Intelligent Digital Tutor to Assemble Puzzles Based on Artificial Intelligence Techniques

Abstract

The potential of applying image processing tools and artificial intelligence in learning processes is identified. This article presents the development of a digital tutor that helps solve a puzzle. Point cloud technology is used to identify each person’s interaction dynamically in space. The proposed methodology is developed in several stages. The first stage consists of the acquisition and pre-processing system for adquaring the user environment. The second stage consists of recognizing the piece in the puzzle, at this stage it is necessary to develop a database of the particular puzzle. In the identification process, the PCA algorithm is implemented as a complementary strategy to the use of the neural network. The last stage implements a general search algorithm as the core of the decision system. This methodology is presented as an iterative process and evolves over time according to the interaction with the user. The results are presented through confusion matrix which exhibits a performance of 92.7% assertiveness. Finally, the potential of using this methodological structure in different cognitive processes with puzzles with different levels of difficulty is raised.