Visual Based Cart Follower Using Artificial Neural Network

Johari, Mohamad Faiz Ahmad (2019) Visual Based Cart Follower Using Artificial Neural Network. Masters thesis, Universiti Sains Malaysia.

Download (424kB) | Preview


A visual based cart follower can benefit as a helper robot. It can track and follow a wheelchair user without having any physical attachment between them. In addition, the low intensity of the surrounding light can affect the tracking performance too. In this study, the cart follower that equipped with all tracking component has been developed. The system was also integrated with Artificial Neural Network (ANN) for good visual navigation. A colour tracking method being used for following task application with Pixy CMUcam5 camera. It gathered the information of the width, height, area, angle, x and y coordination of the colour pattern board which situated behind the wheelchair and translate this information into relative position information which enable the cart to follow the wheelchair. The activation function being used is saturating linear (satlin). The Field of View (FOV) of Pixy CMUcam5 is from 69.98o to 76.83o with vertical distance of 20cm to 150cm. The optimum target colour size for maximum distance 150cm is 98.07cm2. The distance from the top view shows that the minimum and maximum distance error is 0.40cm and 2.30cm while the maximum and minimum angle error is 5.30o and 21.30o from point P0 to P1, P2 and P3 respectively. The most ideal tracking condition is at 205 Lux since the error rate for each R, G and B value is the lowest. The final error simulation test shows that there is 0.65% and 4.27% of error in minimum distance 20cm and -15o angle while 1.93% and 5.57% of error in maximum distance 69cm and 30o angle. The overall test performance shows that the error occurred in distance is 1.62% meanwhile 5.39% in angle. As a conclusion, the tracking system for cart follower has been developed and integration of ANN has achieved its deserved accuracy with the final error test.

Item Type: Thesis (Masters)
Subjects: T Technology
T Technology > TK Electrical Engineering. Electronics. Nuclear Engineering
Divisions: Kampus Kejuruteraan (Engineering Campus) > Pusat Pengajian Kejuruteraaan Elektrik & Elektronik (School of Electrical & Electronic Engineering) > Thesis
Depositing User: Mr Mohamed Yunus Mat Yusof
Date Deposited: 09 Jul 2020 01:26
Last Modified: 17 Nov 2021 03:42

Actions (login required)

View Item View Item