Text
Approximation of point clouds in two dimensions using chebyshev polynomials
This thesis investigates the approximation of point clouds by utilizing chebyshev polynomials in two dimensions. In particular, a two dimensional chebyshev polynomial functtion is defined as an approximated surface of the point clouds. Moreover, by incorporating fast fourier transform in the determination of Chebyshev coefficients, Chebyshev polynomials can be accomplished with the low computation times. According to the numerical investigation, two dimensional chebyshev polynomial approximation via fast fourier transform can be determined in a fraction of asecond.
Equally important, we proposed an approach to optimized the chebyshev polynomial approximations by introducing a threshold which acts as the break-off value in the automated loop condition and to complete the truncations of the Chebyshev coefficients. Thus, unnecessary coefficients can be excluded witch reduce the computation complexity and proven to raise the Chebyshev polynomial approximations precision.
A Further investigation of chebyshev polynomial approximations was applied to deformation analysis. Detection and localization of deformation were assessd from two different epochof point clouds. A statistical test was applied on Chebyshev coefficients of the two epochs in order to measure its statistical differences between the two coefficients. Accordingly, the Chebyshev polynomial approximations of the two epochs of point clods were performed to localized the deformation.
B20190930331 | 526.1 YOR a | Perpustakaan BIG (500-) | Tersedia |
Tidak tersedia versi lain