Transplantation Of Improved Denoising Algorithm On Digital Hearing Aids Based On The Android Platform

Although there is a great emphasis on the study of the relevant content of the hearing aid, the design of hearing aid on the operating system has not been widely studied. In the non operating system, the development of hearing aids not only caused the trans-platform algorithm is poor, but also cannot be interactive. In order to solve the problem of the construction of the system caused by the intelligent digital hearing aids,this disseration will study the improvement of the algorithm and the architecture of the software transplantation.In this disseration, we have a deep understanding of the single language development of the traditional way. This method is relatively simple, but it cannot get feedback in the operating system of cross software interface. Even mature algorithms have been developed still needed to continue to use different development language.Resulting in a long development cycle, very poor portability, for this kind of problem,the current solution is to adopt a hierarchical software architecture design.Although the layered software architecture solves the hardware control of the software hierarchy, but this way, it is still a complicated and time-consuming process.On this basis, this paper studies the mechanism and implementation of JNI, and realizes the system development structure of cross platform software design by using Java language system structure independent and JNI. This paper points out the shortcomings of the ETSI202050V1.1.5 denoise specification in the European Telecommunication Standardization Institute in the application of the hearing aid, and then improve it.According to the human ear hearing system, the algorithm of the gain adjustment factor based on MFCC feature extraction and a priori SNR is designed, and the adjustment mechanism of AGC is added in the process. Compared to the original algorithm, the accuracy of the algorithm is improved and the requirement of the hearing aids is satisfied.The process of transplantation at first analysis the algorithm in the Matlab platform,and compared with the original algorithm. Then the algorithm function is developed based on the PC platform and the characteristics of audio processing data stream using the C++ language. Finally, through the design software architecture, the C++development program is transplanted to the middle layer of Android, the upper layer is called by Java. The whole algorithm is designed included the hardware driver layer, the middleware interface and the upper interface function. Compared to the development process of the design and the traditional mode of transplantation, without copying program language, realizing the multi-language development cross application, make full use of the characteristics of the two kinds of language development, and improvethe development efficiency and algorithm processing speed. This cross platform approach is more suitable for the development of digital hearing aids.