| With the development of the network and the robotics, telepresence robotcan do lots of real-time work in remote or dangerous areas instead of humans.In telepresence robot system, the video quality on the control side decides thecurrent satisfaction of the teleoperation. Because of the uncertainty of thenetwork, especially the limited bandwidth, delay and jitter in the network, canhave a serious influence on the video quality and operation experience, whichwill bring challenge to the real-time teleoperation. Moreover, the telepresencerobot system differs from the video conferencing. Because of the mobility ofthe robot, the video which is captured by the robot moves heavily and theimage data is tremendous; the network conditions which the robot locateschanges heavily, which means the bandwidth between the user and the robotare usually limited; the target information which appears in the scope of thecamera can have an effect on the demand of the video quality for the user. So,the research on the video quality on the control side in robot telepresencesystem is of great importance.In order to solve the problems mentioned above, the design oftelepresence robot system is conducted. The system based on FLASH andWebRTC is explained and the advantages and disadvantages of the twosystems are given. After the analysis, the system based on FLASH is chosenfor the video transmission optimization algorithm based on QoE.To improve the video quality on the operation side in telepresence robotsystem, an optimization algorithm of video transmission in telepresence robotsystem based on QoE(Quality of Experience) was presented in this paper.Firstly, with the mobility of the telepresence robot and the target velocity inthe scope of camera, an identification of the video parameters is conducted forcomputing the QoE of network’s and user’s. Then the optimization function is established and the video parameters which affect QoE are optimized inmultivariable optimization method in order to improve the video quality onthe operation side. The user tests showed that this algorithm can reduce thepacket loss in video transmission, improve the video quality and operationexperience in limited bandwidth.For the system based on WebRTC，the experiments for the transmissionrate and video quality are conducted. The final data transmission test and usertest show that the system have lower transmission rate and higher videoquality than the FLASH system. Finally, the test on the internet between thetwo systems is conducted and the result shows that the system based onWebRTC has higher video quality and better control experience than thesystem based on FLASH.The main content of this thesis is summarized as bellow:1．A introduction to the two robot telepresence systems. The designtarget and system architecture are illustrated. The advantages anddisadvantages of the two systems are summarized.2．An introduction to the QoE theis and calculation method. Theoptimization question of the QoE in robot telepresence system is analyzed.For the QoENand QoEU, an optimization method is given and the flow chartof the algorithm is summaried.3. For the uncertain parameters in optimization method, the identificationof the video parameters is conducted. Describing the robot’s motion statevariables on the relationship between the parameters of video, and theapplication of the parameters is given.4. For the access of the mobile devices to our telepresence robot system,a system based on WebRTC is proposed and the hardware and the softwareblock diagram are given.5. Experiments for the QoE algorithm and the bandwidth test for themobile devices. The analysis of the results is given.6. Summary of the paper and the scientific research directions. |
PDF Full Text Request