Object recognition using Kinect sensor

• Main Author: Chia, Dillon Jia Ming
• Other Authors: Chau Lap Pui
• Format: Final Year Project
• Language: English
• Published: 2019
• Subjects: DRNTU::Engineering::Electrical and electronic engineering
• Online Access: http://hdl.handle.net/10356/77360
• Institution: Nanyang Technological University

Enhanced Seatbelt reminder system(ESRS) has been a mainstay in the auto mobile industry for many years. It has become common practice among companies to ship their vehicles to consumers with some form of the ESRS installed on board for the end user. The system includes various forms of reminders for the vehicle to remind the passengers on board to put on their seatbelts when the vehicle is in operation. However, this system relies very heavily on weight sensors underneath the seats, which only comes shipped in most cars in the front two passengers, due to the airbag that is used for the front two passengers in the case of an accident. Therefore, those passengers in the back seats do not receive any of such service from the ESRS system. In this project, we will conduct an in depth look into what the ESRS does and how it works. We also implemented a simple image recognition system through the use of a Kinect motion sensor to collect data from passengers riding in the car and then build a deep learning model in python which was fed with these data to be trained to fulfil the same purpose of the ESRS in modern day vehicles and establish such a service provided for all the passengers in the car. We will also discuss the results obtained through the project and future implementations in future implementations of this project that could be taken up by other Final year project students.