METHOD AND DEVICE FOR USE IN VECTOR QUANTIZATION ENCODING/DECODING OF AUDIO SIGNAL

摘要

A method and device for use in vector quantization coding/decoding of an audio signal. The encoding method comprises: audio perception analyzing a transform domain spectrum of an audio signal and amplitude adjusting the transform domain spectrum of the audio signal on the basis of the analysis result to acquire a to-be-quantified weighted spectrum; organizing the to-be-quantified weighted spectrum to acquire multiple to-be-quantified vectors; and, quantization coding the multiple to-be-quantified vectors to acquire vector quantization coded data. The method and device refer to audio perception characteristics for amplitude adjustment of a transform domain signal, thus allowing for elimination of perception redundancy and increased coding efficiency. By means of signal characteristic analysis, the time-frequency plane of an audio is divided reasonably and is organized into to-be-quantified vectors. A time-frequency plane division and vector organization scheme allowing for maximum coding gain can be selected, thus facilitating highly efficient quantization coding of a signal.