Font Size: a A A

A Design Of Intelligent Terminal Of Voice Control Based On Arm9

Posted on:2011-06-30Degree:MasterType:Thesis
Country:ChinaCandidate:J H ZhouFull Text:PDF
GTID:2198330338486096Subject:Software engineering
Abstract/Summary:PDF Full Text Request
Cell phones, MP3 etc terminal equipment has been more and more popular to people's lives, the occurrence of this circumstance has offered great convenience for people's communication and enriched our life; the speech recognition technology is a hot topic of current research, which has applied to many industrial control field and has a role to play. If speech recognition technology is applied to the mobile terminal, we can inject new vitality to the terminal equipment whit more simplified operation and a large market prospect, as well provide the users whole new experience. The embedded system has the characteristics of short research process, high performance, To be a development platform ,it can play out the advantages of the above two.Based on the review and analysis on the background of the subject, we make a system implementation scheme of speech recognition +ARM9 + + LINUX, then begin to carry on the research, first we study the algorithm of Speech recognition. This paper discusses the method of sampling and window frame of speech signal ,detecting method of two threshold endpoint, as well as the detection method of two endpoint threshold. Hereby we further discusses the method of the abstraction of characteristics of speech signals and the key parameters, including the short-term energy and power, ZCR, the LPCC, MFCC in the cepstrum. Then, this paper puts these conclusions as the premise, focuses on the two big main algorithms of speech recognition: DTW, HMM. After the completion of the above work, we turn to build the software framework of the embedded system, including the bootloader, kernel, file system.The last part of this paper is the experiment. In order to verify the feasibility of the control system, in front of this part ,the speech program is verified, including the endpoint detection procedures and the DTW algorithm. Then, this paper designs a voice control MP3 with analysising of the control system structure and realization scheme. After that, we realize the scheme by the mechanism of signal and slot.
Keywords/Search Tags:Speech recognition, Embedded Linux, ARM9, DTW, QT/Embedded, drivers
PDF Full Text Request
Related items