Research And Realization Of Audio Processing System Based On C5545

Embedded audio devices have been integrated into all aspects of production and life.With the increasing number of application scenarios,the audio applications using microphones as input sensors are more or less disturbed by environmental noise,and the fidelity performance of the device,including voice recognition and wireless voice transmission as the main function,has decreased significantly.This paper is based on the full study of C5545 DSP and related voice processing algorithms,designs and implements a pre-audio processing system based on microphone array and C5545 DSP.The system simplifies the voice collection and quality improvement of audio applications.For the noise and echo problems in the field of acoustics,this paper studies the post-filter beamforming algorithm and the LMS adaptive echo cancellation algorithm,combines the two algorithms with the harmonic regenerative noise reduction(HRNR)algorithm,proposes an improved post-filtering algorithm,and finally implement it in the processing system.Simulation experiments and actual tests show that the algorithm enhances the ability of speech enhancement and echo cancellation,effectively eliminates clutter and background noise,and improves the quality of speech collected by the microphone.Based on the processing system,this paper also designs a cloud-based speech recognition system,which has a high level of recognition accuracy even in the noisy environments.Based on the studies of protocol named ASBP.RSL10 and the front-end processing system are combined to realize a wireless voice transmission system suitable for low power consumption and high-fidelity applications.The tests of the system show that high-fidelity and stable voice transmission of the system in noisy environments.The above two system solutions have basically met the development expectations,providing a new optional reference solution for the Internet of Things,especially the smart home field.