The Design And Implementation Of Streaming Media Player With Identification Based On Android

With the popularity of smart phones and the rapid development of mobile Internet, mobile phone users have a convenient and rich life with the mobile terminals. Android plays an important role in the mobile terminal because of Android’s popularity and openness, it makes the study and design on streaming media player in Android has certain significance. The player with a text-independent speaker recognition function was designed, in addition to the function of playing in local and network, speaker recognition function was used to distinguish whether the speaker has permissions.In this paper, with the background of the video monitoring, the basic function of video monitoring which based on RTSP and RTP protocol and the control of yuntai were realized. For video on demand, HTTP protocol is to be supported. The player should support the local video playback and other function such as identification. Under these requirements, this paper mainly completed the following work:(1) In this paper, design and implementation of client UI interface has be carried on the Android platform with Android SDK. The control of yuntai and the function of plist resolution were also completed with Android SDK. The related functions such as play were encapsulated in the native layer.(2) FFmpeg was transplanted to Android in this paper to realize software decoding and the OpenMAX was used in Android multimedia module to achieve the function of hardware decodeing. Then the player would get the YUV data and convert YUV data into the RGB data which is supported in Android. The efficiency of the two kinds of color space conversion was compared. This paper also made a contrast between several kinds of rendering and calculated the average time. Also NEON optimization and other related technology were illustrated. Finally the audio and video were synchronized.(3) A text-independent speaker recognition function was added in this paper and algorithm was used in the paper are mature, using MEL cepstrum coefficient to extract the voiceprint characteristics and using the gaussian mixture model to model training, finally the function was realized on the Android.(4) All the function of each part of the streaming media player was tested at last.