits the data packet containing the voice information to the receiving end, and then synthesizes the voice signal at the receiving end. If a packet loss occurs during the transmission, it is necessary to recover the voice information contained in the lost packet by using the previously received data packet or other methods, which may cause an auditory discomfort.
  • IP Internet Protocol
  • FEC forward error correction
  • Forward Error Correction strategy, which better solves the problem of voice quality degradation when frames are lost.
  • the forward error correction strategy for the narrowband and wideband parts is used for the ultra-wideband part effect. Not necessarily good, there is currently no better forward error correction strategy for the signal of the ultra-wideband part.
  • An existing method for processing an ultra-wideband signal is such that if a frame is lost or damaged, the decoded parameter of the previous frame is repeated, that is, the decoded parameter of the previous frame is used as the parameter of the current frame. Subsequent decoding processing is performed.
  • the speech signal has a short-term stationary characteristic
  • the signals of two adjacent frames have a certain correlation, but the above method directly uses the parameters of the previous frame as the parameters of the current frame for decoding, that is, does not contain any information that the current frame should have. , the signal finally decoded will still have a more obvious discontinuity, and the hearing is uncomfortable.
  • Embodiments of the present invention provide an encoding method and apparatus, a decoding method, and an apparatus, which are implemented by using the present invention.
  • the technical solution provided by the example can use part of the information of the current frame for decoding when the current frame is lost or damaged.
  • An embodiment of the present invention provides an encoding method, including:
  • Encoding the current frame requires time domain envelope information for encoding transmission
  • Transmitting the encoded current frame requires encoding the transmitted time domain envelope information.
  • the embodiment of the invention provides a decoding method, including:
  • Decoding to obtain the time domain envelope information of the currently received frame code transmission, and the time domain envelope information transmitted by the currently received frame code includes the current frame part in the time domain envelope information of the currently received frame. And a subsequent frame portion of the time domain envelope information of the frame received before receiving the currently received frame; extracting the current received from the time domain envelope information transmitted by the currently received frame code The current frame portion of the time domain envelope information of the frame;
  • An embodiment of the present invention provides another decoding method, including:
  • Decoding to obtain time domain envelope information transmitted by the frame encoding received after the currently received frame, the encoded time domain envelope information including the current frame in the time domain envelope information of the subsequently received frame And a subsequent frame portion of the time domain envelope information of the currently received frame;
  • the time domain envelope information of the currently received frame is obtained using the subsequent frame portion.
  • An embodiment of the present invention provides an encoding apparatus, including:
  • a time domain envelope information obtaining unit configured to obtain quantized time domain envelope information of the current frame
  • a splitting unit configured to divide the quantized time domain envelope information acquired by the time domain envelope information acquiring unit into at least two parts, including at least one current frame part and at least one subsequent frame part;
  • a combining unit configured to combine the current frame portion and a subsequent frame portion of the previous frame data of the current frame to obtain time domain envelope information that the current frame needs to be encoded and transmitted, the number of the previous frame data and the subsequent frame The same number;
  • the code sending unit is configured to encode the time domain envelope information of the current frame obtained by combining the combination unit, and transmit the encoded time domain envelope information of the current frame.
  • An embodiment of the present invention provides a decoding apparatus, including:
  • a decoding unit configured to decode time domain envelope information of the currently received frame code transmission, where time domain envelope information transmitted by the currently received frame code includes time domain envelope information of the currently received frame The current frame portion of the current frame portion of the time domain envelope information of the frame received before receiving the currently received frame;
  • an extracting unit configured to extract, from the time domain envelope information transmitted by the currently received frame encoding, a current frame portion in the time domain envelope information of the currently received frame;
  • An acquiring unit configured to acquire a subsequent frame portion in time domain envelope information of the currently received frame; a combining unit, configured to combine a current frame portion and a current frame portion in the time domain envelope information of the currently received frame The subsequent frame part in the time domain envelope information of the currently received frame obtains time domain envelope information of the currently received frame.
  • An embodiment of the present invention provides another decoding apparatus, including:
  • a decoding unit configured to decode time domain envelope information transmitted by the frame encoding received after the currently received frame, where the encoded time domain envelope information includes a time domain envelope of the subsequently received frame a current frame portion of the information and a subsequent frame portion of the time domain envelope information of the currently received frame;
  • an extracting unit configured to extract, in the time domain envelope information transmitted by the frame encoding received later, a subsequent frame portion in the time domain envelope information of the currently received frame;
  • the time domain envelope information obtaining unit is configured to obtain time domain envelope information of the currently received frame by using the subsequent frame portion.
  • the embodiment of the invention provides a codec system, including:
  • An encoding device configured to obtain quantized time domain envelope information of a current frame; and divide the quantized time domain envelope information into at least two parts, including at least one current frame part and at least one subsequent frame part And combining the current frame portion and the subsequent frame portion of the previous frame data of the current frame to obtain time domain envelope information that the current frame needs to be encoded and transmitted, the number of the previous frame data being the same as the number of the subsequent frames
  • the encoded current frame needs to encode the transmitted time domain envelope information, and the encoded current frame needs to encode the transmitted time domain envelope information;
  • a decoding device configured to decode time domain envelope information of the currently received frame code transmission; extract, from the time domain envelope information transmitted by the currently received frame code, the currently received frame a current frame portion in the time domain envelope information; acquiring a subsequent frame portion in the time domain envelope information of the current frame; combining a current frame portion of the currently received frame with a subsequent of the currently received frame The frame portion obtains time domain envelope information of the currently received frame.
  • An embodiment of the present invention provides another codec system, including:
  • An encoding device configured to obtain quantized time domain envelope information of a current frame; dividing the quantized time domain envelope information into at least two parts, including at least one current frame part and at least one subsequent frame part; a current frame portion and a subsequent frame portion of the previous frame data of the current frame, obtaining time domain envelope information that the current frame needs to be encoded and transmitted, the number of the previous frame data being the same as the number of the subsequent frames;
  • the frame needs to encode the transmitted time domain envelope information
  • the encoded current frame needs to encode the transmitted time domain envelope information
  • Decoding means configured to decode time domain envelope information of the frame code transmission received after the currently received frame; extracting the current receipt from the time domain envelope information transmitted by the frame coding received thereafter a subsequent frame portion of the time domain envelope information of the received frame; obtaining the time domain envelope information of the currently received frame using a subsequent frame portion of the time domain envelope information of the currently received frame.
  • the present invention also provides a computer program product comprising computer program code, wherein when the computer program code is executed by a computer, the computer program code can cause the computer to execute an encoding method or a decoding method Any of the steps.
  • the present invention also provides a computer readable storage medium, the computer storing computer program code, when the computer program code is executed by a computer, the computer program code can cause the computer to execute an encoding method or a decoding method Any of the steps.
  • this embodiment divides each frame into at least two parts, including a current frame part and a subsequent frame part, and the current frame part and the subsequent frame part.
  • the sub-distribution is performed in the encoding of at least two frames, such that only the current frame portion of the time domain envelope information of the encoding transmission belongs to the current frame, and the remaining portion is sent in the time domain envelope information of the subsequent frame, thereby If the decoding end cannot correctly decode the current frame, the time domain envelope information belonging to the current frame can also be extracted from the subsequent frame, and then the time domain envelope information belonging to the current frame can be used for decoding, because the current frame belongs to the current frame.
  • the decoding of the time domain envelope information allows the continuity of the signal to be maintained as much as possible to make the hearing feel comfortable.
  • FIG. 1 is a flowchart of Embodiment 1 of an encoding method according to an embodiment of the present invention
  • Embodiment 2 is a flowchart of Embodiment 2 of an encoding method according to an embodiment of the present invention
  • Embodiment 3 is a flowchart of Embodiment 1 of a decoding method according to an embodiment of the present invention.
  • Embodiment 4 is a flowchart of Embodiment 2 of a decoding method according to an embodiment of the present invention.
  • FIG. 5 is a flowchart of Embodiment 4 of a decoding method according to an embodiment of the present invention.
  • Embodiment 5 is a flowchart of Embodiment 5 of a decoding method according to an embodiment of the present invention.
  • FIG. 7 is a structural diagram of an embodiment of an encoding apparatus according to an embodiment of the present invention.
  • FIG. 8 is a structural diagram of Embodiment 1 of a decoding apparatus according to an embodiment of the present invention.
  • FIG. 9 is a structural diagram of Embodiment 2 of a decoding apparatus according to an embodiment of the present invention.
  • FIG. 10 is a structural diagram of Embodiment 1 of a codec system according to an embodiment of the present invention.
  • FIG. 11 is a structural diagram of Embodiment 2 of a codec system according to an embodiment of the present invention.
  • FIG. 1 is a flowchart of a first embodiment of an encoding method according to an embodiment of the present invention, including: Step 101: Obtain a quantized time domain envelope information of a current frame.
  • the time domain signal of the current frame needs to be acquired first, and then the time domain envelope information of the time domain signal is extracted, and the extracted time domain envelope information is quantized, thereby obtaining the quantized Time domain envelope information;
  • the time domain signal according to the embodiment of the present invention may be a time domain signal of a narrowband portion, a time domain signal of a broadband portion, a time domain signal of an ultra wideband portion, and the like;
  • time domain envelope information There may be multiple ways to extract the time domain envelope information.
  • the embodiment of the present invention does not limit the extraction time domain packet.
  • the specific manner of the network information; the time domain envelope information is a group of information, the number of the group information may be preset, the number of the group of information of each frame is the same; the number of the group information may also be according to different frames The characteristics of the input signal determine that the number of sets of information of different frames may be different at this time;
  • Step 102 Divide the quantized time domain envelope information into at least two parts, including a current frame part and a subsequent frame part;
  • the quantized time domain envelope information is a set of signals, and the embodiment of the present invention does not limit how to divide it into a current frame part and a subsequent frame part, and does not limit the number of subsequent frame parts divided; however, the embodiment of the present invention preferably
  • the quantized time domain envelope information is divided into two parts, that is, one current frame part and one subsequent frame part; when the quantized time domain envelope information is divided into two parts, the embodiment of the present invention preferably quantizes in a parity manner.
  • the subsequent time domain envelope information is divided into a current frame portion and a subsequent frame portion, for example, the current frame portion may be an odd portion or an even portion of the quantized time domain envelope information, and the remaining portion is a subsequent frame portion.
  • time domain envelope information When the time domain envelope information is divided into two parts, only one subsequent frame carries the subsequent frame part in the time domain envelope information of the current frame; when the time domain envelope information is divided into more than two parts, there may be at least The two frames carry the subsequent frame portion of the time domain envelope information of the current frame.
  • Step 103 Combine the current frame part and the subsequent frame part of the previous frame data to obtain time domain envelope information that the current frame needs to be coded and transmitted;
  • the combination may be performed in a cross-combination manner, or may be directly combined in a segment, and the embodiment of the present invention does not limit how to combine specifically;
  • the number of envelope information in the time domain is determined according to the characteristics of the input signal, in order to enable the decoding end to determine the number of time domain envelope information of each frame, it is possible to combine the time domain envelope information that needs to be encoded and transmitted in the current frame. And combining the number of time domain envelope information of the current frame and the number of time domain envelope information of the next frame in the time domain envelope information that the current frame needs to be encoded and transmitted;
  • the subsequent frame portion of the previous frame data of the first frame data may be set in advance, for example, may be set to all 0s or all 1s, etc.; , encoding and transmitting the current frame requires time domain envelope information for transmission.
  • this embodiment will divide each frame into at least two parts, including the current frame part and the subsequent frame part, and distribute the current frame part and the subsequent frame part in the coding of at least two frames, which is sent.
  • the time domain envelope information sent by the sample code only the current frame part belongs to the current frame, and the rest part is sent in the time domain envelope information of the subsequent frame, so that if the decoding end cannot correctly decode the current frame, it can also The time domain envelope information belonging to the current frame is extracted in the subsequent frame, and then the time domain envelope information belonging to the current frame is used for decoding. Since the decoding is performed using the time domain envelope information belonging to the current frame, it is possible to try to The continuity of the signal is maintained, making the hearing feel comfortable.
  • the embodiment of the present invention further provides the second embodiment of the encoding method.
  • Figure 2 depicts the flow of the second embodiment, including:
  • Step 201 Acquire a time domain signal to be encoded of a current frame.
  • Step 202 Extract time domain envelope information of the time domain signal.
  • Step 203 Quantify time domain envelope information, and obtain quantized time domain envelope information.
  • Step 204 Divide the quantized time domain envelope information into at least two parts, including a current frame part and a subsequent frame part;
