Large area interactive browsing for high resolution digitized Dunhuang murals

Abstract

The Dunhuang Motao Grottoes consist of about 45000 square meters murals, which are the most important part of the Dunhuang Art. As Dunhuang murals are rich in contents, large in sizes and amounts, high resolution and high fidelity acquisition techniques are required for storage. It is significantly important to exhibit the murals vividly and tell the recondite stories to the public. How to browse these high resolution mural images effectively is an important technical issue. In the paper, a novel multi-resolution image partition method is presented, thus the system can efficiently load and splice these partitioned images with proper resolution during browsing. A natural gesture interaction approach is also brought forward for convenient browsing of the murals. Gesture detection is used to get the moving and zooming operation instructions from the user. Using gesture approaches, the user can interact with the system without wearing or operating any device, but swinging the arms. The results show that the system can effectively exhibit high resolution mural images with convenient interactions. © 2009 Springer-Verlag.