A 3D Solid Texture Feature Computation Method using Multi-Core CPU Architectures

M.T. Suzuki, Y. Yaginuma, and H. Kodama (Japan)

Keywords

MultiCore CPU Architectures, Image segmentation, Image texture analysis, Fractals

Abstract

This paper describes a pattern feature extraction system for 3D volumetric data. The uses of 3D volumetric are increasing in various applications. Pattern feature extraction is an important task for many computer vision software programs, and the extracted pattern features can be used for 3D volumetric data analysis including segmentation, classification, and retrievals. We have developed an experimental system to estimate fractal dimension of 3D volumetric data. Since the multi-threaded program can be executed concurrently on multi-core CPU computers, the 3D volumetric can be analyzed in parallel. The efficiency of the multi-threaded program to handle 3D volumetric data was tested using a popular desktop PC which uses a multi-core CPU architecture. Preliminary experiments show fairly efficient computation time for analyzing 3D volumetric data.

Important Links:



Go Back