A Tangible Programming Solution To Develop Computational Thinking: Programming Block Recognition Using Convolution Neural Network

Chin, Khai Hong (2017) A Tangible Programming Solution To Develop Computational Thinking: Programming Block Recognition Using Convolution Neural Network. Project Report. Universiti Teknikal Malaysia Melaka, Melaka, Malaysia. (Submitted)

Text (24 Pages) A Tangible Programming Solution To Develop Computational Thinking Programming Block Recognition Using Convolution Neural Network.pdf - Submitted Version Download (433kB)

Abstract

Computational thinking development with the word-based programming language is not efficient enough for children in early age due to the complexity in format and syntax. Tangible programming is a good approach to develop computational thinking for children at early age.In this project, an algorithm is developed to localize and classify τ-block using Convolutional Neural Network (CNN) and a program is developed to determine the arrangement of τ-block and transfer the command sequentially to the τ-bot wirelessly.In this project,an approach known as Region-with Convolutional Neural Network (R-CNN) is used to develop the detection algorithm.The developed detection model is evaluated with several cases and the result is promising.The average accuracy of the detection with the developed model is 99.59%.Besides that,a web application is developed to simulate the overall system of the tangible programming tools using HTML,CSS and JavaScript.

Actions (login required)

View Item