| Embedded speech synthesis system is based on the audio driver development, combining the voice synthesis technology with embedded system to make the system speech synthesis pronounce. It has been widely used in automobile industry. Therefore, studying on this system has important significance.This paper is based on the analysis of the audio driver technology and theory of speech synthesis technology. Through study on the Freescale company i.MX6 Q processor which is used in speech processing technology research, completed to design implementation depend on i.MX6 Q and TLV320AIC3106 which respectively based on audio driver and the speech synthesis engine of iflytek Aisound5.0 speech synthesis system Finally the speech synthesis system is tested and the testing result is analyzed. This paper is based on i.MX6 Q quad-core processors as the hardware environment to develop audio driver. It has more significance of application. Combination with quad-core embedded as the audio data stream processing hardware equipment, make the speech synthesis output naturalness is better and without noise.The research work of this paper is as follows:(1) First discuss the source of this topic, explain the research purpose and the significance of this topic in detail. This paper research the current speech synthesis technology to design speech synthesis system with better nature degree.(2) Then based on a detailed study of i.MX6 Q quad-core processor-based, research the designing of audio driver and building of speech synthesis system, developed the audio driver and designed the framework of speech synthesis system under the environment of Linux. On the basis of above, sort out the overall design scheme of the system.(3) Combining i.MX6 Q which as the processor and TLV320AIC3106 audio codec chip which has excellent performance to design the hardware connection. This includes the design of bus connection circuit and the audio output circuit. The bus architecture design depends on the I2 C bus driver.(4) The next work is design the audio driver. Audio driver development build reference with ALSA audio architecture, create registered sound card related equipment, including the sound card instance, component management, PCM equipment and control devices. Write TLV320AIC3106 audio codec chip driver code and debug compilation, then finally install the driver to the application of plate test. So we complete audio driver development(5) Design the speech synthesis system combined with the actual hardware platform design by calling the speech engine Aisound5.0. Including access of voice resources packet, create TTS instances, calls of the callback function and synthesis parameter settings. Transplanting the speech synthesis system to the application board, then all the speech synthesis system has been completed. Through equipment connection and communication protocol of the command frame control the speech synthesis system, test the pronunciation of synthesis and the effect of the speech synthesis system, finally analyzed the results of speech synthesis. |
