3D personal navigation in smart phone using geocoded images

Abstract

Three-dimensional personal navigation in smart phones develops in a fantastic speed nowadays. This topic has already attracts many attentions both from the scientific and industrial field. Such kind of 3D navigation application is mostly fulfilled by importing 3D models constructed in a computer into a mobile phone. However, comparing the computer platform, the mobile devices typically have the several limitations such as a weaker CPU, less memory and other hardware restrictions. All these limitations severely restrict the performance of the 3D personal navigation. The paper proposes a novel solution which is based on geocoded images rather than the 3D models to mitigate these technical restrictions in a smart phone platform and achieves a better performance and smoother human-computer interaction. We design a program base on S60 platform running on a Nokia 6710 handset. A five layers configuration is applied in this program. An arc-direction-distance geocoding method is adopted for the file nomination. Several lab and field tests were carried out to compare various aspects between the geocoded image based solution with 3D model based solution and evaluate the performance of the proposed solution. In addition, this solution provides an optional way to decrease the labor cost and working time by building up a similar solution using photos instead of complex 3D model construction.