Parallel Research On Surface Reconstruction Algorithm Based On Machine Learning

Machine learning and related algorithms are becoming more and more popular in recent years. With the development of 3D surface reconstruction technology, it is widely used in more and more fields. Used in this machine learning algorithm of radial basis function interpolation and radial basis function neural network are discussed in the description of the surface reconstruction algorithm, serial algorithm in the selection of basis function interpolation data center point, computing local surface information function, and the global interpolation calculation of implicit surface process time-consuming part of the analysis, based on the OpenMP multi-core parallel technology and CUDA parallel technology respectively, parallel design is carried out for surface reconstruction algorithm based on radial basis function interpolation. Finally, it puts forward and CUDA OpenMP hybrid parallel architecture of parallel algorithm based on. The experimental results show that the parallel algorithm can obtain a better speedup without affecting the reconstruction effect.In this paper, we first introduce the related research situation in recent years at home and abroad of 3D surface reconstruction, to explain the radial basis function of the basic concepts of the theory and radial basis function interpolation method, and introduces the artificial neural network and the basic theory of learning style, and describes the basic process of implicit surface reconstruction. On existing parallel computing techniques are also described, focusing on the description of the radial basis function neural network (RBFNN) surface reconstruction and radial basis function (RBF) implicit surface function reconstruction, and the design and implementation of a parallel algorithm based on CUDA OpenMP and radial basis function (RBF) implicit surface reconstruction of parallel algorithms, and parallel hybrid architecture of. Finally, the acceleration effects of different parallel technologies are analyzed by experiments, and the related performance analysis is made.