Nowadays human-centered multimedia (HCM) systems are growing rapidly, and HCM systems are broadly accessible to users over mobile networks. As one of the most important human-centered computing services, video service constitutes the main part of the overall mobile network traffic in an HCM system. However, video service currently faces significant technical challenges. In particular, despite significant improvements in mobile network bandwidth and processor performance, the online video services on smart devices (e.g., smartphones, tablets) are still constrained by limited battery capacity. This contradiction underscores the inadequacies of existing battery energy consumption optimization strategies in dealing with various usage scenarios, especially in meeting the diverse battery-saving needs of online video stream users. In response to this issue, we have proposed an efficient energy-aware adaptive video service scheme. This scheme divides the battery power into different energy levels, and then dynamically adjusts the playing parameters (including the threshold of the warm-up stage, capacity of the video buffer, etc.) according to the energy level, while helping the edge client to flexibly capture and play back the required videos. Based on designed experiments already conducted on the proposed method, the results demonstrate the effectiveness of this method in optimizing energy consumption for video services.