Voice recognition character string processing comparison method based on Pinyin

Abstract

The invention relates to a voice recognition character string processing comparison method based on Pinyin. For application of an existing voice recognition technology to certain special occasions ofperson name recognition, equipment name recognition and the like, errors are generated easily due to incorrect comparison. The method is "secondary processing" based on a general Chinese character recognition algorithm; and recognized Chinese character strings are converted into Pinyin strings, and then the Pinyin strings are compared with target Pinyin strings. The method comprises the followingsteps of 1, performing Pinyin coding: performing coding on all Chinese character Pinyin, wherein the coding is similar to coding of unicode; and enumerating all Chinese character Pinyin combinations;2, performing code conversion: converting the character strings, with coding modes of GBK, Unicode, UTF-8 and the like, for expressing Chinese characters converted into the Pinyin strings; and 3, performing polyphone processing: enumerating polyphones of all family names; performing special processing; and distributing the same Pinyin codes. According to the method, accurate recognition can be rapidly realized, so that misjudgment is avoided.

Description

technical field [0001] The invention relates to the field of digital electronic products, in particular to a phonetic-based phonetic recognition character string processing and comparison method. Background technique [0002] In general, speech recognition is a technology that converts input speech into text through feature recognition. In some special occasions, this general-purpose technology cannot meet people's needs, such as the recognition of people's names, device names, etc. Perhaps the voice recognition algorithm recognizes the word "Yu Guoquan" through the input voice. character string, and the real user may say "Yu Guoquan". Since the name itself has little contextual relevance, the comparison will be incorrect. Contents of the invention [0003] The object of the present invention is to provide a method for processing and comparing character strings for speech recognition based on Pinyin. In the case of small character sets that require "special nouns" to ide...

AI Technical Summary

Problems solved by technology

In some special occasions, this general-purpose technology cannot meet people's needs, such as the recognition of people's names, device names, etc. Perhaps the voice recognition algorithm recognizes the word "Yu Guoquan" through the input voice. character string, and the real user may say "Yu Guoquan", because the name itself has little contextual relevance, it will cause the comparison to be incorrect