  • Step 205 Combine the current frame part and the subsequent frame part of the previous frame data to obtain time domain envelope information that the current frame needs to be encoded and transmitted.
  • Step 206 Obtain a frequency domain signal of the current frame by using the quantized time domain envelope information.
  • the quantized time domain envelope information may be smoothly interpolated to obtain a smoothed time domain envelope.
  • the subsequent time domain envelope normalizes the time domain signal to obtain a normalized time domain signal; and uses the normalized time domain signal and the normalized time domain signal of the previous frame data to form Array; transforming and encoding the array to obtain a frequency domain signal of the current frame;
  • the embodiment of the present invention does not limit the smooth interpolation method, the normalization processing method, and the transform coding manner;
  • the preferred transform coding mode in the embodiment of the present invention is a 50% overlapping modified discrete cosine transform (MDCT: Modified Discrete Cosine Transform );
  • Step 207 Extract frequency domain envelope information of the frequency domain signal, and select a frequency domain language parameter.
  • the frequency domain envelope information of the frequency domain signal can be extracted.
  • the embodiment of the present invention does not limit how to extract the frequency domain envelope information of the frequency domain signal.
  • the selection can be made according to the human ear masking effect
  • Step 208 Quantize the frequency domain envelope information and the frequency domain language parameter, and obtain the quantized frequency domain envelope information and the quantized frequency domain spectrum parameter.
  • Step 209 Encode and transmit the time domain envelope information of the current frame, the quantized frequency domain envelope information, and the quantized frequency domain spectrum parameter.
  • this embodiment will divide each frame into at least two parts, including the current frame part and the subsequent frame part, and distribute the current frame part and the subsequent frame part in the coding of at least two frames, so that the coding is performed.
  • the time domain envelope information sent only the current frame part belongs to the current frame, and the rest part is sent in the time domain envelope information of the subsequent frame, so that if the decoding end cannot correctly decode the current frame, it can also be from the subsequent frame.
  • the time domain envelope information belonging to the current frame is extracted, and then the time domain envelope information belonging to the current frame is used for decoding. Because it is decoded by using the time domain envelope information belonging to the current frame, it can be kept as much as possible. The continuity of the signal makes the hearing feel comfortable.
  • the embodiment of the present invention further provides a third embodiment of the encoding method, and describes a process for encoding an ultra-wideband signal according to an embodiment of the present invention.
  • the embodiment describes that the time domain envelope information to be quantized is divided into a current frame portion and In the case of a subsequent frame part, the flow is as follows:
  • the current frame is the kth frame
  • the ultra-wideband input time domain signal of the kth frame is ⁇ y _ hi k (0), y _ hi k (1), ..., _ (319) ⁇
  • the extracted time domain envelope information is quantized, and the quantized time domain envelope information can be obtained.
  • the quantized time domain envelope information of the kth frame signal is divided into odd parts.
  • dividing the quantized time domain envelope information into an odd part and an even part is a preferred implementation manner provided by the embodiment of the present invention.
  • embodiments of the present invention are not limited to dividing the quantized time domain envelope information into odd and even portions. Among them, the even part is taken as the current frame part, and the odd part is taken as the subsequent frame part.
  • the time domain envelope information quantized by the frame signal is subjected to smooth interpolation processing to obtain a smoothed time domain envelope ⁇ . ⁇ 4 & 4 19 ⁇ .
  • the embodiment of the present invention does not limit the specific method of the smooth interpolation.
  • the smooth interpolation method provided by the embodiment of the present invention is as follows
  • the normalized time domain signal ⁇ y _hi k ⁇ 0), y_hi k '(1), ..., y — hi k '(319) ⁇ and the k-1th frame signal stored in the encoding status register 'J UI),... '(319) ⁇ constitutes a set of 640-point arrays:
  • the encoded k-th frame needs to encode the transmitted time domain envelope information, the frequency domain envelope information, and the frequency domain speech parameters into the final code stream for packet transmission.
  • the number of time domain envelope information is determined by the input signal characteristics
  • the two frames involved in the time domain envelope information that needs to be encoded and transmitted may be used.
  • the number of time domain envelope information is also combined in the time domain envelope information that the current frame needs to be encoded and transmitted;
  • the time domain envelope information of the code transmission is encoded.
  • the ..., ⁇ of the frame signal is stored in the code status register for time domain envelope coding of the next frame signal.
  • the combination of the time domain envelope information of the current frame and the coded transmission is only an implementation manner of the embodiment of the present invention, and is not limited to the combination of the foregoing manner; Differentiating different combinations of information does not affect the implementation of the embodiments of the present invention.
  • the foregoing describes only the case where the time domain envelope information is divided into a current frame part and a subsequent frame part. The process of dividing the time domain envelope information into one current frame part and multiple subsequent frame parts is similar, and details are not described herein again. .
  • FIG. 3 is a flowchart of a first embodiment of a decoding method according to an embodiment of the present invention, including: Step 301: Decoding to obtain time domain envelope information of a current frame coding transmission;
  • Step 302 Extract, from the time domain envelope information transmitted by the current frame encoding, a current frame portion in the time domain envelope information of the current frame.
  • the time domain envelope information transmitted by the current frame coding in the embodiment of the present invention is a current frame part in the time domain envelope information of the current frame, and a subsequent frame part in the time domain envelope information of the previous frame of the current frame; Therefore, the current frame portion in the time domain envelope information of the current frame may be extracted from the time domain envelope information transmitted by the current frame encoding;
  • the time domain envelope information transmitted by the current frame coding further includes the number of time domain envelope information of the current frame and the previous frame of the current frame.
  • the number of time domain envelope information; at this time, the time domain of the current frame may be extracted from the time domain envelope information transmitted by the current frame encoding according to the number of time domain envelope information of the current frame.
  • Step 303 Obtain a subsequent frame part of time domain envelope information of the current frame.
  • the subsequent frame portion of the time domain envelope information of the current frame may be obtained by decoding: obtaining time domain envelope information of the subsequent frame coding transmission of the current frame; extracting the current frame from the time domain envelope information transmitted by the subsequent frame coding
  • the subsequent frame portion of the time domain envelope information; when the time domain envelope information is only divided into two parts, the subsequent frame portion may be obtained from a subsequent frame; when the time domain envelope information is divided into more than two parts, The subsequent frame portion may be obtained from at least two subsequent frames;
  • the subsequent frame portion of the current frame is not transmitted, and the subsequent frame portion of the time domain envelope information of the current frame may be acquired by using a preset manner
  • the body may use the current frame portion of the time domain envelope information of the current frame to obtain the entire time domain envelope information of the current frame.
  • the position of the time domain envelope information of the subsequent frame portion may be filled with 0 or 1 or may be used.
  • the information of the current frame portion is interpolated to obtain the entire time domain envelope information of the current frame;
  • Step 304 Combine the current frame part and the subsequent frame part to obtain time domain envelope information of the current frame.
  • the time domain envelope information of the current frame received by this embodiment only the current frame part belongs to the current frame, and the remaining part is transmitted by other frames, so even if the current frame cannot be correctly decoded or lost, it can be obtained from other frames.
  • the remaining portion of the time domain envelope information of the current frame which in turn can be decoded using the remaining portion of the time domain envelope information belonging to the current frame, is decoded by using the time domain envelope information belonging to the current frame, so The continuity of the signal is maintained to make the hearing feel comfortable; at the same time, when the current frame is normally transmitted, the information of the current frame can be correctly decoded without affecting the continuity of the signal.
  • FIG. 4 is a flowchart of a second embodiment of a decoding method according to an embodiment of the present invention, including: Step 401: Decoding, obtaining time domain envelope information, frequency domain envelope information, and frequency domain spectrum parameter information of a current frame coding transmission;
  • Step 402 Extract, from the time domain envelope information transmitted by the current frame encoding, the current frame portion in the time domain envelope information of the current frame.
  • Step 403 Obtain a subsequent frame portion of the time domain envelope information of the current frame.
  • Step 404 Combine the current frame part and the subsequent frame part to obtain time domain envelope information of the current frame.
  • Step 405 Obtain a reconstructed frequency domain signal of the current frame by using frequency domain envelope information and frequency domain parameter information.
  • the step may specifically include: generating a random noise spectrum, performing energy adjustment on the random noise spectrum by using the frequency domain envelope information, and calculating spectral envelope information of the energy-adjusted random noise spectrum; and replacing the frequency envelope parameter information with the spectral envelope Information, obtaining a reconstructed frequency domain signal of the current frame;
  • the spectral envelope information of the random noise spectrum can be adjusted to the envelope information conforming to the original signal
  • the reconstructed frequency domain signal of the current frame After the reconstructed frequency domain signal of the current frame is obtained, the reconstructed frequency domain signal of the current frame may be further saved;
  • Step 406 Perform an inverse transform reconstruction on the reconstructed frequency domain signal and the intermediate signal obtained by performing inverse transform reconstruction on the previous frame of the saved current frame to obtain a reconstructed time domain signal. After the reconstructed time domain signal is obtained, the intermediate signal of the current frame obtained during the inverse transform reconstruction may be further saved;
  • the intermediate signal of the current frame obtained by the inverse transform can be saved, which is convenient for the next frame to be inverse transformed and reconstructed;
  • the inverse transform reconstruction method corresponds to the transform coding method used in coding. For example, 50% discrete cosine transform transform is used in coding, and corresponding inverse discrete cosine is used in inverse transform reconstruction. Transform transformation and 50% overlap addition;
  • Step 407 Perform a smooth interpolation operation on the time domain envelope information of the current frame to obtain smoothed time domain envelope information.
  • Step 408 Perform a denormalization operation to obtain the reconstructed current frame information by using the reconstructed time domain signal and the smoothed time domain envelope information.
  • the remaining part of the time domain envelope information of the frame received in this embodiment belongs to the current frame, and the remaining part is transmitted by other frames. Therefore, even if the current frame cannot be correctly decoded or lost, the current frame can be obtained from other frames.
  • the remaining part of the time domain envelope information of the frame can be decoded by using the remaining part of the time domain envelope information belonging to the current frame. Since it is decoded using the time domain envelope information belonging to the current frame, it can be as much as possible. Keeping the continuity of the signal makes the hearing feel comfortable. At the same time, when the current frame is normally transmitted, the information of the current frame can be correctly decoded without affecting the continuity of the signal.
  • Embodiment 3 of the decoding method further provides Embodiment 3 of the decoding method, and describes a case where the time domain envelope information is divided into two parts, and the flow is as follows:
  • the time domain envelope information of the current kth frame is decoded from the received code stream ⁇ 4 4 4 ⁇ 4 ⁇ q k - ⁇ frequency domain envelope information and frequency domain spectral parameter information, because the discrete cosine is performed at the decoding end Transforming the inverse transform will have a delay of one frame, so when the decoding end decodes the time domain envelope information of the kth frame, the time domain envelope information of the k+1th frame can be decoded to obtain the time domain envelope information IA q A q A q A q A q A q A q A q A q A q A q A q A q A q A q q ⁇ ⁇
  • a set of random noise spectra is generated, and the noise spectrum is energy-adjusted according to the decoded frequency domain envelope information, so that the spectral envelope of the noise spectrum conforms to the envelope information of the original signal, and the spectral envelope information of the energy-adjusted noise spectrum is calculated. .
  • the decoded frequency domain language parameter is substituted for the spectral envelope information of the noise spectrum of the corresponding position to obtain the current
  • the reconstructed frequency domain signal of the frame ⁇ _ 3 ⁇ 4(0), _ 3 ⁇ 4(1), ..., y_swb k (3 ⁇ 9) ⁇ ; further, the set of signals can be retained in the decoding status register for needs Used when the frame is lost or cannot be decoded.
  • the reconstructed frequency i or the signal ⁇ _ 3 ⁇ 4 (0), y _ swb k (1), ..., _ 3 ⁇ 4 (319) ⁇ and the previous frame signal are inverse transformed and reconstructed in the decoding status register.
  • Group signal > , imdct k _, (319) ⁇ , perform inverse transform reconstruction, get reconstructed time domain signal ⁇ y_hi k X0ly_hi k X ⁇ l...
  • the transform coding method used in coding corresponds to, for example, a discrete cosine transform transform is used for encoding, and an inverse transform cosine transform transform can be used for inverse transform reconstruction during decoding. Further, a set of intermediate signals ⁇ imdct k (0), imdct k (1), ..., imdct k (319) ⁇ generated by the inverse of the inverse discrete cosine transform may be retained in the decoding status register for The next frame signal is used for inverse transform reconstruction.
  • FIG. 5 is a flowchart of a fourth embodiment of a decoding method according to an embodiment of the present invention, including: Step 501: Decode time domain envelope information of a subsequent frame encoding transmission of a current frame; Step 502: Time domain of transmission from a subsequent frame encoding Extracting, in the envelope information, a subsequent frame portion in the time domain envelope information of the current frame;
  • the subsequent frame part in the time domain envelope information of the current frame may be extracted from the time domain envelope information transmitted by the subsequent frame coding according to the fixed value;
  • the number of time domain envelope information of the current frame may be extracted from the time domain envelope information transmitted by the subsequent frame encoding; and then according to the time domain envelope of the current frame. The amount of information, using the subsequent frame portion to obtain the time domain envelope information of the current frame;
  • Step 503 Obtain time domain envelope information of the current frame by using a subsequent frame portion.
  • the frame part recovers the time domain envelope information of the current frame. Since the subsequent frame part is only part of the time domain envelope information, in order to enable this part of the information to restore the entire time domain envelope information as much as possible, the subsequent frame part may be performed internally. Inserting processing to obtain time domain envelope information of the current frame;
