Abstract
Users need to record whole motions on their computer screens in many applications, such as distance learning, distance conference, thin client, share desktop and remote control etc. In order to conveniently use these applications, a problem, called asymmetric screen resolution (ASR), needs to be solved. The ASR problem means that the resolution of a remote computer is different from that of a client computer. Usually, one computer, called ¡§remote computer¡¨, records its screen and sends this recorded data to another computer, called ¡§client computer¡¨, for display through network transmission. Previously, resolutions of computers are only two categories, either 800x600 or 1024x768. Users can easily switch their computer¡¦s resolution between these two kinds of resolutions when adjusting resolution is necessary. However, resolutions of current computers are more than thirteen categories. Furthermore, many computers have their own specific resolutions and not all computer screens support these specific resolutions. If resolution of a remote computer is larger than that of a client computer, the ASR problem will leads to roll scroll bars to explore uncovered area of a computer screen while displaying a remote computer screen in a client computer with small resolution. Conversely, if resolution of a remote computer is smaller than that of a client computer, the ASR problem will lead to only use a small area in a large client computer screen. Traditionally, handling the ASR problem uses resizing whole screen images with system API, such as stretchblt ( ) of Microsoft windows system, in a client computer. However, this resizing whole image method requires too much network bandwidth and CPU utilization. This study addresses the R2A and R2A+ methods to handle real-time resolution adaptation for dealing with the ASR problem in screen recording applications. This paper tests the resolution adaptation quality and CPU utilization of the R2A and R2A+ methods in four different application scenarios, web browsing, slide show, text editing, and video playback. The experimental results show that integrating R2A and R2A+ methods utilizes less CPU utilization and guarantees the display quality at the same time in different applications. Compared with that of resizing whole image method, the integration method reduces CPU utilization by more than 30%.