Implementation of a surface reconstruction algorithm using dictionary learning

• Main Author: Luong, Viet Hoang
• Other Authors: Qian Kemao
• Format: Final Year Project
• Language: English
• Published: 2018
• Subjects: DRNTU::Engineering::Computer science and engineering::Computing methodologies::Computer graphics
• Online Access: http://hdl.handle.net/10356/74347
• Institution: Nanyang Technological University

Triangular mesh reconstruction from point cloud data has always been an important practice in computer graphic. There are several existing reconstruction methods to choose from, all with their own strengths and limitations. However, one solution, proposed in the paper: Robust Surface Reconstruction via Dictionary Learning [Shiyao Xiong et al. 2014], interests us the most due to its outstanding accuracy and robustness comparing to other algorithms. It makes use of the dictionary learning framework, where the dictionary is the positions of the reconstructed mesh vertices, and the sparse coding matrix is the connectivity of those vertices. In this final year project, we tried to create a program to implement this algorithm, eventually provide NTU researchers with a powerful tool for their surface reconstructing works. As we developed the program, we encountered many practical problems that were not mentioned in the original paper and came up with different solutions for them. The purpose of this report is to summarize what we have achieved so far, explain the reasoning behind our code and provide a reference document for anyone who will continue working on this project in the future.