基于WEB的智能语音信息展示系统

随着通信技术与互联网技术的快速发展，使用电话和网络进行语音交流的用户越来越多，生成的语音数据也越来越多。与这些语音数据相关的业务仍需要人工逐条播放收听来进行处理。数据量的增大使得这些工作人员依然逐条收听这些语音变得不再可能，如果在收听之前可以对这些语音进行过滤和检索，尽可能保证重要的语音不被遗漏，并能优先处理重要的数据。而仅仅通过时间等外部信息来对所有语音进行筛选，在一定程度，对上述问题有所改善，然而过滤方式不涉及语音内容导致过滤结果不够准确。因此，利用语音特征信息和内容信息对语音数据进行有效过滤，具有十分广泛的应用需求和发展前景。
本论文主要是根据上述需求，针对海量语音信息处理难题，通过话音检出、性别识别、语种识别、声纹识别、关键词识别六种引擎技术手段，实现话音高效检索过滤，并将其结果展示在系统上；通过连续话音写技术，实现语音到文本的自动转化，并实现语音和转写结果的同步展示。
首先对语音识别技术进行概述，论述该系统的市场应用需求进行论述，并进行系统的功能结构的需求分析，同时介绍技术以及相关的开发平台；
其次对该系统进行详尽的系统设计、功能设计、数据库设计；
最后摘录了部分根据需求以及前期的设计进行的详细流程设计及编写。

关键字：语音识别；信息处理；结果展示；转写

Abstract：
With the rapid development of communication technology and Internet technology, more and more users are using telephone and network for voice communication, and more and more voice data is generated. Associated with these voice and data business still need artificial to play to listen to for a detailed processing. The increase in the amount of data makes it impossible for these workers to listen to these sounds, and if they can be filtered and retrieved before they are listened to, as much as possible to ensure that important speech is not missed and that important data is prioritized. And just by just waiting for the outside information to filter all of the sounds, and to a certain extent, it's an improvement on the problem, but the filter doesn't involve the voice content and the filter is not accurate enough. So, using the voice signature information and the content information to effectively filter the voice data, it has a very wide application of demand and development prospects.
This thesis is mainly based on the above requirements, in view of the massive amounts of speech information processing problem, silver detected by words, gender identification, language recognition, voice print recognition and keywords to identify six engine technology, achieve efficient retrieval voice filtering, and results show it on the system; Through continuous voice writing technology, the automatic transformation of speech to text is realized, and the simultaneous display of voice and transfer results is realized.
Firstly, the speech recognition technology is summarized, the market application demand of the system is discussed, and the demand analysis of the functional structure of the system is carried out, as well as the technology and related development platform.
Secondly, detailed system design, function design and database design are carried out.
Finally, the code is compiled based on the requirements and the previous design.

Keywords: Speech recognition; information processing; the results show; the transcription

*若需了解更多与协助请咨询↓→[电脑QQ][手机QQ]【数据协助】