WO2010102446A1 - 一种线性预测分析方法、装置及*** - Google Patents

一种线性预测分析方法、装置及*** Download PDF

Info

Publication number
WO2010102446A1
WO2010102446A1 PCT/CN2009/070729 CN2009070729W WO2010102446A1 WO 2010102446 A1 WO2010102446 A1 WO 2010102446A1 CN 2009070729 W CN2009070729 W CN 2009070729W WO 2010102446 A1 WO2010102446 A1 WO 2010102446A1
Authority
WO
WIPO (PCT)
Prior art keywords
signal
input signal
characteristic information
linear predictive
windowed
Prior art date
Application number
PCT/CN2009/070729
Other languages
English (en)
French (fr)
Inventor
许剑峰
苗磊
齐峰岩
张德军
张清
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
Priority to CN2009800001092A priority Critical patent/CN102067211B/zh
Priority to PCT/CN2009/070729 priority patent/WO2010102446A1/zh
Priority to KR1020117023175A priority patent/KR101397512B1/ko
Priority to EP09841315.6A priority patent/EP2407963B1/en
Publication of WO2010102446A1 publication Critical patent/WO2010102446A1/zh
Priority to US13/228,965 priority patent/US8812307B2/en

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
    • G10L19/02Speech 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 spectral analysis, e.g. transform vocoders or subband vocoders
    • G10L19/022Blocking, i.e. grouping of samples in time; Choice of analysis windows; Overlap factoring
    • 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
    • G10L19/06Determination or coding of the spectral characteristics, e.g. of the short-term prediction coefficients
    • HELECTRICITY
    • H03ELECTRONIC CIRCUITRY
    • H03MCODING; DECODING; CODE CONVERSION IN GENERAL
    • H03M7/00Conversion of a code where information is represented by a given sequence or number of digits to a code where the same, similar or subset of information is represented by a different sequence or number of digits
    • H03M7/30Compression; Expansion; Suppression of unnecessary data, e.g. redundancy reduction
    • 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 the field of communications, and in particular, to a linear prediction analysis method, apparatus, and system. Background technique
  • the corresponding speech and audio coding technology has been widely used. At present, it is mainly divided into lossy coding and lossless coding.
  • the reconstructed signal of lossy coding cannot be completely consistent with the original signal. However, the redundant information of the signal can be minimized according to the characteristics of the sound source and the sensing characteristics of the person.
  • Lossless coding must ensure that the reconstructed signal is exactly the same as the original signal, so that the final decoding quality is not damaged.
  • the lossy coding compression rate is relatively high, but the reconstructed speech quality is not guaranteed. Lossless coding can guarantee voice quality, but compression. The rate is low, about 50%.
  • the Linear Prediction Coding (LPC) model is widely used in the field of speech coding.
  • LPC Linear Prediction Coding
  • the code excitation linear prediction coding model is the success of its typical application.
  • the basic principle is: firstly use short-term linear prediction to remove the near-point redundancy of the speech signal, then use the long-term predictor to remove the far-point redundancy of the speech signal, and finally the parameters generated in the prediction process and the two-stage
  • the residual signal obtained by the pre-J is encoded and transmitted.
  • most linear predictive analysis of lossy and lossless audio codecs generally includes three modules: windowing, autocorrelation and Levinson algorithm.
  • the residual signal is obtained by linear prediction, and the residual signal is encoded by entropy coding.
  • the linear prediction analysis is performed twice on the input signal, and the short window is added to the signal once, and the long window is added to the signal.
  • the linear prediction analysis is performed twice, which makes the linear prediction analysis more complicated. Summary of the invention
  • Embodiments of the present invention provide a linear prediction analysis method, apparatus, and system, which can improve linear prediction performance and reduce analysis operation complexity.
  • a linear predictive analysis method including:
  • the window function is adaptively windowed to obtain the windowed signal; the windowed signal is processed to obtain the linear predictive coding coefficient for linear prediction.
  • the linear prediction analysis method provided by the embodiment of the present invention obtains the result by analyzing the input signal, and adaptively allocates the window function required for windowing according to the analysis result, thereby being able to increase the coding complexity. , improved prediction performance of linear predictive coding.
  • a linear predictive analysis device includes:
  • An acquiring unit configured to acquire signal characteristic information of at least one sample of the input signal
  • the analyzing unit is configured to compare and analyze the signal characteristic information to obtain an analysis result;
  • the windowing unit is configured to adaptively window the input signal according to the analysis result selection window function to obtain the windowed signal;
  • a processing unit is configured to process the windowed signal to obtain a linear predictive coding coefficient for linear prediction.
  • the linear predictive analysis device obtained by the embodiment of the present invention obtains a result by analyzing an input signal, and adaptively allocates a window function required for windowing according to the analysis result, thereby being able to increase the coding complexity by adding less coding complexity. , improved prediction performance of linear predictive coding.
  • a linear predictive coding system comprising:
  • a linear predictive analysis device configured to acquire signal characteristic information of at least one sample of the input signal; compare and analyze the signal characteristic information to obtain an analysis result; and select a window function according to the analysis result to adaptively window the input signal to obtain a windowed window Signal; processing the windowed signal to obtain a linear predictive coding coefficient;
  • An encoding device for compiling the linear predictive coding coefficients obtained by the linear predictive analysis device code.
  • the linear predictive coding system provided by the embodiment of the present invention can firstly analyze the input signal to obtain a result, and adaptively allocate a window function required for windowing according to the analysis result, thereby obtaining a linear predictive coding coefficient; and then according to the linear
  • the prediction coding coefficients are encoded. Therefore, the prediction performance of the linear predictive coding can be improved with less coding complexity.
  • FIG. 1 is a flow chart of a linear prediction analysis method according to an embodiment of the present invention.
  • Embodiment 2 is a flow chart of a linear prediction analysis method according to Embodiment 1 of the present invention.
  • Embodiment 3 is a flow chart of a linear prediction analysis method according to Embodiment 2 of the present invention.
  • FIG. 5 is a flow chart of a linear prediction analysis method according to Embodiment 4 of the present invention.
  • FIG. 6 is a flow chart of a linear prediction analysis method according to Embodiment 5 of the present invention.
  • FIG. 7 is a flow chart of a linear prediction analysis method according to Embodiment 6 of the present invention.
  • Embodiment 8 is a flow chart of a linear prediction analysis method according to Embodiment 7 of the present invention.
  • FIG. 9 is a structural block diagram of a linear prediction analysis apparatus according to an embodiment of the present invention.
  • FIG. 10 is a structural block diagram of a linear predictive analysis apparatus according to another embodiment of the present invention
  • FIG. 11 is a structural block diagram of a linear predictive coding system according to an embodiment of the present invention
  • FIG. 12 is a structural block diagram of a linear predictive coding system according to another embodiment of the present invention.
  • Embodiments of the present invention provide a linear prediction analysis method, apparatus, and system, which can improve linear prediction performance and reduce analysis operation complexity.
  • the linear prediction analysis method provided by the embodiment of the present invention, as shown in FIG. 1, the steps of the method include:
  • the linear prediction analysis method obtained by the embodiment of the present invention obtains the result by analyzing the input signal, and adaptively allocates the window function required for windowing according to the analysis result, thereby being able to increase the coding complexity. , improved prediction performance of linear predictive coding.
  • the signal characteristic information includes any one or any of a plurality of amplitudes, energies, zero-crossing rates, signal types, frame lengths, and encoding modes.
  • Embodiment 1 is a diagrammatic representation of Embodiment 1:
  • the linear prediction analysis method provided in the first embodiment of the present invention is as shown in FIG. 2, and the method steps include:
  • the input signal here refers to the signal input for LPC analysis, which may be a frame
  • the signal may also be a signal of one frame plus a segment of the history buffer (such as L samples of the history buffer, L may use different positive integers according to different codecs, such as 40, 80, 160 , 240, 320, etc.);
  • the first 4 points of the window function are set to:
  • Another example is when the number of input samples is 80:
  • the first 8 points of the window function are set to:
  • S203 Process the windowed signal to obtain a linear predictive coding coefficient for linear prediction.
  • the linear predictive analysis method provided by the first embodiment of the present invention obtains the amplitude of the first sample and the last sample of the input signal, and adaptively windowing the input signal according to the amplitude of the sample, thereby enabling The prediction performance of linear predictive coding is improved with less coding complexity.
  • Embodiment 2 is a diagrammatic representation of Embodiment 1:
  • the linear prediction analysis method provided by the second embodiment of the present invention is as shown in FIG. 3, and the method steps include:
  • the input signal here refers to the signal input to the LPC analysis, which may be a frame signal, or a signal of a frame plus a history buffer (such as L samples of the history buffer, L can use different positive integers according to different codecs, such as 40, 80, etc.);
