A Study on the Design and Production of 3d Traditional Stone Pagoda Dataset for Artificial Intelligence Learning

Abstract

Big data in the form of DataSets is crucial for artificial intelligence learning. Currently, while 2D artificial intelligence restoration algorithms are extensively studied and applied, there is still a deficiency in 3D artificial intelligence training data. In this research, we studied the process of designing and creating AI learning data for traditional towers configured in 3D. In the case of 3D AI learning data, the number of points, lines, and surfaces significantly influences the learning speed and accuracy of AI. In this paper, we analyzed and designed traditional stone towers in 3D, resulting in the creation of an optimized dataset based on points, lines, and surfaces. Thus, we propose a method of optimizing the production of datasets required for AI learning from 3D objects. This approach is anticipated to contribute significantly to future 3D artificial intelligence learning.