  • the time domain envelope information of the current frame can be restored by using the subsequent frame portion of the time domain envelope information of the current frame, so that the time domain packet belonging to the current frame can be used.
  • the remaining part of the network information is decoded, and since it is decoded using the time domain envelope information belonging to the current frame, the continuity of the signal can be maintained as much as possible, and the hearing feeling is comfortable.
  • FIG. 6 is a flowchart of a fifth embodiment of a decoding method according to an embodiment of the present invention, including: Step 601: Decode time domain envelope information of a subsequent frame encoding transmission of a current frame; Step 602: Time domain of transmission from a subsequent frame encoding Extracting, in the envelope information, a subsequent frame portion in the time domain envelope information of the current frame;
  • Step 603 Obtain time domain envelope information of the current frame by using a subsequent frame part.
  • Step 604 Perform inverse transform reconstruction on the reconstructed frequency domain signal of the previous frame of the saved current frame and the intermediate signal of the previous frame obtained by reconstructing the inverse transform, to obtain a reconstructed time domain signal of the current frame;
  • obtaining the reconstructed time domain signal can further preserve the intermediate signal of the current frame generated when the inverse transform is reconstructed;
  • the intermediate signal of the current frame generated when performing the inverse transform reconstruction may be used when performing inverse transform reconstruction in the next frame;
  • the inverse transform reconstruction method corresponds to the transform coding method used in coding. For example, 50% discrete cosine transform transform is used in coding, and corresponding inverse discrete cosine is used in inverse transform reconstruction. Transform transformation and 50% overlap addition;
  • Step 605 Perform a smooth interpolation operation on the time domain envelope information of the current frame, and obtain the smoothed time domain envelope information of the current frame.
  • Step 606 Perform anti-normalization operation on the reconstructed time domain signal by using the smoothed time domain envelope information to obtain the reconstructed current frame information.
  • the time domain envelope information of the current frame can be restored by using the subsequent frame portion of the time domain envelope information of the current frame, so that the time domain packet belonging to the current frame can be used.
  • the remaining part of the network information is decoded, and since it is decoded using the time domain envelope information belonging to the current frame, the continuity of the signal can be maintained as much as possible, and the hearing feeling is comfortable.
  • Embodiment 6 of the decoding method describes the case when the time domain envelope information is divided into two parts, and the flow is as follows:
  • the frequency domain envelope and frequency domain parameters of the kth frame cannot be obtained, and the time domain envelope information of the k+1th frame can be decoded from the code stream ⁇ +1 , +1 , + 1 , 4 , , 4 , , where there are subsequent frame portions of the time domain envelope information belonging to the kth frame.
  • imdct k _, (319) ⁇ Perform the inverse transform reconstruction together to obtain the reconstructed time domain signal ⁇ y_hi k ⁇ 0 ⁇ y_hi k '(1), whil, y-hi k '(319) ⁇ , And the intermediate signal ⁇ imdct k (0), imdct k (1), ..., imdct k (319) ⁇ generated when the k- th frame signal is reconstructed in the inverse transform is reserved in the decoding status register for the next frame signal. Used when performing inverse transform reconstruction.
  • FIG. 7 illustrates a structure of an encoding device according to an embodiment of the present invention, which includes:
  • the time domain envelope information obtaining unit 701 is configured to obtain the quantized time domain envelope information of the current frame;
  • the splitting unit 702 is configured to divide the quantized time domain envelope information acquired by the time domain envelope information acquiring unit 701 into at least two parts, including a current frame part and a subsequent frame part;
  • the combining unit 703 is configured to combine the current frame portion and the subsequent frame portion of the previous frame data to obtain time domain envelope information that the current frame needs to be encoded and transmitted;
  • the code sending unit 704 is configured to encode and transmit the time domain envelope information of the current frame that is combined by the combining unit 703.
  • this embodiment will divide each frame into at least two parts, including the current frame part and the subsequent frame part, and distribute the current frame part and the subsequent frame part in the coding of at least two frames, so that the coding is performed.
  • the time domain envelope information sent only the current frame part belongs to the current frame, and the rest part is sent in the time domain envelope information of the subsequent frame, so that if the decoding end cannot correctly decode the current frame, it can also be from the subsequent frame.
  • the time domain envelope information belonging to the current frame is extracted, and then the time domain envelope information belonging to the current frame is used for decoding. Because it is decoded by using the time domain envelope information belonging to the current frame, it can be kept as much as possible. The continuity of the signal makes the hearing feel comfortable.
  • the encoding apparatus may further include:
  • the frequency domain signal acquiring unit is configured to obtain the frequency domain signal of the current frame by using the quantized time domain envelope information after the time domain envelope information acquiring unit 701 obtains the quantized time domain envelope information;
  • a frequency domain envelope information extracting unit configured to extract frequency domain envelope information of a frequency domain signal, and select a frequency domain spectrum parameter
  • a quantization unit configured to quantize the frequency domain envelope information extracted by the frequency domain envelope information extraction unit and the selected frequency domain language parameter, to obtain the quantized frequency domain envelope information and the quantized frequency domain language parameter;
  • the code transmitting unit 704 is further configured to encode and transmit the quantized frequency domain envelope information and the quantized frequency domain language parameter.
  • the frequency domain signal acquiring unit may include: a smoothing unit, configured to: after the quantized unit obtains the quantized time domain envelope information, perform smooth interpolation processing on the quantized time domain envelope information to obtain a smoothed a time domain envelope; a normalization unit for normalizing the time domain signal using the smoothed time domain envelope to obtain a normalized time domain signal; a transform coding unit for using normalization The subsequent time domain signal forms an array with the normalized time domain signal of the previous frame data; the array is transformed and encoded to obtain the frequency domain signal of the current frame.
  • the embodiment of the present invention further provides a decoding device, and FIG. 8 depicts the decoding device provided by the embodiment of the present invention.
  • the structure of the first embodiment includes:
  • a decoding unit 801 configured to decode time domain envelope information of the current frame coding transmission
  • the extracting unit 802 is configured to extract, from the time domain envelope information transmitted by the current frame encoding, the current frame portion in the time domain envelope information of the current frame;
  • the obtaining unit 803 is configured to obtain a subsequent frame portion in the time domain envelope information of the current frame
  • the combining unit 804 is configured to combine the current frame portion and the subsequent frame portion to obtain time domain envelope information of the current frame.
  • the time domain envelope information of the current frame received by this embodiment only the current frame part belongs to the current frame, and the remaining part is transmitted by other frames, so even if the current frame cannot be correctly decoded or lost, it can be obtained from other frames.
  • the remaining portion of the time domain envelope information of the current frame which in turn can be decoded using the remaining portion of the time domain envelope information belonging to the current frame, is decoded by using the time domain envelope information belonging to the current frame, so The continuity of the signal is maintained to make the hearing feel comfortable; at the same time, when the current frame is normally transmitted, the information of the current frame can be correctly decoded without affecting the continuity of the signal.
  • the decoding unit 801 is further configured to: decode the frequency domain envelope information and the frequency domain language parameter information of the current frame.
  • the decoding apparatus provided by the embodiment of the present invention may further include:
  • a reconstruction unit configured to obtain a reconstructed frequency domain signal of the current frame by using frequency domain envelope information and frequency domain parameter information;
  • An inverse transform reconstruction unit is configured to perform inverse transform reconstruction on the reconstructed frequency domain signal and the intermediate signal of the previous frame of the saved current frame to obtain a reconstructed time domain signal;
  • a smoothing interpolation unit configured to perform a smooth interpolation operation on the time domain envelope information of the current frame to obtain smoothed time domain envelope information
  • the inverse normalization unit is configured to perform the inverse normalization operation to obtain the reconstructed current frame information by using the reconstructed time domain signal and the smoothed time domain envelope information.
  • the first embodiment of the decoding apparatus may further include a saving unit, configured to save the reconstructed frequency domain signal of the current frame after the reconstructing unit obtains the reconstructed frequency domain signal of the current frame; After the inverse transform reconstruction unit obtains the reconstructed time domain signal, the intermediate signal of the current frame obtained when the inverse transform is reconstructed is saved;
  • FIG. 9 is a schematic diagram of a structure of a second embodiment of a decoding apparatus according to an embodiment of the present invention, including:
  • the decoding unit 901 is configured to decode the time domain envelope information of the subsequent frame encoding transmission of the current frame.
  • the extracting unit 902 is configured to extract the time domain envelope information of the current frame from the time domain envelope information transmitted by the subsequent frame encoding. Subsequent frame portion in ;
  • the time domain envelope information obtaining unit 903 is configured to obtain time domain envelope information of the current frame by using the subsequent frame portion.
  • the time domain envelope information of the current frame can be restored by using the subsequent frame portion of the time domain envelope information of the current frame, so that the time domain packet belonging to the current frame can be used.
  • the remaining part of the network information is decoded, and since it is decoded using the time domain envelope information belonging to the current frame, the continuity of the signal can be maintained as much as possible, and the hearing feeling is comfortable.
  • the second embodiment of the decoding apparatus provided by the embodiment of the present invention may further include:
  • the reconstructed time domain signal acquiring unit is configured to inversely transform the reconstructed frequency domain signal of the previous frame of the saved current frame and the intermediate signal of the previous frame obtained by reconstructing the inverse transform to obtain a current frame. Reconstructing the time domain signal;
  • a smoothing unit configured to perform a smooth interpolation operation on the time domain envelope information of the current frame, to obtain smoothed time domain envelope information of the current frame
  • the inverse normalization unit is configured to perform an inverse normalization operation on the reconstructed time domain signal by using the smoothed time domain envelope information to obtain the reconstructed current frame information.
  • the second embodiment of the decoding apparatus may further include a saving unit, configured to save, after the reconstructed time domain signal acquiring unit obtains the reconstructed time domain signal of the current frame, save the current generated when the inverse transform is reconstructed.
  • the intermediate signal of the frame may further include a saving unit, configured to save, after the reconstructed time domain signal acquiring unit obtains the reconstructed time domain signal of the current frame, save the current generated when the inverse transform is reconstructed.
  • the intermediate signal of the frame may further include a saving unit, configured to save, after the reconstructed time domain signal acquiring unit obtains the reconstructed time domain signal of the current frame, save the current generated when the inverse transform is reconstructed.
  • FIG. 10 depicts a structure of the first embodiment of the codec system, including:
  • the encoding device 1001 is configured to obtain the quantized time domain envelope information of the current frame, and divide the quantized time domain envelope information into at least two parts, including at least one current frame part and at least one subsequent frame part; combining the current frame part And the subsequent frame part of the previous frame data of the current frame, the time domain envelope information that needs to be encoded and transmitted by the current frame is obtained, and the number of previous frame data is the same as the number of subsequent frames; the encoded current frame needs the time domain envelope of the coded transmission. Information, the time domain envelope information of the encoded transmission is required to transmit the encoded current frame;
  • the decoding device 1002 is configured to decode and obtain time domain envelope information of the current frame coding transmission; In the time domain envelope information transmitted by the frame encoding, extracting a current frame portion in the time domain envelope information of the current frame; acquiring a subsequent frame portion in the time domain envelope information of the current frame; combining the current frame portion and the subsequent frame portion to obtain Time domain envelope information for the current frame.
  • the time domain envelope information of the current frame sent by the encoding apparatus of this embodiment only the current frame part belongs to the current frame, and the remaining part is transmitted by other frames, so even if the current frame cannot be correctly decoded or lost, it can be from other frames.
  • the information of the current frame can be correctly decoded without affecting the continuity of the signal.
  • Figure 11 depicts the structure of the second embodiment of the codec system, including:
  • the encoding device 1101 is configured to obtain the quantized time domain envelope information of the current frame, and divide the quantized time domain envelope information into at least two parts, including at least one current frame part and at least one subsequent frame part; combining the current frame part And the subsequent frame part of the previous frame data of the current frame, the time domain envelope information that needs to be encoded and transmitted by the current frame is obtained, and the number of previous frame data is the same as the number of subsequent frames; the encoded current frame needs the time domain envelope of the coded transmission. Information, the time domain envelope information of the encoded transmission is required to transmit the encoded current frame;
  • the decoding device 1102 is configured to decode time domain envelope information of a subsequent frame coding transmission of the current frame; and extract, from the time domain envelope information transmitted by the subsequent frame encoding, a subsequent frame portion in the time domain envelope information of the current frame; The time domain envelope information of the current frame is obtained using the subsequent frame portion.
  • the decoding device when the current frame fails to be correctly transmitted, the decoding device can restore the time domain envelope information of the current frame by using the subsequent frame portion of the time domain envelope information of the current frame, so that the current frame can be used.
  • the remaining part of the time domain envelope information is decoded, and since it is decoded using the time domain envelope information belonging to the current frame, the continuity of the signal can be maintained as much as possible, and the hearing feeling is comfortable.
  • Encoding and transmitting the current frame requires encoding the transmitted time domain envelope information.
  • Combining the current frame portion and the subsequent frame portion obtains time domain envelope information of the current frame. And / or include the following steps:
  • the time domain envelope information of the current frame is obtained using the subsequent frame portion.
  • the above-mentioned storage medium may be a read only memory, a magnetic disk or an optical disk or the like.

