期刊名称：International Journal of Innovative Research in Computer and Communication Engineering
印刷版ISSN：2320-9798
电子版ISSN：2320-9801
出版年度：2015
卷号：3
期号：6
DOI：10.15680/ijircce.2015.0306090
出版社：S&S Publications
摘要：Reconstructing a 3D digital object from a set of points sampling real world objects has gained largeattention from the computer graphics researchers. However, the 3D surface reconstruction of large model requires theprocessing of large-scale dataset which is indeed a very computationally expensive task. Therefore, a parallel algorithmmust be adopted to improve the performance efficiency of the 3D surface reconstruction process. Even the parallelalgorithms for the 3D surface reconstruction require high performance computing environment. The use of expensivehigh performance computing machines, such as supercomputers, is not practical for educational and research applications.Recently, Cloud Computing offers an outstanding potential of low cost high performance computing environment forsuch distributed large-scale applications. This paper presents a framework for high-performance execution of a parallelsurface reconstruction of 3D objects on a private Cloud environment. First, a parallel algorithm for the 3D surfacereconstruction is designed using the Master-Slave technique. The master computation node partitions the main datasetinto a number subsets of data for the slaves to perform the parallel computations. The slave computation nodes simplyperforms the same computations on different datasets received from the master computation node. Finally, the slavessend the results back to the master node to finalize the computations. Experimental evaluation shows that the frameworkeffectively enhances the time required to sketch the overall 3D reconstruction process.
