Design And Implementation Of Speaker Tracking System Based On Android

With the development of the times and the advancement of science and technology,more and more devices can be applied to video recording,such as cameras,digital video cameras,mobile phones,etc.These devices have been widely used in various fields,such as video surveillance,large conference shooting,travel photography,etc,and they have become an integral part of people's lives and work.However,the common shooting methods are usually fixed in a specific direction or need to be controlled by people.These shooting methods do not have the effect of tracking video.Therefore,in terms of how to track video,this paper designs and implements a “speaker tracking” system by using the Android mobile phone as video equipment,the speaker's face as the video tracking target,and combining face detection technology.This system mainly includes four parts: Android mobile client,pan-tilt part,server and PC web client.The function realized by the Android mobile phone client is to detect the position of the human face when the mobile phone is recording video,and send the control word instruction to the pan-tilt part through the mobile phone Bluetooth.The function of pan-tilt part is to control the rotation of the mobile phone according to the control word instruction so as to achieve the effect of tracking the human face.The function of the server is to realize the data interaction with the Android phone client and PC web client.The function of the PC web client is to manage the photos and video files uploaded to the server by the Android mobile client through a page mode.The main work of this system is as follows:(1)The face detection technology is implemented on the Android phone client by analyzing the existing face detection technology.Through the analysis and comparative study,the video recording and face detection have been achieved by using the new Camera2 API+MediaRecoder after Android 5.0.The MediaRecoder realizes the function of video recording,and Camera2 realizes the detection of face data of each frame of video data in the process of recording video on mobile phone.The position of the face in each frame of video is framed by analyzing and transforming the obtained face data,and adding Surface View.Finally,the coordinates of the midpoint of the face position are calculated to form the pan-tilt control word,and then the control word command is sent to the pan-tilt by the mobile phone Bluetooth.(2)The pan-tilt uses steering pan-tilt to track people's faces.The pan-tilt is composed of a STM32 SCM minimum system,a bluetooth transceiver module and two steering engines.The Bluetooth module is responsible for receiving the control word commands sent by the Android mobile client,then handing over to the STM32 SCM for processing.The STM32 SCM controls the rotation of the four directions of two steering engines by output PWM wave.Since the Android mobile phone is placed on the bracket of steering engines,the mobile phone can be rotated by the rotation of the steering engines to achieve the effect of tracking the human face.(3)The server is implemented by SSH framework +MySQL database +Tomcat server.The function of the server is to implement data interaction with the Android mobile client and the PC web client.After the Android mobile client completes video recording or taking pictures,the video files or photos that have been stored on the Android phone can be chose to upload to the server.(4)The PC web client is mainly implemented by JSP technology + the Struts2 tag in the SSH framework.The main function of PC web client is to manage video files or photos in the server through the page,and to download or delete such files.(5)Various functions of Android mobile phone client,pan-tilt and PC web client are comprehensively tested,and the test results reach the design requirements of the system.