Landscapes

  • Engineering & Computer Science (AREA)
  • Computational Linguistics (AREA)
  • Signal Processing (AREA)
  • Health & Medical Sciences (AREA)
  • Audiology, Speech & Language Pathology (AREA)
  • Human Computer Interaction (AREA)
  • Physics & Mathematics (AREA)
  • Acoustics & Sound (AREA)
  • Multimedia (AREA)
  • Compression, Expansion, Code Conversion, And Decoders (AREA)

Description

编码方法及装置、 解码方法及装置
本申请要求于 2008年 4月 1日提交中国专利局、申请号为 200810089148.1、 发明名称为 "编码方法及装置、 解码方法及装置" 的中国专利申请的优先权, 其全部内容通过引用结合在本申请中。
技术领域
本发明涉及数据编码技术, 具体涉及编码方法及装置、 解码方法及装置。 背景技术
在通信系统中, 两个会话终端之间会建立起一个回路, 通过这一回路发送 端将包含语音信息的数据包传送到接收端,再在接收端合成语音信号。如果在 传送过程中出现了数据包的丟失,需要通过以前接收到的数据包或者其他方法 来恢复出丟失数据包所包含的语音信息, 否则会带来听觉的不适感。 随着通信 技术的发展, 现在很多语音通信很多是通过网际协议(IP: Internet Protocol ) 网络、 无线网络进行连接、 传输, 在这种通信环境下, 出现数据包丟失的概率 更大。
目前在窄带语音编码器、 宽带语音编码器中釆用了各种前向纠错(FEC:
Forward Error Correction )策略, 较好的解决了帧丟失时语音质量下降的问题。 但是对于超宽带部分的信号,由于信号在这部分带宽内的特征、参数等和窄带、 宽带部分的有较大的区别, 用于窄带、 宽带部分的前向纠错策略用于超宽带部 分效果不一定好,目前对于超宽带部分的信号则还没有一种较好的前向糾错策 略。
现有的一种对于超宽带信号的处理方法是这样的:如果出现帧丟失或者受 损的情况, 重复前一帧已解码的参数, 即用前一帧已解码的参数作为当前帧的 参数, 再进行后续解码处理。
虽然语音信号具有短时平稳特性, 相邻两帧的信号存在着一定的相关性, 但是上述方法直接利用前一帧的参数作为当前帧的参数进行解码,即不包含任 何当前帧应有的信息, 则最终解码得到的信号仍会有较明显的不连续感, 听觉 感受不舒服。
发明内容
本发明实施例提供了编码方法及装置、解码方法及装置,使用本发明实施 例提供的技术方案, 可以在当前帧丟失或受损时,使用当前帧的部分信息进行 解码。
本发明实施例提供了一种编码方法, 包括:
获取当前帧的量化后的时域包络信息;
将所述量化后的时域包络信息分成至少两部分,包括至少一个当前帧部分 和至少一个后续帧部分;
组合所述当前帧部分和所述当前帧的之前帧数据的后续帧部分,得到当前 帧需要编码传送的时域包络信息;
编码所述当前帧需要编码传送的时域包络信息;
发送编码后的所述当前帧需要编码传送的时域包络信息。
本发明实施例提供了一种解码方法, 包括:
解码得到当前收到的帧编码传送的时域包络信息,所述当前收到的帧编码 传送的时域包络信息包括所述当前收到的帧的时域包络信息中的当前帧部分 以及在收到当前收到的帧之前收到的帧的时域包络信息中的后续帧部分; 从所述当前收到的帧编码传送的时域包络信息中 ,提取所述当前收到的帧 的时域包络信息中的当前帧部分;
获取所述当前收到的帧的时域包络信息中的后续帧部分;
组合所述当前收到的帧的时域包络信息中的当前帧部分和所述当前收到 的帧的时域包络信息中的后续帧部分, 得到所述当前收到的帧的时域包络信 息。
本发明实施例提供了另一种解码方法, 包括:
解码得到在当前收到的帧之后收到的帧编码传送的时域包络信息,所述编 码传送的时域包络信息包括所述之后收到的帧的时域包络信息中的当前帧部 分和所述当前收到帧的时域包络信息中的后续帧部分;
从所述之后收到的帧编码传送的时域包络信息中,提取所述当前收到的帧 的时域包络信息中的后续帧部分;
使用所述后续帧部分获得所述当前收到的帧的时域包络信息。
本发明实施例提供了一种编码装置, 包括:
时域包络信息获取单元, 用于获取当前帧的量化后的时域包络信息; 拆分单元,用于将所述时域包络信息获取单元获取的量化后的时域包络信 息分成至少两部分, 包括至少一个当前帧部分和至少一个后续帧部分;
组合单元,用于组合所述当前帧部分和所述当前帧的之前帧数据的后续帧 部分,得到当前帧需要编码传送的时域包络信息, 所述之前帧数据的数量与所 述后续帧的数量相同;
编码发送单元,用于编码所述组合单元组合得到的当前帧需要编码传送的 时域包络信息, 发送编码后的当前帧需要编码传送的时域包络信息。
本发明实施例提供了一种解码装置, 包括:
解码单元, 用于解码得到当前收到的帧编码传送的时域包络信息, 所述当 前收到的帧编码传送的时域包络信息包括所述当前收到的帧的时域包络信息 中的当前帧部分以在收到当前收到的帧之前收到的帧的时域包络信息中的后 续帧部分;
提取单元, 用于从所述当前收到的帧编码传送的时域包络信息中,提取所 述当前收到的帧的时域包络信息中的当前帧部分;
获取单元, 用于获取所述当前收到的帧的时域包络信息中的后续帧部分; 组合单元,用于组合所述当前收到的帧的时域包络信息中的当前帧部分和 所述当前收到的帧的时域包络信息中的后续帧部分,得到所述当前收到的帧的 时域包络信息。
本发明实施例提供了另一种解码装置, 包括:
解码单元,用于解码得到在当前收到的帧之后收到的帧编码传送的时域包 络信息 ,所述编码传送的时域包络信息包括所述之后收到的帧的时域包络信息 中的当前帧部分和所述当前收到帧的时域包络信息中的后续帧部分;
提取单元, 用于从所述之后收到的帧编码传送的时域包络信息中,提取所 述当前收到的帧的时域包络信息中的后续帧部分;
时域包络信息获取单元,用于使用所述后续帧部分获得所述当前收到的帧 的时域包络信息。
本发明实施例提供了一种编解码系统, 包括:
编码装置, 用于获取当前帧的量化后的时域包络信息; 将所述量化后的时 域包络信息分成至少两部分, 包括至少一个当前帧部分和至少一个后续帧部 分; 组合所述当前帧部分和所述当前帧的之前帧数据的后续帧部分,得到当前 帧需要编码传送的时域包络信息,所述之前帧数据的数量与所述后续帧的数量 相同; 编码得到的当前帧需要编码传送的时域包络信息,发送编码后的当前帧 需要编码传送的时域包络信息;
解码装置, 用于解码得到所述当前收到的帧编码传送的时域包络信息; 从 所述当前收到的帧编码传送的时域包络信息中,提取所述当前收到的帧的时域 包络信息中的当前帧部分; 获取所述当前帧的时域包络信息中的后续帧部分; 组合所述当前收到的帧的当前帧部分和所述当前收到的帧的后续帧部分得到 所述当前收到的帧的时域包络信息。
本发明实施例提供了另一种编解码系统, 包括:
编码装置, 用于获取当前帧的量化后的时域包络信息; 将所述量化后的时 域包络信息分成至少两部分, 包括至少一个当前帧部分和至少一个后续帧部 分; 组合所述当前帧部分和所述当前帧的之前帧数据的后续帧部分,得到当前 帧需要编码传送的时域包络信息,所述之前帧数据的数量与所述后续帧的数量 相同; 编码得到的当前帧需要编码传送的时域包络信息,发送编码后的当前帧 需要编码传送的时域包络信息;
解码装置,用于解码得到在当前收到的帧之后收到的帧编码传送的时域包 络信息; 从所述之后收到的帧编码传送的时域包络信息中,提取所述当前收到 的帧的时域包络信息中的后续帧部分;使用所述当前收到的帧的时域包络信息 中的后续帧部分获得所述当前收到的帧的时域包络信息。
本发明还提供一种计算机程序产品,所述计算机程序产品包括计算机程序 代码, 当所述计算机程序代码被一个计算机执行的时候, 所述计算机程序代码 可以使得所述计算机执行编码方法或解码方法中的任意一项步骤。
本发明还提供一种计算机可读存储介质, 所述计算机存储计算机程序代 码, 当所述计算机程序代码被一个计算机执行的时候, 所述计算机程序代码可 以使得所述计算机执行编码方法或解码方法中的任意一项步骤。
从本发明实施例提供的以上技术方案可以看出,由于本实施例将每个帧都 分成至少两部分, 包括当前帧部分和后续帧部分, 并将当前帧部分和后续帧部 分放在至少两个帧的编码中分发送,这样编码发送的时域包络信息中只有当前 帧部分是属于当前帧的, 而其余部分放在后续帧的时域包络信息中发送,从而 如果解码端不能对当前帧正确解码,也能从后续帧中提取到属于当前帧的时域 包络信息, 进而可以使用属于当前帧的时域包络信息进行解码, 由于是使用的 属于当前帧的时域包络信息进行的解码, 因此可以尽量的保持信号的连续性, 使听觉感受舒服。
附图说明
图 1为本发明实施例中编码方法实施例一的流程图;
图 2为本发明实施例中编码方法实施例二的流程图;
图 3为本发明实施例中解码方法实施例一的流程图;
图 4为本发明实施例中解码方法实施例二的流程图;
图 5为本发明实施例中解码方法实施例四的流程图;
图 6为本发明实施例中解码方法实施例五的流程图;
图 7为本发明实施例中编码装置实施例的结构图;
图 8为本发明实施例中解码装置实施例一的结构图;
图 9为本发明实施例中解码装置实施例二的结构图;
图 10为本发明实施例中编解码系统实施例一的结构图;
图 11为本发明实施例中编解码系统实施例二的结构图。
具体实施方式
为使本发明的目的、 技术方案、 及优点更加清楚明白, 以下参照附图并举 实施例, 对本发明进一步详细说明。
图 1描述了本发明实施例提供的编码方法实施例一的流程, 包括: 步骤 101、 获取当前帧的量化后的时域包络信息;
在获取量化后的时域包络信息过程中需要先获取当前帧的时域信号,然后 提取时域信号的时域包络信息,对提取的时域包络信息进行量化,从而得到量 化后的时域包络信息;
本发明实施例所述的时域信号可以是窄带部分的时域信号、宽带部分的时 域信号、 超宽带部分的时域信号等;
时域包络信息的提取方式可以有多种,本发明实施例并不限定提取时域包 络信息的具体方式; 时域包络信息是一组信息, 该组信息的数量可以是预先设 定的,每一帧的该组信息的数量都相同; 该组信息的数量也可以根据不同帧的 输入信号特征决定, 此时不同帧的该组信息的数量可能不同;
量化获得量化后的时域包络信息的量化方式也可以有多种,本发明实施例 并不限定;
步骤 102、 将量化后的时域包络信息分成至少两部分, 包括当前帧部分和 后续帧部分;
量化后的时域包络信息是一组信号,本发明实施例并不限定如何将其分成 当前帧部分和后续帧部分, 并且也不限定分成的后续帧部分的数量; 但是本发 明实施例优选将量化后的时域包络信息分成两部分,即一个当前帧部分和一个 后续帧部分; 在将量化后的时域包络信息分成两部分时, 本发明实施例优选按 照奇偶的方式将量化后的时域包络信息分成当前帧部分和后续帧部分,例如当 前帧部分可以是量化后的时域包络信息的奇数部分或偶数部分,剩余部分作为 后续帧部分。在将时域包络信息分成两部分时, 只有一个后续帧会携带当前帧 的时域包络信息中的后续帧部分; 在将时域包络信息分成多于两部分时, 则可 以有至少两个帧携带当前帧的时域包络信息中的后续帧部分。
步骤 103、 组合当前帧部分和之前帧数据的后续帧部分, 得到当前帧需要 编码传送的时域包络信息;
在将当前帧的当前帧部分和之前帧的后续帧部分组合时, 可以交叉组合, 也可以分段直接组合, 本发明实施例并不限定具体如何组合;
在时域包络信息的数量是根据输入信号特征决定时,为了使解码端能够确 定每个帧的时域包络信息的数量,可以在组合得到当前帧需要编码传送的时域 包络信息时,将当前帧的时域包络信息的数量以及下一帧的时域包络信息的数 量, 也组合在当前帧需要编码传送的时域包络信息中;
在当前帧是第 1帧时, 由于没有之前帧数据的后续帧部分, 因此可以预先 设置第 1帧数据的之前帧数据的后续帧部分, 例如可以设定为全 0或全 1等; 步骤 104、 编码并发送当前帧需要编码传送的时域包络信息。
从上可知, 本实施例将当每个帧都分成至少两部分, 包括当前帧部分和后 续帧部分, 并将当前帧部分和后续帧部分放在至少两个帧的编码中分发送, 这 样编码发送的时域包络信息中只有当前帧部分是属于当前帧的,而其余部分放 在后续帧的时域包络信息中发送,从而如果解码端不能对当前帧正确解码,也 能从后续帧中提取到属于当前帧的时域包络信息,进而可以使用属于当前帧的 时域包络信息进行解码, 由于是使用的属于当前帧的时域包络信息进行的解 码, 因此可以尽量的保持信号的连续性, 使听觉感受舒服。
进一步, 为了使传输的信号更准确,在得到量化后的时域包络信息后可以 进一步获取并发送频域信号和频域语参数,因此本发明实施例还提供了编码方 法的实施例二, 图 2描述了该实施例二的流程, 包括:
步骤 201、 获取当前帧的待编码的时域信号;
步骤 202、 提取时域信号的时域包络信息;
步骤 203、 量化时域包络信息, 得到量化后的时域包络信息;
步骤 204、 将量化后的时域包络信息分成至少两部分, 包括当前帧部分和 后续帧部分;
步骤 205、 组合当前帧部分和之前帧数据的后续帧部分, 得到当前帧需要 编码传送的时域包络信息;
步骤 206、 使用量化后的时域包络信息获得当前帧的频域信号; 具体的, 可以对量化后的时域包络信息进行平滑内插处理,得到平滑后的 时域包络; 使用平滑后的时域包络对时域信号进行归一化处理,得到归一化后 的时域信号;使用归一化后的时域信号与上一帧数据的归一化后的时域信号构 成数组; 对数组进行变换编码得到所述当前帧的频域信号;