  • the window functions ⁇ ] and ⁇ 2[ ] can be selected according to different speech and audio encoders through a large number of experiments, respectively, for different signals, such as a sine window, v2[] is a Hamming window, or a Hamming window. W2[ ] is a sine window.
  • S303 Process the windowed signal to obtain a linear predictive coding coefficient, and use it for linear prediction.
  • the linear prediction analysis method provided by the second embodiment of the present invention obtains the amplitude of the first sample of the input signal, and adaptively windowing the input signal according to the amplitude of the sample, thereby being able to increase less In the case of coding complexity, the prediction performance of linear predictive coding is improved.
  • Embodiment 3 is a diagrammatic representation of Embodiment 3
  • Nl is the input signal
  • N is the number of samples of the input signal
  • the input signal is here In the input signal for LPC analysis, it may be a frame signal, or it may be a frame signal plus a segment of the history buffer (such as L samples of the history buffer, L can be based on different codecs) Use different positive integers, such as 40, 80, etc.);
  • the window functions ⁇ ] and ⁇ 2[] can be selected according to different speech and audio encoders through a large number of experiments, respectively, for different signals, such as ⁇ ] is a sine window, v2[] is a Hamming window; or ⁇ ] is Hamming window, w2[] is a sine window.
  • S403. Process the windowed signal to obtain a linear predictive coding coefficient for linear prediction.
  • the linear prediction analysis method provided by the third embodiment of the present invention is obtained before (or after) the input signal is obtained.
  • Embodiment 4 is a diagrammatic representation of Embodiment 4:
  • the linear prediction analysis method provided in Embodiment 4 of the present invention is as shown in FIG. 5, and the method steps include:
  • the input signal here refers to the signal input to the LPC analysis, which may be a frame signal, or a signal of one frame plus a history buffer (such as L samples of the history buffer, L may be different according to The codec uses different positive integers, such as 40, 80, etc.);
  • the window functions ⁇ ] and ⁇ 2[ ] can be selected according to different speech and audio encoders through a large number of experiments, respectively, for different signals, such as ⁇ ] is a sine window, v2[ ] is a Hamming window; or ⁇ ] is Hamming window, w2[i] is a sine window.
  • S503 Processing the windowed signal to obtain a linear predictive coding coefficient, which is used for linear prediction.
  • the linear prediction analysis method provided by Embodiment 4 of the present invention is obtained before (or after) acquiring an input signal.
  • the average energy of the M samples, and adaptively windowing the input signal according to the average energy therefore, the prediction performance of the linear predictive coding can be improved with less coding complexity.
  • Embodiment 5 is a diagrammatic representation of Embodiment 5:
  • the input signal here refers to the signal input for LPC analysis, possibly Is a frame signal, or a frame signal plus a segment of the history buffer (such as L samples of the history buffer, L can use different positive integers according to different codecs, such as 40, 80, etc.);
  • the window function and can be selected according to different speech and audio encoders through a large number of experiments, respectively, for different signals, such as a sine window, w2 [''] is a Hamming window; or a Hamming window, w 2 . ''] is a sine window.
  • the linear prediction analysis method provided in the fifth embodiment of the present invention can obtain the zero-crossing rate of the input signal and adaptively window the input signal according to the zero-crossing rate, thereby being able to increase the coding complexity. , improved prediction performance of linear predictive coding.
  • the linear prediction analysis method provided in Embodiment 6 of the present invention is as shown in FIG. 7, and the method steps include:
  • the number of samples into the signal, ® is the AND operation.
  • the input signal here refers to the input for LPC
  • the analyzed signal may be a frame signal, or it may be a frame signal plus a segment of the history buffer (such as L samples of the history buffer, L may be different according to different codecs) Integer, such as 40, 80, etc.);
  • the window functions ⁇ ] and ⁇ 2[ ] can be selected according to different speech and audio encoders through a large number of experiments, respectively, for different signals, such as a sine window, v2[] is a Hamming window, or a Hamming window. W2[ ] is a sine window.
  • S703 Processing the windowed signal to obtain a linear predictive coding coefficient for linear prediction.
  • the linear prediction analysis method provided in Embodiment 6 of the present invention acquires the zero-crossing rate of the input signal and the average energy of the front (or after) A samples, and adaptively adds the input signal according to the zero-crossing rate and the average energy.
  • the window therefore, can improve the predictive performance of linear predictive coding with less coding complexity added.
  • the linear prediction analysis method provided in the seventh embodiment of the present invention is as shown in FIG. 8, and the method steps include:
  • the window functions ⁇ ] and v2[ ] can be selected according to different speech and audio encoders through a large number of experiments, respectively, for different signals, such as a sine window, v2[] is a Hamming window, or a Hamming window, 2 [ ] is a sine window;
  • S803 Processing the windowed signal to obtain a linear predictive coding coefficient, which is used for linear prediction.
  • the linear prediction analysis method provided in Embodiment 7 of the present invention acquires an input signal coding mode, converts an input signal, and obtains a PCM signal, and adaptively windowing the input signal according to the signal coding mode, thereby being able to increase less In the case of coding complexity, the prediction performance of linear predictive coding is improved.
  • the linear predictive analysis device provided by the embodiment of the present invention, as shown in FIG. 9, includes:
  • the obtaining unit 901 is configured to acquire signal characteristic information of at least one sample of the input signal; the analyzing unit 902 is configured to perform comparative analysis on the signal characteristic information to obtain an analysis result; and the windowing unit 903 is configured to select a window function pair according to the analysis result.
  • the input signal is adaptively windowed to obtain a windowed signal;
  • the processing unit 904 is configured to process the windowed signal to obtain a linear predictive coding coefficient, which is used for linear prediction.
  • the linear predictive analysis device obtained by the embodiment of the present invention obtains a result by analyzing an input signal, and adaptively allocates a window function required for windowing according to the analysis result, thereby being able to increase the coding complexity by adding less coding complexity. , improved prediction performance of linear predictive coding.
  • the analyzing unit 902 includes: a calculating module 902A, configured to calculate a value of the signal characteristic information acquired by the acquiring unit 901, where the value of the signal characteristic information includes a certain sample point. The value of the signal characteristic information and/or the average of the values of the signal characteristic information of a plurality of samples;
  • the determining module 902B is configured to determine whether the value of the signal characteristic information obtained by the calculating module 902A is greater than or equal to a certain threshold; or for determining the signal type and/or encoding mode of the input signal acquired by the acquiring unit 901. Further, in the analyzing unit 902, the method further includes:
  • the conversion module 902C is configured to convert the input signal acquired by the obtaining unit 901 into a pulse code modulation signal.
  • the linear predictive analysis device obtained by the embodiment of the present invention obtains a result by analyzing an input signal, and adaptively allocates a window function required for windowing according to the analysis result, thereby being able to increase the coding complexity by adding less coding complexity. , improved prediction performance of linear predictive coding.
  • the linear predictive coding system provided by the embodiment of the present invention, as shown in FIG. 11, includes:
  • the linear predictive analysis device 1101 is configured to acquire signal characteristic information of at least one sample of the input signal; compare and analyze the signal characteristic information to obtain an analysis result; and select a window function according to the analysis result to adaptively window the input signal to obtain windowing. After signal; processing the windowed signal to obtain a linear predictive coding coefficient;
  • the encoding device 1102 is configured to encode the linear predictive coding coefficients obtained by the linear predictive analysis device 1101.
  • the linear predictive coding system provided by the embodiment of the present invention can firstly analyze the input signal to obtain a result, and adaptively allocate a window function required for windowing according to the analysis result, thereby obtaining a linear predictive coding coefficient; and then according to the linear
  • the prediction coding coefficients are encoded. Therefore, the prediction performance of the linear predictive coding can be improved with less coding complexity.
  • the linear predictive analysis device 1101 has the same configuration as the linear predictive analysis device in the above embodiment, and details are not described herein again.
  • the linear predictive coding system provided by the embodiment of the present invention can firstly analyze the input signal to obtain a result, and adaptively allocate a window function required for windowing according to the analysis result, thereby obtaining a linear predictive coding coefficient; and then according to the linear
  • the prediction coding coefficients are encoded. Therefore, the prediction performance of the linear predictive coding can be improved with less coding complexity.
  • the storage medium may be a magnetic disk, an optical disk, or a read-only storage memory (Read-Only) Memory, ROM) or random access memory, random access memory (RAM), etc.
  • the above is only a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto, any familiar with this A person skilled in the art can easily conceive changes or substitutions within the scope of the invention as disclosed in the technical scope of the present invention. Therefore, the scope of protection of the present invention should be quasi.

