Application Of Stereoscopic Display Technology In Mobile Terminal

Mobil terminal industry is growing toward the trend of intelligent, entertainmentand multi-functional, with the development of latest communication technology andchip integration technology. Reappearing of the three-dimension sense with depthinformation on a mobile terminal platform is now becoming a new market demandand a new business growing point. This paper aims to develop a standard andpractical method in both software and hardware to solve the problem of stereoscopicimage acquisition and display, as well as the implementation and optimization ofdual-view synthesis algorithms in mobile terminal platform.To solve the problem of EMI noise, complex interface design, high powerconsumption bring by parallel interface, in the area of multimedia, MDDI and MIPIhas appear and occupied the market. MIPI specification has a strong integratedsupport team, it’s modular, hierarchical design shows a powerful ascendancy in theprocess of completing. The method of acquisition of double view stereo image shownin this article is based on MIPI interface specification, whose single-channel datatransfer rate meets up to1Gbps. With the extending of MIPI interface and the controlof the data flow between memory and processor using the method of Time DivisionMultiplexing, this article has reached the stated objectives.Based on Linux kernel, Android in recent years has become the dominantoperating system in smart mobile terminal platform, and its open source feature,hierarchical design and a strong technical support team has helped to become theleader in embedded operation system. The video driver framework in Android iscalled V4L2, the implementation of driver in both Linux kernel and hardwareabstraction layer in this article has all been based on V4L2. With the integration ofdouble view stereo image synthesis algorithm, the system has achieved its target.To improve the3D display effect on mobile terminal, this article also proposed amethod of optimization. With the matching of Scale-Invariant-Feature-Transformfeature between images at double viewpoints, the coordinate mapping model can bebuilt by solving the basic matrix, using this model3D material without a high qualitycan be reconstructed.