其中, 本发明实施例对平滑内插方法、 归一化处理的方法以及变换编码的 方式都不作限定;本发明实施例优选的变换编码方式是 50%重叠的修正的离散 余弦变换(MDCT: Modified Discrete Cosine Transform );
步骤 207、 提取频域信号的频域包络信息, 选取频域语参数;
得到频域信号后, 可以提取频域信号的频域包络信息, 本发明实施例并不 限定如何提取频域信号的频域包络信息;
在选取频域语参数时, 具体可以根据人耳掩蔽效应进行选择;
步骤 208、 量化频域包络信息和所述频域语参数, 得到量化后的频域包络 信息和量化后的频域谱参数; 步骤 209、 编码并发送当前帧需要编码传送的时域包络信息、 量化后的频 域包络信息和量化后的频域谱参数。
从上可知, 本实施例将当每个帧都分成至少两部分, 包括当前帧部分和后 续帧部分, 并将当前帧部分和后续帧部分放在至少两个帧的编码中分发送, 这 样编码发送的时域包络信息中只有当前帧部分是属于当前帧的,而其余部分放 在后续帧的时域包络信息中发送,从而如果解码端不能对当前帧正确解码,也 能从后续帧中提取到属于当前帧的时域包络信息,进而可以使用属于当前帧的 时域包络信息进行解码, 由于是使用的属于当前帧的时域包络信息进行的解 码, 因此可以尽量的保持信号的连续性, 使听觉感受舒服。
本发明实施例还提供了编码方法的实施例三,描述了本发明实施例对超宽 带信号进行编码的过程,本实施例描述的是将待量化的时域包络信息分成一个 当前帧部分和一个后续帧部分的情况, 流程如下:
假设当前帧是第 k 帧, 且第 k 帧的超宽带输入时域信号为 {y _ hik (0), y _ hik (1),…… , _ (319) } , 提取该超宽带信号在时域的一组包络信息 {Aok,Alk,…… vJ; 提取的时域包络的个数可以是固定的, 每一帧都相同; 也 可以根据当前的输入信号特征确定。本发明实施例并不限定时域包络的计算方 式, 假设在本实施例中 N=9, 对应于时域包络的时间分辨率是 2ms, 则时域包 络信息可以釆用
Figure imgf000010_0001
等方式进行计算。
对提取的这组时域包络信息进行量化, 可以得到量化后的时域包络信息 将第 k 帧信号的量化后的 时域包络信息分为 奇数部分
{K d A 和偶数部分 { , , ,^, } , 需要说明的是, 将 量化后的时域包络信息分成奇数部分和偶数部分是本发明实施例提供的一种 较优的实现方式,但是本发明实施例并不限定只能将量化后的时域包络信息分 成奇数部分和偶数部分。 其中, 将偶数部分作为当前帧部分, 将奇数部分作为 后续帧部分。 将第 k
Figure imgf000010_0002
和上一帧信号保 存在编码状态器中的奇数部分 {4^,4^,4—,Λ^, 4 }重新组合为第 k 帧需 要编码发送的时域包络信息 { , , , , , ΑΚ-^^,Λ^,^ ^将本 帧信号的奇数部分 保存在编码状态寄存器中,用于第 k+1 帧信号的时域包络编码。
对本帧信号量化后的时域包络信息 …… 进行平滑内插处理,得 到平滑后的时域包络 μ。^4 …… 419^。 其中, 本发明实施例并不限定平滑内 插的具体方法, 本发明 实施例提供的一种平滑内插方法如下
= 。
Figure imgf000011_0001
N + l
再利用平滑之后的时域包络信息对时域信号进行归一化处理,可以釆用如 下公式进行归一化处理: y_hik(i) = y-hik(i); 从而得到归一化后的时域信号
_ 2Α'*
{y _hik\0 y _hik\\ ······ ,W (319)}。
将归一化后的时域信号 {y _hik \0),y_hik '(1),…… ,y— hik '(319)}和保存在编码状 态寄存器中的第 k-1帧信号 ' J UI),…… '(319)}构成一组 640 点的数组:
{y _ hik_, \0\y_ ik_x '(1),…… ,y_ hik_x 3\9\y_hik\0\y_hik\\\…… ,y_ hik '(319)}, 然后对该数组进行 50%重叠的离散余弦变换变换, 得到第 k 帧的频域信号 {y _swb{0\y _swb{\\……,_y_ 6(319)}。
提取第 k 帧的频域信号 { _ 6(0), _ 6(1),…… ,_y_ 6(319)}的频域包络信 息,根据人耳掩蔽效应挑选一组频域谱参数; 对频域包络信息和频域谱参数进 行量化和编码。
将编码后的第 k帧需要编码传送的时域包络信息、频域包络信息、频域语 参数加入最后的码流中打包传输。
进一步,在时域包络信息的数量由输入信号特征决定时, 在组合得到当前 帧需要编码发送的时域包络信息时,可以将需要编码发送的时域包络信息所涉 及的两个帧的时域包络信息的数量也组合在当前帧需要编码发送的时域包络 信息中;
假设, 第 k帧的时域包络信息的数量为 4,第 k-1帧的时域包络信息的数 量为 Jw , 则将本帧信号的
Figure imgf000011_0002
以及指示当前帧提取时域包络个数 和上一帧信号保存在编码状态器中的 ,,Λ 1…… }以及指示上一帧提取时域 包络个数^重新组合为 m …… ,^,4^,4 ,…… },作为当前帧需要编 码传送的时域包络信息进行编码。将本帧信号的 …… ,}保存在编码状 态寄存器中, 用于下一帧信号的时域包络编码。
可以理解的是,上述的当前帧需要编码传送的时域包络信息的组合方式只 是本发明实施例提出的一种实施方式, 并不是限定只能釆用上述方式进行组 合;只要解码端能够正确的区分不同的组合信息就不会影响本发明实施例的实 现。上述仅描述了将时域包络信息分成一个当前帧部分和一个后续帧部分的情 况, 将时域包络信息分成一个当前帧部分和多个后续帧部分的处理流程类似, 此处不再赘述。
图 3描述了本发明实施例提供的解码方法实施例一的流程, 包括: 步骤 301、 解码得到当前帧编码传送的时域包络信息;
步骤 302、 从当前帧编码传送的时域包络信息中, 提取当前帧的时域包络 信息中的当前帧部分;
本发明实施例中的当前帧编码传送的时域包络信息是包括当前帧的时域 包络信息中的当前帧部分、 和当前帧的之前帧的时域包络信息中的后续帧部 分; 因此, 可以从当前帧编码传送的时域包络信息中, 提取当前帧的时域包络 信息中的当前帧部分;
进一步,如果当前帧的时域包络信息的数量由输入信号特征决定, 则当前 帧编码传送的时域包络信息还会包括当前帧的时域包络信息的数量、和当前帧 的之前帧的时域包络信息的数量; 此时, 可以根据所述当前帧的时域包络信息 的数量,从所述当前帧编码传送的时域包络信息中,提取所述当前帧的时域包 络信息中的当前帧部分;
步骤 303、 获取当前帧的时域包络信息的后续帧部分;
当前帧的时域包络信息的后续帧部分可以通过如下流程获得:解码得到当 前帧的后续帧编码传送的时域包络信息; 从后续帧编码传送的时域包络信息 中,提取当前帧的时域包络信息中的后续帧部分; 在时域包络信息只分成了两 部分时, 可以从一个后续帧中获取后续帧部分; 在时域包络信息分成了多于两 部分时, 可以从至少两个后续帧中获取后续帧部分;
如果当前帧是最后一帧, 即没有后续帧时, 当前帧的后续帧部分并没有传 送过来, 此时可以使用预置方式获取当前帧的时域包络信息的后续帧部分, 具 体可以使用当前帧的时域包络信息的当前帧部分得到当前帧的整个时域包络 信息, 例如可以在后续帧部分的时域包络信息的位置填充 0或 1等方式,也可 以使用当前帧部分的信息进行内插处理得到当前帧的整个时域包络信息;
步骤 304、 组合当前帧部分和后续帧部分得到当前帧的时域包络信息。 从上可知,本实施例接收的当前帧的时域包络信息中只有当前帧部分属于 当前帧, 剩余部分由其他帧传送, 因此即使当前帧不能正确解码或丟失, 也可 以从其他帧中获取当前帧的时域包络信息的剩余部分,进而可以使用属于当前 帧的时域包络信息的剩余部分进行解码,由于是使用的属于当前帧的时域包络 信息进行的解码, 因此可以尽量的保持信号的连续性,使听觉感受舒服; 同时, 在当前帧正常传输时,也可以对当前帧的信息进行正确解码, 不会影响信号的 连续性。
图 4描述了本发明实施例提供的解码方法实施例二的流程, 包括: 步骤 401、 解码得到当前帧编码传送的时域包络信息、 频域包络信息和频 域谱参数信息;
步骤 402、 从当前帧编码传送的时域包络信息中, 提取当前帧的时域包络 信息中的当前帧部分;
步骤 403、 获取当前帧的时域包络信息的后续帧部分;
步骤 404、 组合当前帧部分和后续帧部分得到当前帧的时域包络信息; 步骤 405、 使用频域包络信息和频域语参数信息, 获得当前帧的重构频域 信号;
该步骤具体可以包括: 生成随机噪声谱,使用频域包络信息对随机噪声谱 进行能量调整,计算经过能量调整后的随机噪声谱的谱包络信息; 将频域语参 数信息替代谱包络信息, 得到当前帧的重构频域信号;
对随机噪声谱进行能量调整时,可以将随机噪声谱的谱包络信息调整至符 合原始信号的包络信息;
在获得了当前帧的重构频域信号后,可以进一步保存当前帧的重构频域信 号;
步骤 406、对重构频域信号和保存的当前帧的上一帧进行反变换重构时得 到的中间信号, 进行反变换重构得到重构时域信号; 得到重构时域信号后,可以进一步保存反变换重构时得到的当前帧的中间 信号;
每一次进行反变换重构后都可以将反变换重构时得到的当前帧的中间信 号保存, 便于下一帧进行反变换重构时使用;
反变换重构的方式与编码时釆用的变换编码的方式对应,例如在编码时釆 用的是 50%的离散余弦变换变换,则在反变换重构时,要釆用对应的反离散余 弦变换变换和 50%重叠相加;
步骤 407、 对当前帧的时域包络信息进行平滑内插操作, 得到平滑后的时 域包络信息;
步骤 408、 使用重构时域信号和平滑后的时域包络信息, 进行反归一化操 作获得重构的当前帧信息。
从上可知,本实施例接收的当前帧的时域包络信息中只有部分信息属于当 前帧, 剩余部分由其他帧传送, 因此即使当前帧不能正确解码或丟失, 也可以 从其他帧中获取当前帧的时域包络信息的剩余部分,进而可以使用属于当前帧 的时域包络信息的剩余部分进行解码,由于是使用的属于当前帧的时域包络信 息进行的解码, 因此可以尽量的保持信号的连续性, 使听觉感受舒服; 同时, 在当前帧正常传输时,也可以对当前帧的信息进行正确解码, 不会影响信号的 连续性。
本发明实施例还提供了解码方法的实施例三,描述了将时域包络信息分成 两部分时的情况, 流程如下:
从接收到的码流中解码得到 当前第 k 帧的时域包络信息 ^ 4 4 4^4 ^ q k-^ 频域包络信息和频域谱参数信 息, 由于在解码端为了进行离散余弦变换反变换, 会有一帧的时延, 所以在解 码端解码第 k帧的时域包络信息时已经可以解码得到第 k+1帧的时域包络信息 I Aq Aq Aq Aq Aq Aq Aq Aq Aq Aq \
生成一组随机噪声谱,根据解码得到的频域包络信息对噪声谱进行能量调 整,使噪声谱的谱包络符合原始信号的包络信息, 计算经过能量调整的噪声谱 的谱包络信息。
将解码得到的频域语参数代替相应位置的噪声谱的谱包络信息,得到当前 帧的重构频域信号{ _ ¾(0), _ ¾(1),…… ,y_swbk(3\9)}; 进一步, 可以将这组 信号保留在解码状态寄存器中, 以供需要进行帧丟失或不能解码时使用。
将重构频 i或信号 { _ ¾ (0), y _ swbk (1),……, _ ¾ (319)}和上一帧信号进行 反 变 换 重 构 时 保 留 在 解 码 状 态 寄 存 器 中 的 一 组 信 号
Figure imgf000015_0001
…… , imdct k_,(319)} , 进行反变换重构,得到重构时域信号 {y_hikX0ly_hikX\l…… ,y_hik\3\9)} , 反变换重构方式与编码时釆用的变换编码 方式对应, 例如编码时釆用的是离散余弦变换变换,解码时的反变换重构可以 釆用反离散余弦变换变换。进一步, 可以将本帧信号在反离散余弦变换变换时 产生的一组中间信号 {imdctk (0), imdctk (1),…… , imdctk (319)}保留在解码状态 寄存器中, 供下一帧信号进行反变换重构时使用。
从第 k帧的时域包络信息 { , , , , , —" — — — ^和第 k+1帧的时域包络信息 , +1,4 , ,4 , , }中抽取出当 前帧的实际时域包络信息 …… J。对 …… J进行和编码端相 同的平滑内插操作, 得到平滑后的时域包络信息 {。^ u,…… 319^。
利 用 平 滑 后 的 时 域 包 络 对 重 构 的 时 域 信 号
{y _ hik '(0), y_hik '(1),…… , _ '(319)}进行反归一化操作, 得到重构的超宽带信号 {y _ hik (0), y _ hik (1), ······ , _y _ (319) }。 进一步 , 在进 # 归一 4匕才喿作后, 可以进# 一些后处理, 如后滤波处理、 后滤波增强处理等。
图 5描述了本发明实施例提供的解码方法实施例四的流程, 包括: 步骤 501、 解码得到当前帧的后续帧编码传送的时域包络信息; 步骤 502、 从后续帧编码传送的时域包络信息中, 提取当前帧的时域包络 信息中的后续帧部分;
在时域包络信息的数量是固定值时,可以根据该固定值从后续帧编码传送 的时域包络信息中, 提取当前帧的时域包络信息中的后续帧部分;
在时域包络信息的数量由输入信号特征决定时,可以从后续帧编码传送的 时域包络信息中提取当前帧的时域包络信息的数量;然后再根据当前帧的时域 包络信息的数量 , 使用后续帧部分获得当前帧的时域包络信息;
