Data processing: speech signal processing – linguistics – language – Speech signal processing – For storage or transmission
Reexamination Certificate
1998-10-14
2001-05-08
Dorvil, Richemond (Department: 2641)
Data processing: speech signal processing, linguistics, language
Speech signal processing
For storage or transmission
C704S228000, C704S222000
Reexamination Certificate
active
06230124
ABSTRACT:
BACKGROUND OF THE INVENTION
1. Field of the Invention
The present invention relates to a coding method and apparatus for dividing an input audio signal into predetermined coding units such as blocks and frames and carrying out a coding processing to each of the coding units, as well as a decoding method and apparatus for decoding the coded signal.
2. Description of the Prior Art
There are known various coding methods carrying out signal compression by utilizing statistical features and human hearing sense characteristic in the time region and frequency region of an audio signal (including a sound signal and an acoustic signal). As such a coding method, there is a so-called CELP (code excited linear prediction) coding such as a VSEL (vector sum excited linear prediction) coding method and a PSI-CELP (pitch synchronous innovation—CELP) coding method which are now in a spotlight as a low bit rate audio coding method.
In a waveform coding method such as the CELP coding method, a predetermined number of samples of an input audio signal are used as a coding unit such as a block r frame, so that each block or frame of audio time axis waveform is subjected to the analysis-by-synthesis method so as to carry out a closed loop search of a optimal vector for quantization of the waveform and output an index of the vector.
A coded bit obtained by the aforementioned audio coding method of a low bit rate such as 2 kbps or 4 kbps can be widely applied as an audio in general to communication, computers, broadcasting, and the like, irrespective of particular audio information contents. Accordingly, it should be strictly protected against an error caused in a transmission path.
If an error is repeatedly caused in a transmission path, audio decoding is accompanied by a long time of missing sound, deteriorating the audio quality.
SUMMARY OF THE INVENTION
It is therefore an object of the present invention to provide a coding method and apparatus and a decoding method and apparatus that are tough against the transmission path error, enabling to improve the quality lowering.
In order to achieve the aforementioned object, the coding method according to the present invention includes: an audio coding step for dividing on a time axis an input audio signal into predetermined coding units and coding each of the coding units so as to output a plurality of types of audio coded parameters; an error detection code creation step for selecting important bits for the hearing sense from the plurality of types of audio coded parameters from the audio coding step, and creating an error check code from the important bit group; and a convolution coding step for executing a convolution coding to the error check code and the important bit group from the error check code creation step. Consequently, it is possible to protect important bits for the hearing sense, from a transmission path error.
Here, the audio coding step includes: a short period prediction residue calculation step for obtaining a short period prediction residue of the input audio signal; a sinusoidal wave analysis coding step for executing a sinusoidal wave analysis coding to the short period prediction residue obtained; and a waveform coding step for coding the input audio signal by way of waveform coding.
Moreover, the error check code creation step selects as the important bits: some or all of the parameters created when the short period prediction residue is obtained in the short period prediction residue calculation step; some or all of the coded outputs of the sinusoidal wave analysis coding step; and some or all of coded outputs from the waveform coding step, and creates an error check code from the important bits.
It should be noted that the aforementioned parameters created to obtain a short period prediction residue by the short period prediction residue calculation step are linear spectrum pair parameters forming an outline of a frequency spectrum of the input audio signal; the coded outputs from the sinusoidal wave analysis coding step include a voiced/unvoiced decision parameter indicating whether the input audio signal is voiced or unvoiced, and a pitch parameter if the input audio signal is voiced, and a spectrum codebook index and gain index indicating a spectrum envelope of the short period prediction coding residue; and the coded output from the waveform coding step is a noise codebook index and gain index based on the short period prediction residue when the input signal is unvoiced.
Moreover, in order to achieve the aforementioned object, the coding apparatus according to the present invention includes: audio coding means for dividing on a time axis an input audio signal into predetermined coding units and coding each of the units so as to output a plurality of types of audio coded parameters; error check code creation means for selecting important bits for the hearing sense from the plurality of types of audio coded parameters from the audio coding means, and creating an error check code from the important bits; and convolution coding means for executing a convolution coding to the error check code and the important bits from the error check code creation means.
Moreover, the decoding method according to the present invention is for decoding a coded data obtained from an input audio signal which has been divided on a time axis into predetermined units which have been coded to obtain a plurality of types of audio coded parameters from which important bits for the hearing sense have been used to create an error check code, wherein the error check code and the important bits have been subjected to a convolution coding to obtain a coded output which has been adjoined with non-important bits for transmission, the decoding method comprising: a convolution decoding step for executing a convolution decoding to the convolution-coded output so as to obtain a decoded output from the important bits added with the error check code and the non-important bits; an error check step for checking a transmission error using the error check code added to the convolution-decoded output from the convolution decoding step; an output adjustment step for adjusting the convolution-decoded output according to a result of the error check in the error check step; and an audio decoding step for executing an audio decoding processing to the convolution-decoded output from the output adjustment step.
Here, if an error is detected in the error check step, the output adjustment step outputs an interpolated data obtained by an interpolation processing instead of the convolution-decoded output.
Moreover, the output adjustment step executes a bad frame masking processing to the convolution-decoded output according to a result of the error check.
Moreover, the decoding apparatus according to the present invention is for decoding a coded data obtained from an input audio signal which has been divided on a time axis into predetermined units which have been coded to obtain a plurality of types of audio coded parameters from which important bits for the hearing sense have been used to create an error check code, wherein the error check code and the important bits have been subjected to a convolution coding to obtain a coded output which has been adjoined with non-important bits for transmission, the decoding apparatus including: convolution decoding means for executing a convolution decoding to the convolution-coded output so as to obtain a decoded output from the important bits added with the error check code and the non-important bits; error check and output adjustment means for checking a transmission error using the error check code added to the convolution-decoded output from the convolution decoding means and adjusting the convolution-decoded output according to a result of the error check; and audio decoding means for executing an audio decoding processing to the convolution-decoded output from the error check and output adjustment means.
REFERENCES:
patent: 5073940 (1991-12-01), Zinser et al.
patent: 5255343 (1993-10-01), Su
patent: 5371853 (1994-12-01), Kao et al.
Armstrong Angela
Dorvil Richemond
Maioli Jay H.
Sony Corporation
LandOfFree
Coding method and apparatus, and decoding method and apparatus does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Coding method and apparatus, and decoding method and apparatus, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Coding method and apparatus, and decoding method and apparatus will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-2453779