| With the prevalence and further development of computer and communication technology,varied researches on speech encoding technology have been maturity at present,because the performance of computer hardware become more and more well,almost encoding technologies can be implemented easily on computer,but it has low desire for the encoding performance on many applied occasion,it just need to synthesize understandable speech,and encoding at the voice collecting time,so it ought to reduce the encoding algorithmic complexity to reach Real-time character, choosing computer hardware system with high performance is wasting resource,for the sake of economical efficiency,it need to choose a micro computer system which have fit performance and cost least—embedded system,and choose a suitable encoding algorithm in given embedded system environment, so it can synthesize continuous,understandable voice,and use the limited storage to load the most compressive speech data,make the economical performance,real time,storage capability to be optimization.For this purpose and sense,the dissertation on the basis of analyzing the practice requirement of speech collected system seriously,and deep realizing the popular speech encoding technology at present, choose the embedded system which based on the ARM processor to be hardware environment implementing the encoding algorithm,and give the holistic software designing scheme, introduce several kinds of classical encoding technology detailedly, analyze the performance and speciality of various encoding technology seriously, then select the LPC-10 encoder after counting the cost, base on the speciality of embedded system,aim at adapting to the implemented environment of hardware,make partly rebuilding optimization to the aigorithm,reach the applied purpose,with that, get the result throughing the emulational experiment, estimating and analyzing the data.Finally,it summarizes the full text and puts forward the prospects of the work in the future. |
PDF Full Text Request