Embedded hardware implementation for real time hand gesture recognition / Farah Farhana Mod Ma'asum

This research is focused in designing a low-cost embedded system which can produce robust marker-less tracking system. The application will benefit medical, disabled person and factories, since it can perform as a replacement of mouse cursor by just gesturing motion in the air. The performance of th...

Full description

Bibliographic Details
Main Author:	Mod Ma'asum, Farah Farhana
Format:	Thesis
Language:	English
Published:	2017
Subjects:	Telecommunication Scanning systems
Online Access:	http://ir.uitm.edu.my/id/eprint/21629/ http://ir.uitm.edu.my/id/eprint/21629/1/TM_FARAH%20FARHANA%20MOD%20MA%27ASUM%20EE%2017_5.pdf

id	uitm-21629
recordtype	eprints
spelling	uitm-216292018-09-26T03:14:14Z http://ir.uitm.edu.my/id/eprint/21629/ Embedded hardware implementation for real time hand gesture recognition / Farah Farhana Mod Ma'asum Mod Ma'asum, Farah Farhana Telecommunication Scanning systems This research is focused in designing a low-cost embedded system which can produce robust marker-less tracking system. The application will benefit medical, disabled person and factories, since it can perform as a replacement of mouse cursor by just gesturing motion in the air. The performance of the hand gesture image recognition, segmentation technique and feature classification technique is established as part of the processes. There are four main phases were set up in achieving the research objectives. Initially, image of hand which are being captured are being segmented using Canny and Otsu threshold technique. Then, the hand image is extracted using convex hull and convexity technique while angle of fingertips is obtained from feature vector representation. Three actions are classified: MOVE, RIGHT CLICK and LEFT CLICK cursor. All these actions are then demonstrated with the Arduino board to verify that all techniques are authenticated based on the signal sent by hand gesture. An experiment is set up for 10 users for validation. Also, the users are trained to familiarize with the gesture system. The results revealed that the users are better trained in controlling their fingertips after five-minute of training in the second trial. The findings show that an increase in the LEFT CLICK action is achieved from 33.3% to 52.6%. The RIGHT CLICK is improved from 46.7% to 61% while 56.7% to 77.3 % for MOVE cursor. The results indicate that the system is capable to replace the multi-touch modalities. In addition, there are three different LED colors: RED, YELLOW and BLACK are embedded to the system to represent the gesture - MOVE, RIGHT CLICK and LEFT CLICK respectively, using serial communication. For that reason, low-cost embedded system for marker-less tracking system has been verified to obtain good gesture recognition. 2017 Thesis NonPeerReviewed text en http://ir.uitm.edu.my/id/eprint/21629/1/TM_FARAH%20FARHANA%20MOD%20MA%27ASUM%20EE%2017_5.pdf Mod Ma'asum, Farah Farhana (2017) Embedded hardware implementation for real time hand gesture recognition / Farah Farhana Mod Ma'asum. Masters thesis, Universiti Teknologi MARA.
repository_type	Digital Repository
institution_category	Local University
institution	Universiti Teknologi MARA
building	UiTM Institutional Repository
collection	Online Access
language	English
topic	Telecommunication Scanning systems
spellingShingle	Telecommunication Scanning systems Mod Ma'asum, Farah Farhana Embedded hardware implementation for real time hand gesture recognition / Farah Farhana Mod Ma'asum
description	This research is focused in designing a low-cost embedded system which can produce robust marker-less tracking system. The application will benefit medical, disabled person and factories, since it can perform as a replacement of mouse cursor by just gesturing motion in the air. The performance of the hand gesture image recognition, segmentation technique and feature classification technique is established as part of the processes. There are four main phases were set up in achieving the research objectives. Initially, image of hand which are being captured are being segmented using Canny and Otsu threshold technique. Then, the hand image is extracted using convex hull and convexity technique while angle of fingertips is obtained from feature vector representation. Three actions are classified: MOVE, RIGHT CLICK and LEFT CLICK cursor. All these actions are then demonstrated with the Arduino board to verify that all techniques are authenticated based on the signal sent by hand gesture. An experiment is set up for 10 users for validation. Also, the users are trained to familiarize with the gesture system. The results revealed that the users are better trained in controlling their fingertips after five-minute of training in the second trial. The findings show that an increase in the LEFT CLICK action is achieved from 33.3% to 52.6%. The RIGHT CLICK is improved from 46.7% to 61% while 56.7% to 77.3 % for MOVE cursor. The results indicate that the system is capable to replace the multi-touch modalities. In addition, there are three different LED colors: RED, YELLOW and BLACK are embedded to the system to represent the gesture - MOVE, RIGHT CLICK and LEFT CLICK respectively, using serial communication. For that reason, low-cost embedded system for marker-less tracking system has been verified to obtain good gesture recognition.
format	Thesis
author	Mod Ma'asum, Farah Farhana
author_facet	Mod Ma'asum, Farah Farhana
author_sort	Mod Ma'asum, Farah Farhana
title	Embedded hardware implementation for real time hand gesture recognition / Farah Farhana Mod Ma'asum
title_short	Embedded hardware implementation for real time hand gesture recognition / Farah Farhana Mod Ma'asum
title_full	Embedded hardware implementation for real time hand gesture recognition / Farah Farhana Mod Ma'asum
title_fullStr	Embedded hardware implementation for real time hand gesture recognition / Farah Farhana Mod Ma'asum
title_full_unstemmed	Embedded hardware implementation for real time hand gesture recognition / Farah Farhana Mod Ma'asum
title_sort	embedded hardware implementation for real time hand gesture recognition / farah farhana mod ma'asum
publishDate	2017
url	http://ir.uitm.edu.my/id/eprint/21629/ http://ir.uitm.edu.my/id/eprint/21629/1/TM_FARAH%20FARHANA%20MOD%20MA%27ASUM%20EE%2017_5.pdf
first_indexed	2023-09-18T23:06:59Z
last_indexed	2023-09-18T23:06:59Z
_version_	1777418548754251776

Embedded hardware implementation for real time hand gesture recognition / Farah Farhana Mod Ma'asum

Similar Items