Big data simulation for surface reconstruction on CPU-GPU platform

Abstract

This paper discusses on CPU-GPU collaboration in surface reconstruction process from CT scan image, in this case human head. This type of image consists hundreds of data points and needs a powerful processor to ensure the accurateness of result. Therefore, an architecture to segment the whole process to CPU and GPU is designed. Comparison of CPU and GPU is done based on execution time and speedup for total number of the data points up to 105. From the analysis, it is shown that the maximum speedup is more than 15 for a process.