研究目的
To propose a method to accelerate data loading for Photo AR operating on a web browser in a mobile device.
研究成果
The proposed method to accelerate the data loading of Photo AR operating on a web browser in mobile devices is effective in shortening the startup time.
研究不足
The acceleration of data loading using either method is not sufficiently effective for mobile devices because of their consumer-grade GPU.
1:Experimental Design and Method Selection:
The study proposes a hybrid method to decode multiple videos for Photo AR in parallel, whose frame contains multiple pictures.
2:Sample Selection and Data Sources:
The performance of the video decoder of GPUs in mobile devices (Apple iPad mini and Google Nexus6P) is analyzed.
3:List of Experimental Equipment and Materials:
Apple iPad mini (CPU: ARMv8-A, 2 core,
4:5 GHz, RAM:
2 GB, browser: Safari v.
5:0) and Google Nexus6P (CPU:
Snapdragon 810, RAM: 3 GB, browser: Chrome v.
6:87). Experimental Procedures and Operational Workflow:
The relationship between the video frame size and the loading time of each frame is measured.
7:Data Analysis Methods:
The data amount of the video file generated by two methods of picture arrangement is compared.
独家科研数据包,助您复现前沿成果,加速创新突破
获取完整内容