步骤 503、 使用后续帧部分获得当前帧的时域包络信息。
由于当前帧没能正确传送,因此需要使用当前帧的时域包络信息中的后续 帧部分恢复当前帧的时域包络信息,由于后续帧部分只是时域包络信息的一部 分, 因此为了使这一部分信息能够尽可能的还原整个时域包络信息, 可以对后 续帧部分进行内插处理得到当前帧的时域包络信息;
从上可知, 本实施例在当前帧没能正确传送时, 可以使用当前帧的时域包 络信息的后续帧部分还原当前帧的时域包络信息,从而可以使用属于当前帧的 时域包络信息的剩余部分进行解码,由于是使用的属于当前帧的时域包络信息 进行的解码, 因此可以尽量的保持信号的连续性, 使听觉感受舒服。
图 6描述了本发明实施例提供的解码方法实施例五的流程, 包括: 步骤 601、 解码得到当前帧的后续帧编码传送的时域包络信息; 步骤 602、 从后续帧编码传送的时域包络信息中, 提取当前帧的时域包络 信息中的后续帧部分;
步骤 603、 使用后续帧部分获得当前帧的时域包络信息;
步骤 604、 将保存的当前帧的上一帧的重构频域信号、 和反变换重构时得 到的上一帧的中间信号进行反变换重构 , 获得当前帧的重构时域信号;
进一步,获得重构时域信号好可以进一步保存进行反变换重构时产生的当 前帧的中间信号;
其中,保存进行反变换重构时产生的当前帧的中间信号, 可以在下一帧进 行反变换重构时使用;
反变换重构的方式与编码时釆用的变换编码的方式对应,例如在编码时釆 用的是 50%的离散余弦变换变换,则在反变换重构时,要釆用对应的反离散余 弦变换变换和 50%重叠相加;
步骤 605、 对当前帧的时域包络信息进行平滑内插操作, 获得当前帧的平 滑后的时域包络信息;
步骤 606、 使用平滑后的时域包络信息对重构时域信号进行反归一化操 作, 获得重构的当前帧信息。
从上可知, 本实施例在当前帧没能正确传送时, 可以使用当前帧的时域包 络信息的后续帧部分还原当前帧的时域包络信息,从而可以使用属于当前帧的 时域包络信息的剩余部分进行解码,由于是使用的属于当前帧的时域包络信息 进行的解码, 因此可以尽量的保持信号的连续性, 使听觉感受舒服。 本发明实施例还提供了解码方法的实施例六,描述的也是将时域包络信息 分成两部分时的情况, 流程如下:
当前第 k帧丟失后, 已经无法得到第 k帧的频域包络和频域语参数, 可以 从 码 流 中 解 码 得 到 第 k+1 帧 的 时 域 包 络 信 息 { +1, +1 , +1,4 , ,4 , ,其中有属于第 k帧的时域包络信 息的后续帧部分。
将 保 留 在 解 码 寄 存 器 中 上 一 帧 重 构 频 域 信 号 {y _ swbk_x (0), y _swbk_x (1),……, _ ¾— , (319)}和上一帧信号在反离散余弦变换变换 时产生的一组保留在解码寄存器中的信号
Figure imgf000017_0001
…… , imdct k_,(319)} 一起进行反变换重构, 得到重构时域信号 {y_hik \0\y_hik '(1),…… ,y— hik '(319)} , 并 将 第 k 帧 信 号 在 反 变 换 重 构 时 产 生 的 中 间 信 号 {imdctk (0), imdctk (1),…… , imdctk (319)}保留在解码状态寄存器中,供下一帧信号进行 反变换重构时使用。
从 k+1 帧的时域包络信息 , +1,4 , ,4 , , }中 抽取出当前第 k帧的时域包络信息 对其进行内插, 得到 对 当 前 第 k 帧 的 一 个 刻 画 较 为 粗糙 的 时 域 包 络信 息
Figure imgf000017_0002
' =(4^+4D = 2, 4,6,8。
Figure imgf000017_0003
进行和编码端相同的反量化、 平 滑操作, 得到平滑后的时域包络信息 {Uu,…… 4^}。
利 用 平 滑 后 的 时 域 包 络 对 重 构 的 时 域 信 号
{y _ hik '(0), y_hik '(1),…… , _ '(319)}进行反归一化操作, 得到重构的超宽带信号 {y _ hik (0), y _ hik (1), ······ , _y _ (319) }。 进一步 , 在进 # 归一 4匕才喿作后, 可以进# 一些后处理, 如后滤波处理、 后滤波增强处理等。
其中, 需要说明的是, 本发明实施例的上述流程仅是本发明实施例提供的 实现方式, 并不是限定各个步骤仅能按照上述流程进行。
本发明实施例还提供了编码装置,图 7描述了本发明实施例提供的编码装 置实施例的结构, 包括:
时域包络信息获取单元 701, 用于获取当前帧的量化后的时域包络信息; 拆分单元 702 ,用于将时域包络信息获取单元 701获取的量化后的时域包 络信息分成至少两部分, 包括当前帧部分和后续帧部分;
组合单元 703 , 用于组合当前帧部分和之前帧数据的后续帧部分,得到当 前帧需要编码传送的时域包络信息;
编码发送单元 704 ,用于编码并发送组合单元 703组合得到的当前帧需要 编码传送的时域包络信息。
从上可知, 本实施例将当每个帧都分成至少两部分, 包括当前帧部分和后 续帧部分, 并将当前帧部分和后续帧部分放在至少两个帧的编码中分发送, 这 样编码发送的时域包络信息中只有当前帧部分是属于当前帧的,而其余部分放 在后续帧的时域包络信息中发送,从而如果解码端不能对当前帧正确解码,也 能从后续帧中提取到属于当前帧的时域包络信息,进而可以使用属于当前帧的 时域包络信息进行解码, 由于是使用的属于当前帧的时域包络信息进行的解 码, 因此可以尽量的保持信号的连续性, 使听觉感受舒服。
进一步, 本发明实施例提供的编码装置还可以包括:
频域信号获取单元,用于在时域包络信息获取单元 701获取量化后的时域 包络信息后, 使用量化后的时域包络信息获得当前帧的频域信号;
频域包络信息提取单元, 用于提取频域信号的频域包络信息,选取频域谱 参数;
量化单元,用于量化频域包络信息提取单元提取的频域包络信息和选取的 频域语参数, 得到量化后的频域包络信息和量化后的频域语参数;
此时, 编码发送单元 704 , 还用于编码并发送量化后的频域包络信息和量 化后的频域语参数。
其中, 上述频域信号获取单元具体可以包括: 平滑单元, 用于在量化单元 得到量化后的时域包络信息后, 对量化后的时域包络信息进行平滑内插处理, 得到平滑后的时域包络; 归一化单元, 用于使用平滑后的时域包络对时域信号 进行归一化处理, 得到归一化后的时域信号; 变换编码单元, 用于使用归一化 后的时域信号与上一帧数据的归一化后的时域信号构成数组;对数组进行变换 编码得到当前帧的频域信号。
本发明实施例还提供了解码装置,图 8描述了本发明实施例提供的解码装 置实施例一的结构包括:
解码单元 801 , 用于解码得到当前帧编码传送的时域包络信息;
提取单元 802 , 用于从当前帧编码传送的时域包络信息中,提取当前帧的 时域包络信息中的当前帧部分;
获取单元 803 , 用于获取当前帧的时域包络信息中的后续帧部分; 组合单元 804 ,用于组合当前帧部分和后续帧部分得到当前帧的时域包络 信息。
从上可知,本实施例接收的当前帧的时域包络信息中只有当前帧部分属于 当前帧, 剩余部分由其他帧传送, 因此即使当前帧不能正确解码或丟失, 也可 以从其他帧中获取当前帧的时域包络信息的剩余部分,进而可以使用属于当前 帧的时域包络信息的剩余部分进行解码,由于是使用的属于当前帧的时域包络 信息进行的解码, 因此可以尽量的保持信号的连续性,使听觉感受舒服; 同时, 在当前帧正常传输时,也可以对当前帧的信息进行正确解码, 不会影响信号的 连续性。
进一步, 上述解码单元 801 , 还用于解码得到当前帧的频域包络信息和频 域语参数信息; 此时, 本发明实施例提供的上述解码装置还可以包括:
重构单元, 用于使用频域包络信息和频域语参数信息, 获得当前帧的重构 频域信号;
反变换重构单元,用于对重构频域信号和保存的当前帧的上一帧的中间信 号, 进行反变换重构得到重构时域信号;
平滑内插单元, 用于对当前帧的时域包络信息进行平滑内插操作,得到平 滑后的时域包络信息;
反归一化单元, 用于使用重构时域信号和平滑后的时域包络信息, 进行反 归一化操作获得重构的当前帧信息。
进一步, 本发明实施例提供的解码装置实施例一还可以包括保存单元, 用 于在重构单元获得当前帧的重构频域信号后,保存当前帧的重构频域信号; 还 用于在反变换重构单元得到重构时域信号后,保存反变换重构时得到的当前帧 的中间信号;
图 9描述了本发明实施例提供的解码装置实施例二的结构, 包括: 解码单元 901 , 用于解码得到当前帧的后续帧编码传送的时域包络信息; 提取单元 902, 用于从后续帧编码传送的时域包络信息中,提取当前帧的 时域包络信息中的后续帧部分;
时域包络信息获取单元 903 ,用于使用后续帧部分获得当前帧的时域包络 信息。
从上可知, 本实施例在当前帧没能正确传送时, 可以使用当前帧的时域包 络信息的后续帧部分还原当前帧的时域包络信息,从而可以使用属于当前帧的 时域包络信息的剩余部分进行解码,由于是使用的属于当前帧的时域包络信息 进行的解码, 因此可以尽量的保持信号的连续性, 使听觉感受舒服。
进一步, 本发明实施例提供的解码装置实施例二还可以包括:
重构时域信号获取单元, 用于将保存的当前帧的上一帧的重构频域信号、 和反变换重构时得到的上一帧的中间信号进行反变换重构 ,获得当前帧的重构 时域信号;
平滑单元, 用于对当前帧的时域包络信息进行平滑内插操作, 获得当前帧 的平滑后的时域包络信息;
反归一化单元,用于使用平滑后的时域包络信息对重构时域信号进行反归 一化操作 , 获得重构的当前帧信息。
进一步, 本发明实施例提供的解码装置实施例二还可以包括保存单元, 用 于在重构时域信号获取单元获得当前帧的重构时域信号后 ,保存进行反变换重 构时产生的当前帧的中间信号。
本发明实施例还提供了编解码系统, 图 10描述了编解码系统实施例一的 结构, 包括:
编码装置 1001 , 用于获取当前帧的量化后的时域包络信息; 将量化后的 时域包络信息分成至少两部分,包括至少一个当前帧部分和至少一个后续帧部 分; 组合当前帧部分和当前帧的之前帧数据的后续帧部分,得到当前帧需要编 码传送的时域包络信息,之前帧数据的数量与后续帧的数量相同; 编码得到的 当前帧需要编码传送的时域包络信息 ,发送编码后的当前帧需要编码传送的时 域包络信息;
解码装置 1002 , 用于解码得到当前帧编码传送的时域包络信息; 从当前 帧编码传送的时域包络信息中, 提取当前帧的时域包络信息中的当前帧部分; 获取当前帧的时域包络信息中的后续帧部分;组合当前帧部分和后续帧部分得 到当前帧的时域包络信息。
从上可知,本实施例编码装置发送的当前帧的时域包络信息中只有当前帧 部分属于当前帧, 剩余部分由其他帧传送, 因此即使当前帧不能正确解码或丟 失,也可以从其他帧中获取当前帧的时域包络信息的剩余部分, 进而可以使用 属于当前帧的时域包络信息的剩余部分进行解码,由于是使用的属于当前帧的 时域包络信息进行的解码, 因此可以尽量的保持信号的连续性,使听觉感受舒 服; 同时, 在当前帧正常传输时, 也可以对当前帧的信息进行正确解码, 不会 影响信号的连续性。
图 11描述了编解码系统实施例二的结构, 包括:
编码装置 1101 , 用于获取当前帧的量化后的时域包络信息; 将量化后的 时域包络信息分成至少两部分,包括至少一个当前帧部分和至少一个后续帧部 分; 组合当前帧部分和当前帧的之前帧数据的后续帧部分,得到当前帧需要编 码传送的时域包络信息,之前帧数据的数量与后续帧的数量相同; 编码得到的 当前帧需要编码传送的时域包络信息 ,发送编码后的当前帧需要编码传送的时 域包络信息;
解码装置 1102 , 用于解码得到当前帧的后续帧编码传送的时域包络信息; 从后续帧编码传送的时域包络信息中,提取当前帧的时域包络信息中的后续帧 部分; 使用后续帧部分获得当前帧的时域包络信息。
从上可知, 本实施例中解码装置在当前帧没能正确传送时, 可以使用当前 帧的时域包络信息的后续帧部分还原当前帧的时域包络信息,从而可以使用属 于当前帧的时域包络信息的剩余部分进行解码,由于是使用的属于当前帧的时 域包络信息进行的解码,因此可以尽量的保持信号的连续性,使听觉感受舒服。
本领域普通技术人员可以理解实现上述实施例方法中的全部或部分步骤 是可以通过程序来指令相关的硬件完成,所述的程序可以存储于一种计算机可 读存储介质中, 该程序在执行时, 包括如下步骤:
获取当前帧的量化后的时域包络信息;
将所述量化后的时域包络信息分成至少两部分,包括当前帧部分和后续帧 部分;
组合所述当前帧部分和之前帧数据的后续帧部分,得到当前帧需要编码传 送的时域包络信息;
编码并发送所述当前帧需要编码传送的时域包络信息。
和 /或包括如下步骤:
解码得到当前帧编码传送的时域包络信息;
从所述当前帧编码传送的时域包络信息中,提取所述当前帧的时域包络信 息中的当前帧部分;
获取所述当前帧的时域包络信息中的后续帧部分;
组合所述当前帧部分和所述后续帧部分得到所述当前帧的时域包络信息。 和 /或包括如下步骤:
解码得到所述当前帧的后续帧编码传送的时域包络信息;
从所述后续帧编码传送的时域包络信息中,提取所述当前帧的时域包络信 息中的后续帧部分;
使用所述后续帧部分获得所述当前帧的时域包络信息。
上述提到的存储介质可以是只读存储器, 磁盘或光盘等。
以上对本发明实施例所提供的编码方法及装置、解码方法及装置进行了详 细介绍, 以上实施例的说明只是用于帮助理解本发明的方法及其思想; 同时, 对于本领域的一般技术人员,依据本发明的思想, 在具体实施方式及应用范围 上均会有改变之处, 综上所述, 本说明书内容不应理解为对本发明的限制。