Landscapes

  • Engineering & Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • Health & Medical Sciences (AREA)
  • Signal Processing (AREA)
  • Audiology, Speech & Language Pathology (AREA)
  • Human Computer Interaction (AREA)
  • Computational Linguistics (AREA)
  • Acoustics & Sound (AREA)
  • Multimedia (AREA)
  • Spectroscopy & Molecular Physics (AREA)
  • Theoretical Computer Science (AREA)
  • Compression, Expansion, Code Conversion, And Decoders (AREA)
  • Compression Or Coding Systems Of Tv Signals (AREA)

Description

一种线性预测分析方法、 装置及*** 技术领域
本发明涉及通信领域, 尤其涉及一种线性预测分析方法、 装置及***。 背景技术
为节省语音与音频信号传输和存储的带宽, 相应的语音与音频编码技术 得到了广泛的应用, 目前主要分为有损编码和无损编码, 有损编码的重建信 号与原始信号并不能保持完全一致, 但可以根据声源特点及人的感知特点最 大程度上减少信号的冗余信息。 无损编码则必须保证重建信号与原始信号完 全一致, 可以使得最后的解码质量没有任何损伤, 一般来讲有损编码压缩率 比较高, 但重建语音质量没有保证, 无损编码可以保证语音质量, 但压缩率 较低, 大约 50%左右。
无论在有损编码或无损编码中, 线性预测编码 ( LPC, Linear Prediction Coding )模型被广泛地应用在语音编码领域中, 在有损编码中码激励线性预 测编码模型是其典型应用的成功。 基本原理为: 先利用短时线性预测去除语 音信号的近样点冗余度, 再用长时预测器去除语音信号的远样点冗余度, 最 后对预测过程中产生的参数以及经过两级预 'J得到的残差信号进行编码传 输。
目前大多数有损和无损音频编解码的线性预测分析一般包括加窗、 求自 相关和 Levinson算法求解三个模块, 通过线性预测来得到残差信号, 再用熵 编码对残差信号进行编码来实现音频压缩。
在实现上述线性预测编码的过程中, 发明人发现现有技术中至少存在以 下问题:
加窗时釆用固定窗函数, 会使得线性预测性能达不到最优;
或者, 对输入信号分别进行两次线性预测分析, 一次给信号加短窗, 另 一次给信号加长窗, 会因为对输入信号进行了两次线性预测分析, 使得线性 预测分析的复杂度较大。 发明内容
本发明的实施例提供一种线性预测分析方法、 装置及***, 能够提高线 性预测性能, 降低分析运算复杂度。
一种线性预测分析方法, 包括:
获取输入信号至少一个样点的信号特性信息;
对信号特性信息进行比较分析, 得到分析结果;
根据分析结果选择窗函数对输入信号进行自适应加窗, 得到加窗后信号; 对加窗后信号进行处理, 得到线性预测编码系数用于线性预测。
本发明实施例提供的线性预测分析方法, 通过对输入信号进行分析, 得 到结果, 并根据分析结果自适应分配加窗所需的窗函数, 因此, 能够在增加 较少的编码复杂度的情况下, 提高了线性预测编码的预测性能。
一种线性预测分析装置, 包括:
获取单元, 用于获取输入信号至少一个样点的信号特性信息;
分析单元, 用于对信号特性信息进行比较分析, 得到分析结果; 加窗单元, 用于根据分析结果选择窗函数对输入信号进行自适应加窗, 得到加窗后信号;
处理单元, 用于对加窗后信号进行处理, 得到线性预测编码系数, 用于 线性预测。
本发明实施例提供的线性预测分析装置, 通过对输入信号进行分析, 得 到结果, 并根据分析结果自适应分配加窗所需的窗函数, 因此, 能够在增加 较少的编码复杂度的情况下, 提高了线性预测编码的预测性能。
一种线性预测编码***, 包括:
线性预测分析装置, 用于获取输入信号至少一个样点的信号特性信息; 对信号特性信息进行比较分析, 得到分析结果; 根据分析结果选择窗函数对 输入信号进行自适应加窗, 得到加窗后信号; 对加窗后信号进行处理, 得到 线性预测编码系数;
编码装置, 用于根据线性预测分析装置得到的线性预测编码系数进行编 码。
本发明实施例提供的线性预测编码***, 能够先通过对输入信号进行分 析, 得到结果, 并根据分析结果自适应分配加窗所需的窗函数, 进而得到线 性预测编码系数; 然后再根据该线性预测编码系数进行编码。 因此, 能够在 增加较少的编码复杂度的情况下, 提高了线性预测编码的预测性能。
附图说明 施例或现有技术描述中所需要使用的附图作简单地介绍, 显而易见地, 下面 描述中的附图仅仅是本发明的一些实施例, 对于本领域普通技术人员来讲, 在不付出创造性劳动性的前提下, 还可以根据这些附图获得其他的附图。
图 1为本发明实施例提供的线性预测分析方法的流程框图;
图 2为本发明实施例一提供的线性预测分析方法的流程框图;
图 3为本发明实施例二提供的线性预测分析方法的流程框图;
图 4为本发明实施例三提供的线性预测分析方法的流程框图;
图 5为本发明实施例四提供的线性预测分析方法的流程框图;
图 6为本发明实施例五提供的线性预测分析方法的流程框图;
图 7为本发明实施例六提供的线性预测分析方法的流程框图;
图 8为本发明实施例七提供的线性预测分析方法的流程框图;
图 9为本发明实施例提供的线性预测分析装置的结构框图;
图 10为本发明另一实施例提供的线性预测分析装置的结构框图; 图 11为本发明实施例提供的线性预测编码***的构造框图;
图 12为本发明另一实施例提供的线性预测编码***的构造框图。
具体实施方式
下面将结合本发明实施例中的附图, 对本发明实施例中的技术方案进行 清楚、 完整地描述, 显然, 所描述的实施例仅仅是本发明一部分实施例, 而 不是全部的实施例。 基于本发明中的实施例, 本领域普通技术人员在没有作 出创造性劳动前提下所获得的所有其他实施例 , 都属于本发明保护的范围。 本发明的实施例提供一种线性预测分析方法、 装置及***, 能够提高线 性预测性能, 降低分析运算复杂度。
下面结合附图对本发明实施例进行详细描述。
本发明实施例提供的线性预测分析方法, 如图 1 所示, 该方法的步骤包 括:
5101、 获取输入信号至少一个样点的信号特性信息;
5102、 对信号特性信息进行比较分析, 得到分析结果;
5103、 根据分析结果选择窗函数对输入信号进行自适应加窗, 得到加窗 后信号;
5104、 对加窗后信号进行处理, 得到线性预测编码系数用于线性预测。 本发明实施例提供的线性预测分析方法, 通过对输入信号进行分析, 得 到结果, 并根据分析结果自适应分配加窗所需的窗函数, 因此, 能够在增加 较少的编码复杂度的情况下, 提高了线性预测编码的预测性能。
其中, 信号特性信息包括幅值、 能量、 过零率、 信号类型、 帧长、 编码 方式中的任意一个或任意多个。
