SPRITE TREE: AN EFFICIENT IMAGE-BASED REPRESENTATION FOR NETWORKED VIRTUAL ENVIRONMENTS

Abstract

WE PROPOSE A NEW AND EFFICIENT IMAGE-BASED REPRESENTATION FOR NETWORKED VIRTUAL ENVIRONMENTS, CALLED THE SPRITE TREE. A SPRITE TREE ORGANIZES MULTIPLE REFERENCE IMAGES EFFICIENTLY AND COMPACTLY FOR ACCELERATING THE RENDERING OF COMPLEX VIRTUAL SCENES. USING OUR BASIC CONSTRUCTION AND RENDERING METHODS, THE RESULTS SHOW THAT A SPRITE TREE CAN EFFICIENTLY ORGANIZE THE PIXELS FROM HUNDREDS OF DISTINCTIVE REFERENCE IMAGES AND PRODUCE ACCEPTABLE VISUAL QUALITY. FURTHERMORE, WE PROPOSE THE SPRITE VIEW SIMILARITY TO (I) LARGELY REDUCE THE LIGHTING ARTIFACTS IN THE RENDERED IMAGES, (II) SIGNIFICANTLY REDUCE THE REDUNDANCY AND THE TREE SIZE WITH LITTLE LOSS OF THE VISUAL QUALITY, AND (III) EFFICIENTLY ACCESS THE VISUAL QUALITY WITHOUT REFERENCES. MOREOVER, WE PROPOSE A REMOTE RENDERING ARCHITECTURE FOR THE SPRITE TREE, AND PROPOSE TO PREFETCH AND CACHE THE MOSTLY LIKELY USED SPRITES BASED ON THE SPRITE ACCESS PROBABILITY. THE RESULTS SHOW THAT OUR METHODS ARE MORE EFFICIENT THAN THE TRADITIONAL