Claims

权 利 要 求
1、 一种编码方法, 其特征在于, 包括:
获取当前帧的量化后的时域包络信息;
将所述量化后的时域包络信息分成至少两部分,包括至少一个当前帧部分 和至少一个后续帧部分;
组合所述当前帧部分和所述当前帧的之前帧数据的后续帧部分,得到当前 帧需要编码传送的时域包络信息;
编码所述当前帧需要编码传送的时域包络信息;
发送编码后的所述当前帧需要编码传送的时域包络信息。
2、 如权利要求 1所述的编码方法, 其特征在于, 得到量化后的时域包络 信息后进一步包括:
使用所述量化后的时域包络信息获得所述当前帧的频域信号;
提取所述频域信号的频域包络信息, 选取频域谱参数;
量化所述频域包络信息和所述频域语参数,得到量化后的频域包络信息和 量化后的频域谱参数;
所述编码并发送所述当前帧需要编码传送的时域包络信息包括: 编码并发送所述当前帧需要编码传送的时域包络信息、量化后的频域包络 信息和量化后的频域谱参数。
3、 如权利要求 2所述的编码方法, 其特征在于, 所述使用所述量化后的 时域包络信息获得所述当前帧的频域信号的步骤包括:
对所述量化后的时域包络信息进行平滑内插处理, 得到平滑后的时域包 络;
使用所述平滑后的时域包络对所述时域信号进行归一化处理,得到归一化 后的时域信号;
使用所述归一化后的时域信号与所述之前帧数据的归一化后的时域信号 构成数组;
对所述数组进行变换编码得到所述当前帧的频域信号。
4、 如权利要求 1至 3任一所述的编码方法, 其特征在于, 将所述量化后 的时域包络信息分成至少两部分具体为: 将所述量化后的时域包络信息分成两部分;所述当前帧部分是所述量化后 的时域包络信息的奇数部分;
所述当前帧部分是所述量化后的时域包络信息的偶数部分。
5、 如权利要求 4所述的编码方法, 其特征在于, 提取的所述时域信号的 时域包络信息的数量由所述时域信号的输入信号特征确定。
6、 如权利要求 5所述的编码方法, 其特征在于, 组合所述当前帧部分和 之前帧数据的后续帧部分,得到当前帧需要编码传送的时域包络信息的步骤包 括:
组合所述当前帧部分、所述之前帧数据的后续帧部分、提取的所述当前帧 的时域包络信息的数量和所述之前帧数据的时域包络信息的数量,得到所述当 前帧需要编码传送的时域包络信息。
7、 如权利要求 4所述的编码方法, 其特征在于, 组合所述当前帧部分和 之前帧数据的后续帧部分,得到当前帧需要编码传送的时域包络信息时,如果 所述当前帧为第一帧,所述之前帧数据的后续帧部分为预先设置的之前帧数据 的后续帧部分。
8、 一种计算机程序产品, 其特征在于, 所述计算机程序产品包括计算机 程序代码, 当所述计算机程序代码被一个计算机执行的时候, 所述计算机程序 代码可以使得所述计算机执行权利要求 1至 7项中任意一项的步骤。
9、 一种计算机可读存储介质, 其特征在于, 所述计算机可读存储介质存 储计算机程序代码, 当所述计算机程序代码被一个计算机执行的时候, 所述计 算机程序代码可以使得所述计算机执行权利要求 1至 7项中任意一项的步骤。
10、 一种解码方法, 其特征在于, 包括:
解码得到当前收到的帧编码传送的时域包络信息,所述当前收到的帧编码 传送的时域包络信息包括所述当前收到的帧的时域包络信息中的当前帧部分 以及在收到当前收到的帧之前收到的帧的时域包络信息中的后续帧部分; 从所述当前收到的帧编码传送的时域包络信息中,提取所述当前收到的帧 的时域包络信息中的当前帧部分;
获取所述当前收到的帧的时域包络信息中的后续帧部分; 组合所述当前收到的帧的时域包络信息中的当前帧部分和所述当前收到 的帧的时域包络信息中的后续帧部分, 得到所述当前收到的帧的时域包络信 息。
11、 如权利要求 10所述的解码方法, 其特征在于, 从所述当前收到的帧 编码传送的时域包络信息中,提取所述当前收到的帧的时域包络信息中的当前 帧部分的步骤包括:
从所述当前收到的帧编码传送的时域包络信息中提取所述当前收到的帧 的时域包络信息的数量;
根据所述当前收到的帧的时域包络信息的数量,从所述当前收到的帧编码 传送的时域包络信息中, 提取所述当前帧的时域包络信息中的当前帧部分。
12、 如权利要求 10所述的解码方法, 其特征在于, 所述获取所述当前收 到的帧的时域包络信息中的后续帧部分具体为:
解码得到在当前收到的帧之后收到的续帧编码传送的时域包络信息; 从所述之后收到的续帧编码传送的时域包络信息中,提取所述当前收到的 帧的时域包络信息中的后续帧部分。
13、 如权利要求 10所述的解码方法, 其特征在于, 所述当前收到的帧为 最后一帧, 所述获取所述当收到的前帧的时域包络信息中的后续帧部分具体 为:
使用预置方式获取所述当前收到的帧的时域包络信息中的后续帧部分。
14、 如权利要求 10或 13所述的解码方法, 其特征在于, 还包括: 解码得到所述当前收到的帧的频域包络信息和频域语参数信息; 使用所述当前收到的帧的频域包络信息和所述频域语参数信息,获得所述 当前收到的帧的重构频域信号;
对所述重构频域信号和保存的在当前收到的帧的之前收到的帧数据的中 间信号, 进行反变换重构得到重构时域信号;
对所述当前收到的帧的时域包络信息进行平滑内插操作,得到平滑后的时 域包络信息;
使用所述重构时域信号和所述平滑后的时域包络信息,进行反归一化操作 获得重构的当前帧信息。
15、 如权利要求 14所述的解码方法, 其特征在于, 所述使用所述频域包 络信息和所述频域谱参数信息获得所述当前收到的帧的重构频域信号具体为: 生成随机噪声谱, 使用所述频域包络信息对所述随机噪声谱进行能量调 整, 计算经过能量调整后的随机噪声谱的谱包络信息;
将所述频域语参数信息替代所述语包络信息,得到所述当前收到的帧的重 构频域信号。
16、 一种解码方法, 其特征在于, 包括:
解码得到在当前收到的帧之后收到的帧编码传送的时域包络信息,所述编 码传送的时域包络信息包括所述之后收到的帧的时域包络信息中的当前帧部 分和所述当前收到帧的时域包络信息中的后续帧部分;
从所述之后收到的帧编码传送的时域包络信息中,提取所述当前收到的帧 的时域包络信息中的后续帧部分;
使用所述后续帧部分获得所述当前收到的帧的时域包络信息。
17、 如权利要求 16所述的解码方法, 其特征在于, 使用所述后续帧部分 获得所述当前收到的帧的时域包络信息的步骤包括:
从所述之后收到的帧编码传送的时域包络信息中提取所述当前收到的帧 的时域包络信息的数量;
根据所述当前收到的帧的时域包络信息的数量,使用所述后续帧部分获得 所述当前收到的帧的时域包络信息。
18、 如权利要求 16或 17所述的解码方法, 其特征在于, 还包括: 将保存的在当前收到的帧的之前收到的帧数据的重构频域信号、与所述之 前收到的帧数据进行反变换重构时得到的所述之前收到的帧数据的中间信号 进行反变换重构, 获得所述当前收到的帧的重构时域信号;
对所述当前收到的帧的时域包络信息进行平滑内插操作,获得所述当前收 到的帧的平滑后的时域包络信息;
使用所述平滑后的时域包络信息对所述重构时域信号进行反归一化操作, 获得重构的当前帧信息。
19、 一种计算机程序产品, 其特征在于, 所述计算机程序产品包括计算机 程序代码, 当所述计算机程序代码被一个计算机执行的时候, 所述计算机程序 代码可以使得所述计算机执行权利要求 10至 18项中任意一项的步骤。
20、 一种计算机可读存储介质, 其特征在于, 所述计算机可读存储介质存 储计算机程序代码, 当所述计算机程序代码被一个计算机执行的时候, 所述计 算机程序代码可以使得所述计算机执行权利要求 10至 18项中任意一项的步 骤。
21、 一种编码装置, 其特征在于, 包括:
时域包络信息获取单元, 用于获取当前帧的量化后的时域包络信息; 拆分单元,用于将所述时域包络信息获取单元获取的量化后的时域包络信 息分成至少两部分, 包括至少一个当前帧部分和至少一个后续帧部分;
组合单元,用于组合所述当前帧部分和所述当前帧的之前帧数据的后续帧 部分,得到当前帧需要编码传送的时域包络信息, 所述之前帧数据的数量与所 述后续帧的数量相同;
编码发送单元,用于编码所述组合单元组合得到的当前帧需要编码传送的 时域包络信息, 发送编码后的当前帧需要编码传送的时域包络信息。
22、 如权利要求 21所述的编码装置, 其特征在于, 还包括:
频域信号获取单元,用于在所述时域包络信息获取单元获取了量化后的时 域包络信息时, 使用所述量化后的时域包络信息获得所述当前帧的频域信号; 频域包络信息提取单元, 用于提取所述频域信号的频域包络信息,选取频 域谱参数;
量化单元,用于量化所述频域包络信息提取单元提取的频域包络信息和选 取的所述频域谱参数, 得到量化后的频域包络信息和量化后的频域谱参数; 所述编码发送单元,还用于编码并发送所述量化后的频域包络信息和量化 后的频域谱参数。
23、 如权利要求 22所述的编码装置, 其特征在于, 所述频域信号获取单 元包括:
平滑单元, 用于在所述量化单元得到量化后的时域包络信息后,对所述量 化后的时域包络信息进行平滑内插处理, 得到平滑后的时域包络;
归一化单元,用于使用所述平滑后的时域包络对所述时域信号进行归一化 处理, 得到归一化后的时域信号; 变换编码单元,用于使用所述归一化后的时域信号与所述之前帧数据的归 一化后的时域信号构成数组;对所述数组进行变换编码得到所述当前帧的频域 信号。
24、 一种解码装置, 其特征在于, 包括:
解码单元, 用于解码得到当前收到的帧编码传送的时域包络信息, 所述当 前收到的帧编码传送的时域包络信息包括所述当前收到的帧的时域包络信息 中的当前帧部分以在收到当前收到的帧之前收到的帧的时域包络信息中的后 续帧部分;
提取单元, 用于从所述当前收到的帧编码传送的时域包络信息中,提取所 述当前收到的帧的时域包络信息中的当前帧部分;
获取单元, 用于获取所述当前收到的帧的时域包络信息中的后续帧部分; 组合单元,用于组合所述当前收到的帧的时域包络信息中的当前帧部分和 所述当前收到的帧的时域包络信息中的后续帧部分,得到所述当前收到的帧的 时域包络信息。
25、 如权利要求 24所述的解码装置, 其特征在于, 所述解码单元, 还用 于解码得到所述当前收到的帧的频域包络信息和频域语参数信息;
所述解码装置还包括:
重构单元,用于使用所述当前收到的帧的频域包络信息和所述频域语参数 信息, 获得所述当前收到的帧的重构频域信号;
反变换重构单元, 用于对所述重构频域信号、和保存的在当前收到的帧的 之前收到的帧进行反变换重构获得的中间信号,进行反变换重构得到重构时域 信号;
平滑内插单元,用于对所述当前收到的帧的时域包络信息进行平滑内插操 作, 得到平滑后的时域包络信息;
反归一化单元, 用于使用所述重构时域信号和所述平滑后的时域包络信 息, 进行反归一化操作获得重构的当前帧信息。
26、 一种解码装置, 其特征在于, 包括:
解码单元,用于解码得到在当前收到的帧之后收到的帧编码传送的时域包 络信息,所述编码传送的时域包络信息包括所述之后收到的帧的时域包络信息 中的当前帧部分和所述当前收到帧的时域包络信息中的后续帧部分; 提取单元, 用于从所述之后收到的帧编码传送的时域包络信息中,提取所 述当前收到的帧的时域包络信息中的后续帧部分;
时域包络信息获取单元,用于使用所述后续帧部分获得所述当前收到的帧 的时域包络信息。
27、 如权利要求 26所述的解码装置, 其特征在于, 还包括:
重构时域信号获取单元,用于将保存的在当前收到的帧的之前收到的帧的 重构频域信号、和所述之前收到的帧进行反变换重构时得到的中间信号进行反 变换重构, 获得所述当前收到的帧的重构时域信号;
平滑单元, 用于对所述当前收到的帧的时域包络信息进行平滑内插操作, 获得所述当前收到的帧的平滑后的时域包络信息;
反归一化单元,用于使用所述平滑后的时域包络信息对所述重构时域信号 进行反归一化操作, 获得重构的当前帧信息。
28、 一种编解码系统, 其特征在于, 包括:
编码装置, 用于获取当前帧的量化后的时域包络信息; 将所述量化后的时 域包络信息分成至少两部分, 包括至少一个当前帧部分和至少一个后续帧部 分; 组合所述当前帧部分和所述当前帧的之前帧数据的后续帧部分,得到当前 帧需要编码传送的时域包络信息,所述之前帧数据的数量与所述后续帧的数量 相同; 编码得到的当前帧需要编码传送的时域包络信息,发送编码后的当前帧 需要编码传送的时域包络信息;
解码装置, 用于解码得到所述当前收到的帧编码传送的时域包络信息; 从 所述当前收到的帧编码传送的时域包络信息中,提取所述当前收到的帧的时域 包络信息中的当前帧部分; 获取所述当前帧的时域包络信息中的后续帧部分; 组合所述当前收到的帧的当前帧部分和所述当前收到的帧的后续帧部分得到 所述当前收到的帧的时域包络信息。
29、 一种编解码系统, 其特征在于, 包括:
编码装置, 用于获取当前帧的量化后的时域包络信息; 将所述量化后的时 域包络信息分成至少两部分, 包括至少一个当前帧部分和至少一个后续帧部 分; 组合所述当前帧部分和所述当前帧的之前帧数据的后续帧部分,得到当前 帧需要编码传送的时域包络信息,所述之前帧数据的数量与所述后续帧的数量 相同; 编码得到的当前帧需要编码传送的时域包络信息,发送编码后的当前帧 需要编码传送的时域包络信息;
解码装置,用于解码得到在当前收到的帧之后收到的帧编码传送的时域包 络信息; 从所述之后收到的帧编码传送的时域包络信息中,提取所述当前收到 的帧的时域包络信息中的后续帧部分;使用所述当前收到的帧的时域包络信息 中的后续帧部分获得所述当前收到的帧的时域包络信息。
PCT/CN2009/071022 2008-04-01 2009-03-26 编码方法及装置、解码方法及装置 WO2009121280A1 (zh)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN2008100891481A CN101552008B (zh) 2008-04-01 2008-04-01 语音编码方法及装置、语音解码方法及装置
CN200810089148.1 2008-04-01

Publications (1)

Publication Number Publication Date
WO2009121280A1 true WO2009121280A1 (zh) 2009-10-08

Family

ID=41134837

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2009/071022 WO2009121280A1 (zh) 2008-04-01 2009-03-26 编码方法及装置、解码方法及装置

Country Status (2)

Country Link
CN (1) CN101552008B (zh)
WO (1) WO2009121280A1 (zh)

Families Citing this family (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN103280222B (zh) * 2013-06-03 2014-08-06 腾讯科技(深圳)有限公司 音频编码、解码方法及其系统
CN113782039A (zh) * 2017-08-10 2021-12-10 华为技术有限公司 时域立体声编解码方法和相关产品
CN120236596A (zh) * 2023-12-29 2025-07-01 北京字跳网络技术有限公司 编码方法、编码装置、解码方法、解码装置和传输系统

Citations (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN1359513A (zh) * 1999-06-30 2002-07-17 松下电器产业株式会社 话音解码装置和编码差错补偿方法
CN1441949A (zh) * 2000-05-11 2003-09-10 艾利森电话股份有限公司 语音编码中的前向纠错
CN1497864A (zh) * 2002-10-24 2004-05-19 松下电器产业株式会社 抗突发错误力强的通信装置和方法及执行程序与记录媒体
US20050049853A1 (en) * 2003-09-01 2005-03-03 Mi-Suk Lee Frame loss concealment method and device for VoIP system
CN1732512A (zh) * 2002-12-31 2006-02-08 诺基亚有限公司 用于隐蔽压缩域分组丢失的方法和装置
CN101071568A (zh) * 2005-11-23 2007-11-14 美国博通公司 音频解码的方法及系统
CN101136201A (zh) * 2006-08-11 2008-03-05 美国博通公司 对音频信号中认为丢失的一部分进行替换的系统及方法

Patent Citations (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN1359513A (zh) * 1999-06-30 2002-07-17 松下电器产业株式会社 话音解码装置和编码差错补偿方法
CN1441949A (zh) * 2000-05-11 2003-09-10 艾利森电话股份有限公司 语音编码中的前向纠错
CN1497864A (zh) * 2002-10-24 2004-05-19 松下电器产业株式会社 抗突发错误力强的通信装置和方法及执行程序与记录媒体
CN1732512A (zh) * 2002-12-31 2006-02-08 诺基亚有限公司 用于隐蔽压缩域分组丢失的方法和装置
US20050049853A1 (en) * 2003-09-01 2005-03-03 Mi-Suk Lee Frame loss concealment method and device for VoIP system
CN101071568A (zh) * 2005-11-23 2007-11-14 美国博通公司 音频解码的方法及系统
CN101136201A (zh) * 2006-08-11 2008-03-05 美国博通公司 对音频信号中认为丢失的一部分进行替换的系统及方法

Also Published As

Publication number Publication date
CN101552008B (zh) 2011-11-16
CN101552008A (zh) 2009-10-07

Similar Documents

Publication Publication Date Title
US20220180881A1 (en) Speech signal encoding and decoding methods and apparatuses, electronic device, and storage medium
CN112119457B (zh) 可截断的预测编码
CN103069484B (zh) 时/频二维后处理
JP6462653B2 (ja) オーディオ・データを処理するための方法、装置、及びシステム
US10734003B2 (en) Noise signal processing method, noise signal generation method, encoder, decoder, and encoding and decoding system
US9443534B2 (en) Bandwidth extension system and approach
JP2011504249A (ja) 信号処理方法及び装置
US11526734B2 (en) Method and apparatus for recurrent auto-encoding
WO2010130225A1 (zh) 一种音频解码方法和音频解码器
JP2010540992A (ja) ノイズ生成装置及び方法
WO2023197809A1 (zh) 一种高频音频信号的编解码方法和相关装置
WO2014117458A1 (zh) 高频带信号的预测方法、编/解码设备
WO2009109120A1 (zh) 一种音频信号的编解码方法和装置
WO2010000179A1 (zh) 频带扩展的方法、系统和设备
WO2009121280A1 (zh) 编码方法及装置、解码方法及装置
WO2008064577A1 (fr) Procédé et appareil permettant d'ajuster la qualité de la quantification dans un codeur et décodeur
BR112016022764B1 (pt) Aparelho e métodos de comutação de tecnologias de codificação em um dispositivo
CN119132318A (zh) 音频编码方法、音频解码方法、装置、可读存储介质
CN116137151B (zh) 低码率网络连接中提供高质量音频通信的系统和方法
WO2024169532A1 (zh) 有损编解码器和无损编解码器之间的切换方法和装置
WO2009152723A1 (zh) 嵌入式编解码方法和装置
HK40052888A (zh) 語音信號的編解碼方法、裝置、電子設備及存儲介質
CN116137151A (zh) 低码率网络连接中提供高质量音频通信的系统和方法

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 09726879

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 09726879

Country of ref document: EP

Kind code of ref document: A1