下面通过具体实施例加以说明。
实施例一:
本发明实施例一提供的线性预测分析方法, 如图 2 所示, 该方法步骤包 括:
S201、 获取输入信号的第一个样点的幅值 |x[0]|和最后一个样点的幅值 \x[N - 1]|,其中 , φ·] , i = 0,1,..., N - 1为输入信号, N为输入信号的样点个数(如 40、 80、 160、 240、 320等); 输入信号在这里是指输入进行 LPC分析的信号, 可 能是一帧信号, 也可能是一帧信号加上历史緩冲区的一段信号 (如历史緩冲 区的 L个样点, L可根据不同的编解码器釆用不同的正整数,如 40、 80、 160、 240、 320等);
S202, 对样点幅值 |x[0]|和 |x[N- l]|进行分析, 并根据分析结果对输入信号 进行自适应加窗: 如当输入样点数为 40时:
如果输入信号的第一个样点的幅值 |x[0]|小于某个预先设定阈值 thr (如 thr = ),则对窗函数的最前面 4个点设置为:
w{n) = 0.23 + 0.77 · cos(2 · ;τ · (31— 8 · ")/ 127) , η = 0,123
否则对窗函数的最前面 4个点设置为:
w(n) = 0.26 + 0.74 - cos(2 · · (31 - 8 · / 127), n = 0,1,2,3 对窗函数第 5至第 36个点都设为 1, 即:
w(n) = 1 " = 4,...,35 如果输入信号的最后一个样点的幅值 IX[39]I小于某个预先设定阈值 thr (如 t//r = l28 ),则对窗函数的最后面 4个点设置为:
w(n) = 0.23 + 0.77 · cos(2 -π-(8-η- 281)/ 127) η = 36,37,38,39 否则对窗函数的最后面 4个点设置为:
w(n) = 0.26 + 0.74 · cos(2 · ·(81281) / 127) n = 36,37,38,39 然后用上述自适应设置后的窗函数 "), " = 1,2,···,3839对信号 (")," = 1,2,...,38,39进行加窗, 即
xd[n] - x[n] - wn] n - 0,1,...,38,39 得到自适应加窗后的信号 xt["]], « = 0,1,..,38,39
又如当输入样点数为 80时:
如果输入信号的第一个样点的幅值 小于某个预先设定阈值 (如 thr = 128 ),则对窗函数的最前面 8个点设置为:
w{n) = 0.26 + 0.74 · cos(2■ π -{3\- -n)l\21) n = 0,1,2,...,7 否则对窗函数的最前面 8个点设置为:
w{n) = 0.16 + 0.84· cos(2 · ;τ · (31 - 4 · ") / 127), η = 0,1,2,...,7 对窗函数第 9至第 72个点都设为 1, 即:
) = 1, " = 8,...,71 如果输入信号的最后一个样点的幅值 ΙΧ[79]Ι小于某个预先设定阈值 thr (如 t//r = l28 ),则对窗函数的最后面 8个点设置为: w(n) = 0.26 + 0.74 · cos(2 -π-(4-η- 285)/127) n = 72,73,74,...,79 否则对窗函数的最后面 8个点设置为:
w(n) = 0.16 + 0.84 · cos(2 -π-(4-η- 285)/127) η = 72,73,74,...,79
然后用上述自适应设置后的窗函数 , «-0,1,...,78,79对信号 (")," = 0,1,...,78,79进行加窗, 即
xd[n] = x[n] - w[n] n = 0,1,...,78,79 得到自适应加窗后的信号 xt["]], " = 0, .,78,79
窗函数 的调整策略可根据不同的语音频编码器通过大量实验来选定, 分别适用于不同的信号。阈值 也是通过大量实验选定,如 thr = 128或 thr = 157 等;
S203、 对加窗后信号进行处理, 得到线性预测编码系数, 用于线性预测。 本发明实施例一提供的线性预测分析方法, 通过获取输入信号的第一个 样点和最后一个样点的幅值, 并根据该样点幅值对输入信号进行自适应加窗, 因此, 能够在增加较少的编码复杂度的情况下, 提高了线性预测编码的预测 性能。
实施例二:
本发明实施例二提供的线性预测分析方法, 如图 3 所示, 该方法步骤包 括:
S301、获取输入信号的第一个样点的幅值 |x[0]|,其中 , φ·] , i = 0,1,..., N - 1为输 入信号, N为输入信号的样点个数; 输入信号在这里是指输入进行 LPC分析 的信号, 可能是一帧信号,也可能是一帧信号加上历史緩冲区的一段信号(如 历史緩冲区的 L个样点, L可根据不同的编解码器釆用不同的正整数, 如 40、 80等);
S302、 对样点幅值 |x[0]|进行分析, 并根据分析结果对输入信号进行自适 应力口窗:
如果输入信号的第一个样点的幅值 |x[o]|大于(或大于等于)某个预先设定 阈 值 ^ , 用 第 一 窗 函 数 对 输入信 号 进行 加 窗 , 即 令 xd[i] = x[i] - w\[i] , i = 0,1,..., N - 1 ,其中 χφ·]为加窗后的信号, φ·]为第一窗函数; 否 则 , 则 用 第 二 窗 函 数对输入信号 进行加 窗 , 即令 xt[ ] = x[ ].w2[ ], = 0,l,...,N-l,其中 w2[ ]为第二窗函数;
窗函数 φ·]和 ν2[ ]可根据不同的语音频编码器通过大量实验来选定,分别 适用于不同的信号, 例如 为正弦窗, v2[ ]为汉明窗; 或者 为海明窗, w2[ ]为正弦窗。 阈值 t//r也是通过大量实验选定, 如 t//r = 128或 t/^ = 157 ;
在一个具体的实现中, thr = , 当帧长 N = 80时, 9 79
Figure imgf000009_0001
当帧长 N = 40时,
0.26 + 0.74 · cos(2 -π-(3\-8-ϊ) / 127), i = 0,1,...,3
wl[i] = 1, = 4,5,...,35
0.26 + 0.74 · cos(2 -; τ · (8 · i— 281)7127), i = 36,37,...,39 ,...,3 7,...,39
Figure imgf000009_0002
S303、 对加窗后信号进行处理, 得到线性预测编码系数, 用于线性预测。 本发明实施例二提供的线性预测分析方法, 通过获取输入信号的第一个 样点的幅值, 并根据该样点幅值对输入信号进行自适应加窗, 因此, 能够在 增加较少的编码复杂度的情况下, 提高了线性预测编码的预测性能。
实施例三:
本发明实施例三提供的线性预测分析方法, 如图 4所示, 该方法步骤包 括:
1
S401、 获取输入信号的前(或后) A 个样点的幅值平均值 =丄.
M
, x[ ], = 0,l,...,N-l为输入信号, N为输入信号的样点个数; 输入信号在这 里是指输入进行 LPC分析的信号, 可能是一帧信号, 也可能是一帧信号加上 历史緩冲区的一段信号 (如历史緩冲区的 L个样点, L可根据不同的编解码 器釆用不同的正整数, 如 40、 80等);
