A vision system for detection of construction materials

With Deep Learning (DL) emerging from Machine Learning (ML) to become one of the greatest technological advancement and invention in today’s day and age. DL methods and techniques are becoming a pivotal part in our initiatives to Industry 4.0. Convolutional Neural Network (CNN) is an important archi...

Full description

Saved in:
Bibliographic Details
Main Author: Kabilan Elangovan
Other Authors: CHEAH Chien Chern
Format: Final Year Project
Language:English
Published: Nanyang Technological University 2020
Subjects:
Online Access:https://hdl.handle.net/10356/139195
Tags: Add Tag
No Tags, Be the first to tag this record!
Institution: Nanyang Technological University
Language: English
Description
Summary:With Deep Learning (DL) emerging from Machine Learning (ML) to become one of the greatest technological advancement and invention in today’s day and age. DL methods and techniques are becoming a pivotal part in our initiatives to Industry 4.0. Convolutional Neural Network (CNN) is an important architecture of DL. CNN has achieved astounding results in the area of image recognition and object detection. However, CNN can be extensive and thus carrying a high load of computational processes. As such You Only Look Once (YOLO), a form of CNN was developed to perform object detection and classification with a smaller architecture and faster computing capabilities. Therefore, the aim of this project is to employ YOLO as main object detection technique to detect concrete structures and various concrete defects as an initiative to improve the productivity in Construction Industries. Furthermore, this project also focuses on a Computer Vision (CV) technique to retrieve the third dimensional parameter of concrete structures via the use of detection results from YOLO.