Machine vision and sensing with an Android

Field, Shaun (2015) Machine vision and sensing with an Android. [USQ Project]

[img]
Preview
Text
Field_S_Billingsley.pdf

Download (2265Kb) | Preview

Abstract

This project investigated the ability for an Android mobile device to run an application that could automate a tractor. The development of such an application would lead to a cost effective, portable, and user friendly device that could easily be transported and installed on a tractor to allow vehicle automation. At the start of this project the
method for automation had not been determined however the specific intent for the design of a machine vision application on an Android device was later defined.

The development of this application began with investigations into machine vision techniques and the Android SDK which identified the machine vision algorithm as
well as the software libraries the application was be built upon. Access to the main video data was then achieved which enabled the manipulation of image data through accessing the pixel array information. Annotations were then added to the screen to allow for the output of data, and the line fitting algorithm selected for identifying crop rows was programmed. These achievements allowed the output of row identification and steering correction data to be added to the device screen.

These accomplishments concluded in an Android based machine vision application that is able to identify crop rows while processing the 30 fps 320x240 resolution image in an average of 34 ms per frame during typical running circumstances. This was done while keeping system RAM usage to an average of about 17 MB on a system that is also very tolerant to light fluctuations and noisy data.


Statistics for USQ ePrint 29207
Statistics for this ePrint Item
Item Type: USQ Project
Item Status: Live Archive
Additional Information: Bachelor of Engineering (Electrical and Electronics) project
Faculty / Department / School: Current - Faculty of Health, Engineering and Sciences - School of Mechanical and Electrical Engineering
Supervisors: Billingsley, John
Date Deposited: 06 Jun 2016 01:16
Last Modified: 06 Jun 2016 01:16
Uncontrolled Keywords: Machine Vision, Android, Image Processing, Row Detection
Fields of Research : 09 Engineering > 0906 Electrical and Electronic Engineering > 090601 Circuits and Systems
URI: http://eprints.usq.edu.au/id/eprint/29207

Actions (login required)

View Item Archive Repository Staff Only