S402、 对前(或后) M个样点的幅值平均值 进行分析, 并根据分析结 果对输入信号进行自适应加窗:
如果前(或后) A 个样点的幅值平均值 大于(或大于等于)某个预先设 定 阈 值 ^ ,用 第 一 窗 函数对输入信 号 进行加 窗 , 即令 xd[i] = x[i] - wl[i] , i = 0,1,..., N - 1 ,其中 χφ·]为加窗后的信号, φ·]为第一窗函数; 否 则 , 则 用 第 二 窗 函 数对输入信号 进行加 窗 , 即令 xt[] = x[].w2[], = 0,l,...,N-l,其中 w2[]为第二窗函数;
窗函数 φ·]和 ν2[]可根据不同的语音频编码器通过大量实验来选定,分别 适用于不同的信号, 例如 κφ]为正弦窗, v2[]为汉明窗; 或者 κφ]为海明窗, w2[]为正弦窗。 阈值 t//r也是通过大量实验选定, 如 thr = thr = 2
在一个具体的实现中, r = 128, 当帧长 N = 80时, 9 79
Figure imgf000010_0001
当帧长 N = 40时,
,...,3 7,...,39
,...,3 7,...,39
Figure imgf000010_0002
S403、 对加窗后信号进行处理, 得到线性预测编码系数, 用于线性预测。 本发明实施例三提供的线性预测分析方法, 通过获取输入信号前(或后) M个样点的幅值平均值, 并根据该平均值对输入信号进行自适应加窗, 因此, 能够在增加较少的编码复杂度的情况下, 提高了线性预测编码的预测性能。
实施例四:
本发明实施例四提供的线性预测分析方法, 如图 5 所示, 该方法步骤包 括:
5501、 获取输入信号的前(或后) M个样点的平均能量 丄 φ·]2,其 中, x[ ], = 0,l,...,N-l为输入信号, W为输入信号的样点个数。输入信号在这里 是指输入进行 LPC分析的信号, 可能是一帧信号, 也可能是一帧信号加上历 史緩冲区的一段信号 (如历史緩冲区的 L个样点, L可根据不同的编解码器 釆用不同的正整数, 如 40、 80等);
5502, 对前(或后) M个样点的平均能量 ^进行分析, 并根据分析结果对 输入信号进行自适应加窗:
如果前(或后) A 个样点的幅值平均值 ^大于(或大于等于)某个预先设 定 阈 值 ,用 第 一 窗 函数对输入信 号 进行加 窗 , 即令 xd[i] = x[i] - wl[i] , i = 0,1,..., N - 1 ,其中 χφ·]为加窗后的信号, φ·]为第一窗函数。
否 则 , 则 用 第 二 窗 函 数对输入信号 进行加 窗 , 即令 xt[ ] = x[ ].w2[ ], = 0,l,...,N-l,其中 w2[ ]为第二窗函数。
窗函数 φ·]和 ν2[ ]可根据不同的语音频编码器通过大量实验来选定,分别 适用于不同的信号, 例如 κφ]为正弦窗, v2[ ]为汉明窗; 或者 κφ]为海明窗, w2[i]为正弦窗。 阈值 thr也是通过大量实验选定, 如 thr = 1024或 thr = 2573。
在一个具体的实现中, r = 1280 , 当帧长 N = 80时, 9 79
Figure imgf000011_0001
当帧长 W = 40时,
0.25 + 0.75 · cos(2 · ·(31 - 8 · ζ·) / 127), i = 0,1,...,3
wl[i] = 1, ζ' = 4,5,...,35
0.25 + 0.75 · cos(2 · ·(8 · i— 281) / 127), i = 36,37,...,39 ,...,3 7,...,39
Figure imgf000012_0001
S503、 对加窗后信号进行处理, 得到线性预测编码系数, 用于线性预测。 本发明实施例四提供的线性预测分析方法, 通过获取输入信号前(或后)
M个样点的平均能量, 并根据该平均能量对输入信号进行自适应加窗, 因此, 能够在增加较少的编码复杂度的情况下, 提高了线性预测编码的预测性能。
实施例五:
本发明实施例五提供的线性预测分析方法, 如图 6所示, 该方法步骤包 括: S601、 获取输入信号的过零率
Figure imgf000012_0002
,其中, x[ ] , = G,l„,N_l为输入信号, N为输入信号的样点个数, ®为与 (AND )操 作。 输入信号在这里是指输入进行 LPC分析的信号, 可能是一帧信号, 也可 能是一帧信号加上历史緩冲区的一段信号 (如历史緩冲区的 L个样点, L可 根据不同的编解码器釆用不同的正整数, 如 40、 80等);
S602、 对过零率 ^进行分析, 并根据分析结果对输入信号进行自适应加 :
如果过零率 大于(或大于等于)某个预先设定阈值 thr , 用第一窗函数 对输入信号进行加窗, 即令^[''] = 4 ^1[ ] , = 0,1,..., - 1 ,其中 Φ·]为加窗后的 信号, 为第一窗函数;
否 则 , 则 用 第 二 窗 函 数对输入信号 进行加 窗 , 即令 xd[i] = x[i] · w2[i] , i = 0,1,..., N _ 1 ,其中 ν2[ ]为第二窗函数; 窗函数 和 可根据不同的语音频编码器通过大量实验来选定,分别 适用于不同的信号, 例如 为正弦窗, w2['']为汉明窗; 或者 为海明窗, w 2.['']为正弦窗。 阈值^ "也是通过大量实验选定, thr = 5^ hr = 23 ;
在一个具体的实现中, thr = l8 , 当帧长 N = 80时, 9 79
Figure imgf000013_0001
当帧长 N = 40时,
,..., 3 7,...,39
,...,3 7,...,39
Figure imgf000013_0002
S603、 对加窗后信号进行处理, 得到线性预测编码系数, 用于线性预测。 本发明实施例五提供的线性预测分析方法, 通过获取输入信号的过零率, 并根据该过零率对输入信号进行自适应加窗, 因此, 能够在增加较少的编码 复杂度的情况下, 提高了线性预测编码的预测性能。
实施例六:
本发明实施例六提供的线性预测分析方法, 如图 7 所示, 该方法步骤包 括:
S701、获取输入信号的过零率 = χ[( []≤ 0)®( [ -1] > 0)],和前(或后) Μ 个样点的平均能量 丄 ·! φ·]2 , 其中, x[ ], = 0,l,...,N-l为输入信号, N为库 ίT
;-0
入信号的样点个数, ®为与(AND)操作。输入信号在这里是指输入进行 LPC 分析的信号, 可能是一帧信号, 也可能是一帧信号加上历史緩冲区的一段信 号(如历史緩冲区的 L个样点, L可根据不同的编解码器釆用不同的正整数, 如 40、 80等);
S702, 对过零率 zc和前(或后) M个样点的平均能量 进行分析, 并根 据分析结果对输入信号进行自适应加窗:
如果过零率 大于(或大于等于)某个预先设定阈值 1或 ^小于等于某 个预先设定的阈值 r2 , 用第一窗函数对输入信号进行加窗, 即令 xd[i] = x[i]-wl[i],i = Ο,Ι,.,.,Ν -1 ,其中 χφ·]为加窗后的信号, φ·]为第一窗函数。
否 则 , 则 用 第 二 窗 函 数对输入信号 进行加 窗 , 即令 xd[i] = x[i] · w2[i] , i = 0,1,..., N - 1 ,其中 ν2[ ]为第二窗函数。
窗函数 φ·]和 ν2[ ]可根据不同的语音频编码器通过大量实验来选定,分别 适用于不同的信号, 例如 为正弦窗, v2[ ]为汉明窗; 或者 为海明窗, w2[ ]为正弦窗。 阈值 rl和 r2也是通过大量实验选定, 如 rl = 15、 thrl = 1023 或 tM = 23、 2 = 1012。
在一个具体的实现中, 1 = 17、 thrl = \0\2 , 当帧长 N = 80时, 79
Figure imgf000014_0001
0.26 + 0.74 · cos(2 -; τ · (31 _ 4 · ζ·)/127), i = 0,1,..., 7
w2[i] 1, ζ = 8,9,...,71
0.26 + 0.74 · cos(2 -; τ · (4 · i - 285) /127), i = 72,73,...,79 当帧长 N = 40时,
,...,3 7,...,39 ,...,3 7,...,39
Figure imgf000014_0002
S703、 对加窗后信号进行处理, 得到线性预测编码系数, 用于线性预测。 本发明实施例六提供的线性预测分析方法, 通过获取输入信号的过零率 和前(或后) A 个样点的平均能量, 并根据该过零率和平均能量对输入信号 进行自适应加窗, 因此, 能够在增加较少的编码复杂度的情况下, 提高了线 性预测编码的预测性能。
实施例七:
本发明实施例七提供的线性预测分析方法, 如图 8 所示, 该方法步骤包 括:
5801、 获取输入信号编码方式, 输入信号为 G.711信号, 可能为 A-law信 号, 也可能为 mu-law信号; 对输入信号进行转换, 得到 PCM信号;
5802、 对输入信号编码方式进行分析, 并根据分析结果对 PCM信号进行 自适应力口窗如:
如果编码方式为 A-law, 用第一窗函数对 PCM 信号进行加窗, 即令 xd[i] = x[i]-wl[i],i = Ο,Ι,.,.,Ν -1 ,其中 χφ·]为加窗后的信号, φ·]为第一窗函数, φ·]为 PCM信号。
否 则 , 则 用 第 二窗 函数对 PCM 信号进行加 窗 , 即令 xd[i] = x[i] · w2[i] , i = 0,1,..., N - 1 ,其中 ν2[ ]为第二窗函数。
窗函数 κφ]和 v2[ ]可根据不同的语音频编码器通过大量实验来选定,分别 适用于不同的信号, 例如 为正弦窗, v2[ ]为汉明窗; 或者 为海明窗, 2[ ]为正弦窗;
在一个具体的实现中,当编码方式为 A-law或 mu-law时,当帧长 N = 80时, 79
Figure imgf000015_0001
0.26 + 0.74 · cos(2 -; r · (31 _ 4 · i)/\21), i = 0,1,..., 7
w2[i] 1, z = 8,9,...,71
0.26 + 0.74 · cos(2 -π-(4 - 285) /127), i = 72,73,...,79 当帧长 N = 40时, ,...,3 7,...,39
,...,3 7,...,39
Figure imgf000016_0001
S803、 对加窗后信号进行处理, 得到线性预测编码系数, 用于线性预测。 本发明实施例七提供的线性预测分析方法, 获取输入信号编码方式, 并 对输入信号进行转换, 得到 PCM信号, 根据该信号编码方式对输入信号进行 自适应加窗, 因此, 能够在增加较少的编码复杂度的情况下, 提高了线性预 测编码的预测性能。
本发明实施例提供的线性预测分析装置, 如图 9所示, 包括:
获取单元 901 , 用于获取输入信号至少一个样点的信号特性信息; 分析单元 902, 用于对信号特性信息进行比较分析, 得到分析结果; 加窗单元 903 , 用于根据分析结果选择窗函数对输入信号进行自适应加 窗, 得到加窗后信号;
处理单元 904, 用于述加窗后信号进行处理, 得到线性预测编码系数, 用 于线性预测。
本发明实施例提供的线性预测分析装置, 通过对输入信号进行分析, 得 到结果, 并根据分析结果自适应分配加窗所需的窗函数, 因此, 能够在增加 较少的编码复杂度的情况下, 提高了线性预测编码的预测性能。
其中, 在本发明另一实施例中, 如图 10所示, 分析单元 902包括: 计算模块 902A, 用于计算获取单元 901获取的信号特性信息的值, 信号 特性信息的值包括某一个样点的信号特性信息的值和 /或某多个样点的信号特 性信息的值的平均值;
判断模块 902B, 用于判断计算模块 902A得出的信号特性信息的值是否 大于或大于等于某一阈值; 或者用于判断获取单元 901 获取的输入信号的信 号类型和 /或编码方式。 进一步地, 在上述分析单元 902中, 还包括:
转换模块 902C, 用于将获取单元 901获取的输入信号转换为脉冲编码调 制信号。
本发明实施例提供的线性预测分析装置, 通过对输入信号进行分析, 得 到结果, 并根据分析结果自适应分配加窗所需的窗函数, 因此, 能够在增加 较少的编码复杂度的情况下, 提高了线性预测编码的预测性能。
本发明实施例提供的线性预测编码***, 如图 11所示, 包括:
线性预测分析装置 1101 , 用于获取输入信号至少一个样点的信号特性信 息; 对信号特性信息进行比较分析, 得到分析结果; 根据分析结果选择窗函 数对输入信号进行自适应加窗, 得到加窗后信号; 对加窗后信号进行处理, 得到线性预测编码系数;
编码装置 1102,用于才艮据线性预测分析装置 1101得到的线性预测编码系 数进行编码。
本发明实施例提供的线性预测编码***, 能够先通过对输入信号进行分 析, 得到结果, 并根据分析结果自适应分配加窗所需的窗函数, 进而得到线 性预测编码系数; 然后再根据该线性预测编码系数进行编码。 因此, 能够在 增加较少的编码复杂度的情况下, 提高了线性预测编码的预测性能。
其中, 在本发明的另一实施例中, 如图 12所示, 线性预测分析装置 1101 与上述实施例中的线性预测分析装置构造相同, 在此就不再赘述。
本发明实施例提供的线性预测编码***, 能够先通过对输入信号进行分 析, 得到结果, 并根据分析结果自适应分配加窗所需的窗函数, 进而得到线 性预测编码系数; 然后再根据该线性预测编码系数进行编码。 因此, 能够在 增加较少的编码复杂度的情况下, 提高了线性预测编码的预测性能。
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流 程, 是可以通过计算机程序来指令相关的硬件来完成, 所述的程序可存储于 一计算机可读取存储介质中, 该程序在执行时, 可包括如上述各方法的实施 例的流程。其中,所述的存储介质可为磁碟、光盘、只读存储记忆体( Read-Only Memory, ROM )或随机存^ "i己忆体 ( Random Access Memory, RAM )等。 以上所述, 仅为本发明的具体实施方式, 但本发明的保护范围并不局限 于此, 任何熟悉本技术领域的技术人员在本发明揭露的技术范围内, 可轻易 想到变化或替换, 都应涵盖在本发明的保护范围之内。 因此, 本发明的保护 范围应所述以权利要求的保护范围为准。

Claims

权利 要 求 书
1、 一种线性预测分析方法, 其特征在于, 包括:
获取输入信号至少一个样点的信号特性信息;
对所述信号特性信息进行比较分析, 得到分析结果;
根据所述分析结果选择窗函数对所述输入信号进行自适应加窗, 得到加窗 后信号;
对所述加窗后信号进行处理, 得到线性预测编码系数用于线性预测。
2、 根据权利要求 1所述的线性预测分析方法, 其特征在于, 所述信号特性 信息包括:
幅值、 能量、 过零率、 帧长中的任意一个或任意多个;
或者包括: 输入信号的信号类型、 输入信号的编码方式中的任意一个或任 意多个。
3、 根据权利要求 2所述的线性预测分析方法, 其特征在于, 所述对所述信 号特性信息进行比较分析, 得到分析结果包括:
计算获取的信号特性信息的值, 并判断所述信号特性信息的值是否大于或 大于等于某一阈值, 其中, 所述信号特性信息的值包括某一个或多个样点的信 号特性信息的值和 /或某多个样点的信号特性信息的值的平均值;
或者, 判断所述输入信号的信号类型和 /或编码方式。
4、 根据权利要求 3所述的线性预测分析方法, 其特征在于, 所述根据所述 分析结果选择窗函数对所述输入信号进行自适应加窗, 得到加窗后信号包括: 当确定所述信号特性信息的值大于或大于等于某一阈值时, 用第一窗函数 对所述输入信号进行自适应加窗, 得到加窗后信号; 否则, 用第二窗函数对所 述输入信号进行自适应加窗, 得到加窗后信号;
或者, 当确定所述输入信号的类型和 /或编码方式时, 根据所述输入信号的 类型和 /或编码方式的不同, 选择第一窗函数或第二窗函数对所述输入信号进行 自适应加窗, 得到加窗后信号。
5、 根据权利要求 1所述的线性预测分析方法, 其特征在于, 获取输入信号 至少一个样点的信号特性信息包括:
将所述输入信号转换为脉冲编码调制信号;
获取转换后的所述输入信号至少一个样点的信号特性信息。
6、 一种线性预测分析装置, 其特征在于, 包括:
获取单元, 用于获取输入信号至少一个样点的信号特性信息;
分析单元, 用于对所述信号特性信息进行比较分析, 得到分析结果; 加窗单元, 用于根据所述分析结果选择窗函数对所述输入信号进行自适应 加窗, 得到加窗后信号;
处理单元, 用于对所述加窗后信号进行处理, 得到线性预测编码系数, 用 于线性预测。
7、 根据权利要求 6所述的线性预测分析装置, 其特征在于, 所述分析单元 包括:
计算模块, 用于计算所述获取单元获取的信号特性信息的值, 所述信号特 性信息的值包括某一个样点的信号特性信息的值和 /或某多个样点的信号特性信 息的值的平均值;
判断模块, 用于判断所述计算模块得出的所述信号特性信息的值是否大于 或大于等于某一阈值; 或者用于判断所述获取单元获取的输入信号的信号类型 和 /或编码方式。
8、 根据权利要求 7所述的线性预测分析装置, 其特征在于, 所述分析单元 还包括:
转换模块, 用于将所述获取单元获取的输入信号转换为脉冲编码调制信号。
9、 根据权利要求 7所述的线性预测分析装置, 其特征在于, 所述加窗单元 包括:
选择模块, 用于根据所述判断模块判断出的结果, 选择窗函数;
加窗计算模块, 用于根据所述选择模块选择的窗函数对所述输入信号进行 自适应加窗, 得到加窗后信号。
10、 一种线性预测编码***, 其特征在于, 包括: 线性预测分析装置, 用于获取输入信号至少一个样点的信号特性信息; 对 所述信号特性信息进行比较分析, 得到分析结果; 根据所述分析结果选择窗函 数对所述输入信号进行自适应加窗, 得到加窗后信号; 对所述加窗后信号进行 处理, 得到线性预测编码系数;
编码装置, 用于根据所述线性预测分析装置得到的所述线性预测编码系数 进行编码。
11、 根据权利要求 10所述的线性预测编码***, 其特征在于, 所述线性预 测分析装置包括:
获取单元, 用于获取输入信号至少一个样点的信号特性信息;
分析单元, 用于对所述信号特性信息进行比较分析, 得到分析结果; 加窗单元, 用于根据所述分析结果选择窗函数对所述输入信号进行自适应 加窗, 得到加窗后信号;
处理单元, 用于对所述加窗后信号进行处理, 得到线性预测编码系数, 用于线 性预测。
PCT/CN2009/070729 2009-03-11 2009-03-11 一种线性预测分析方法、装置及*** WO2010102446A1 (zh)

Priority Applications (5)

Application Number Priority Date Filing Date Title
CN2009800001092A CN102067211B (zh) 2009-03-11 2009-03-11 一种线性预测分析方法、装置及***
PCT/CN2009/070729 WO2010102446A1 (zh) 2009-03-11 2009-03-11 一种线性预测分析方法、装置及***
KR1020117023175A KR101397512B1 (ko) 2009-03-11 2009-03-11 선형 예측 코딩 분석을 위한 방법, 장치 및 시스템
EP09841315.6A EP2407963B1 (en) 2009-03-11 2009-03-11 Linear prediction analysis method, apparatus and system
US13/228,965 US8812307B2 (en) 2009-03-11 2011-09-09 Method, apparatus and system for linear prediction coding analysis

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
PCT/CN2009/070729 WO2010102446A1 (zh) 2009-03-11 2009-03-11 一种线性预测分析方法、装置及***

Related Child Applications (1)

Application Number Title Priority Date Filing Date
US13/228,965 Continuation US8812307B2 (en) 2009-03-11 2011-09-09 Method, apparatus and system for linear prediction coding analysis

Publications (1)

Publication Number Publication Date
WO2010102446A1 true WO2010102446A1 (zh) 2010-09-16

Family

ID=42727779

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2009/070729 WO2010102446A1 (zh) 2009-03-11 2009-03-11 一种线性预测分析方法、装置及***

Country Status (5)

Country Link
US (1) US8812307B2 (zh)
EP (1) EP2407963B1 (zh)
KR (1) KR101397512B1 (zh)
CN (1) CN102067211B (zh)
WO (1) WO2010102446A1 (zh)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US8812307B2 (en) 2009-03-11 2014-08-19 Huawei Technologies Co., Ltd Method, apparatus and system for linear prediction coding analysis

Families Citing this family (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN105378836B (zh) * 2013-07-18 2019-03-29 日本电信电话株式会社 线性预测分析装置、方法、以及记录介质
CN103700386B (zh) * 2013-12-16 2017-09-29 联想(北京)有限公司 一种信息处理方法及电子设备
EP4095854A1 (en) * 2014-01-15 2022-11-30 Samsung Electronics Co., Ltd. Weight function determination device and method for quantizing linear prediction coding coefficient
US9928850B2 (en) * 2014-01-24 2018-03-27 Nippon Telegraph And Telephone Corporation Linear predictive analysis apparatus, method, program and recording medium

Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US5848391A (en) * 1996-07-11 1998-12-08 Fraunhofer-Gesellschaft Zur Forderung Der Angewandten Forschung E.V. Method subband of coding and decoding audio signals using variable length windows
CN1338096A (zh) * 1998-12-30 2002-02-27 诺基亚移动电话有限公司 用于分析-合成celp型语音编码的自适应窗
CN1387131A (zh) * 2001-05-18 2002-12-25 株式会社东芝 预测参数分析装置与预测参数的分析方法
US20040083096A1 (en) * 2002-10-29 2004-04-29 Chu Wai C. Method and apparatus for gradient-descent based window optimization for linear prediction analysis
CN1732530A (zh) * 2002-11-07 2006-02-08 三星电子株式会社 运动图像专家组音频编码方法及装置

Family Cites Families (17)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
DE3902948A1 (de) * 1989-02-01 1990-08-09 Telefunken Fernseh & Rundfunk Verfahren zur uebertragung eines signals
US5305421A (en) * 1991-08-28 1994-04-19 Itt Corporation Low bit rate speech coding system and compression
US5642464A (en) 1995-05-03 1997-06-24 Northern Telecom Limited Methods and apparatus for noise conditioning in digital speech compression systems using linear predictive coding
FR2742568B1 (fr) * 1995-12-15 1998-02-13 Catherine Quinquis Procede d'analyse par prediction lineaire d'un signal audiofrequence, et procedes de codage et de decodage d'un signal audiofrequence en comportant application
CA2246532A1 (en) * 1998-09-04 2000-03-04 Northern Telecom Limited Perceptual audio coding
US6978236B1 (en) * 1999-10-01 2005-12-20 Coding Technologies Ab Efficient spectral envelope coding using variable time/frequency resolution and time/frequency switching
EP1343143B1 (en) * 2000-12-14 2011-10-05 Sony Corporation Analysis-synthesis of audio signal
AUPS102902A0 (en) * 2002-03-13 2002-04-11 Hearworks Pty Ltd A method and system for reducing potentially harmful noise in a signal arranged to convey speech
AU2003208517A1 (en) * 2003-03-11 2004-09-30 Nokia Corporation Switching between coding schemes
US7148415B2 (en) * 2004-03-19 2006-12-12 Apple Computer, Inc. Method and apparatus for evaluating and correcting rhythm in audio data
EP1775718A4 (en) * 2004-07-22 2008-05-07 Fujitsu Ltd AUDIOCODING DEVICE AND AUDIOCODING METHOD
KR200421226Y1 (ko) 2006-04-18 2006-07-10 솔렌스(주) 기능성 조경 나무명패 설치장치
US7987089B2 (en) * 2006-07-31 2011-07-26 Qualcomm Incorporated Systems and methods for modifying a zero pad region of a windowed frame of an audio signal
US8630863B2 (en) * 2007-04-24 2014-01-14 Samsung Electronics Co., Ltd. Method and apparatus for encoding and decoding audio/speech signal
US9653088B2 (en) * 2007-06-13 2017-05-16 Qualcomm Incorporated Systems, methods, and apparatus for signal encoding using pitch-regularizing and non-pitch-regularizing coding
KR101441896B1 (ko) * 2008-01-29 2014-09-23 삼성전자주식회사 적응적 lpc 계수 보간을 이용한 오디오 신호의 부호화,복호화 방법 및 장치
EP2407963B1 (en) 2009-03-11 2015-05-13 Huawei Technologies Co., Ltd. Linear prediction analysis method, apparatus and system

Patent Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US5848391A (en) * 1996-07-11 1998-12-08 Fraunhofer-Gesellschaft Zur Forderung Der Angewandten Forschung E.V. Method subband of coding and decoding audio signals using variable length windows
CN1338096A (zh) * 1998-12-30 2002-02-27 诺基亚移动电话有限公司 用于分析-合成celp型语音编码的自适应窗
CN1387131A (zh) * 2001-05-18 2002-12-25 株式会社东芝 预测参数分析装置与预测参数的分析方法
US20040083096A1 (en) * 2002-10-29 2004-04-29 Chu Wai C. Method and apparatus for gradient-descent based window optimization for linear prediction analysis
CN1732530A (zh) * 2002-11-07 2006-02-08 三星电子株式会社 运动图像专家组音频编码方法及装置

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US8812307B2 (en) 2009-03-11 2014-08-19 Huawei Technologies Co., Ltd Method, apparatus and system for linear prediction coding analysis

Also Published As

Publication number Publication date
US20110320195A1 (en) 2011-12-29
EP2407963A1 (en) 2012-01-18
EP2407963B1 (en) 2015-05-13
EP2407963A4 (en) 2012-08-01
KR20110132435A (ko) 2011-12-07
US8812307B2 (en) 2014-08-19
CN102067211A (zh) 2011-05-18
KR101397512B1 (ko) 2014-05-22
CN102067211B (zh) 2013-04-17

Similar Documents

Publication Publication Date Title
KR101513184B1 (ko) 계층적 디코딩 구조에서의 디지털 오디오 신호의 송신 에러에 대한 은닉
CA2827000C (en) Apparatus and method for error concealment in low-delay unified speech and audio coding (usac)
US9275644B2 (en) Devices for redundant frame coding and decoding
CN102449689B (zh) 编码方法、编码装置、编码程序、以及它们的记录介质
WO2009092309A1 (zh) 一种量化噪声泄漏控制方法及装置
RU2004133032A (ru) Кодирование стереофонических сигналов
JP5456097B2 (ja) 符号化方法、装置及び機器、及び復号化方法
WO2010102446A1 (zh) 一种线性预测分析方法、装置及***
CN112489665B (zh) 语音处理方法、装置以及电子设备
US9325544B2 (en) Packet-loss concealment for a degraded frame using replacement data from a non-degraded frame
KR102480710B1 (ko) 다중 채널 오디오 신호 처리 방법, 장치 및 시스템
US8380495B2 (en) Transcoding method, transcoding device and communication apparatus used between discontinuous transmission
KR20120096541A (ko) 디지털 신호의 압축 또는 압축해제를 위한 방법, 시스템 및 장치
JP6510566B2 (ja) オーディオ信号の時間包絡線を処理するための方法および装置、ならびにエンコーダ
CN105047200A (zh) 一种基于fpga的flac硬件解码器及解码方法
US11176954B2 (en) Encoding and decoding of multichannel or stereo audio signals
CN102930871B (zh) 一种线性预测分析方法、装置及***
US11380343B2 (en) Systems and methods for processing high frequency audio signal
JP5320508B2 (ja) 符号化装置、復号装置、これらの方法、プログラム及び記録媒体
KR100349329B1 (ko) 엠펙-2 고품질 오디오 처리 알고리즘의 병렬 처리 방법
WO2020001569A1 (zh) 立体声信号的编码、解码方法、编码装置和解码装置
CN114613375A (zh) 一种针对音频信号的时域噪声整形方法及装置
KR20120069906A (ko) 묵음 구간 정보를 사용한 오디오 신호의 부호화/복호화 장치 및 방법
JP2001100797A (ja) 音声符号化復号装置

Legal Events

Date Code Title Description
WWE Wipo information: entry into national phase

Ref document number: 200980000109.2

Country of ref document: CN

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

Ref document number: 09841315

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

WWE Wipo information: entry into national phase

Ref document number: 2009841315

Country of ref document: EP

ENP Entry into the national phase

Ref document number: 20117023175

Country of ref document: KR

Kind code of ref document: A