Identification of the spectral signature of noxious weed St. John’s wort and development of a tracking methodology via remote sensing

Collins, Matthew (2016) Identification of the spectral signature of noxious weed St. John’s wort and development of a tracking methodology via remote sensing. [USQ Project]

[img]
Preview
Text (Main Project)
Collins_M_Apan.pdf

Download (8Mb) | Preview

Abstract

Through geographical and evolutionary isolation, Australia is highly susceptible to invasive plants including noxious weeds. Significant environmental and economic impacts created by the presence of noxious weeds became the driving factors in targeting effective and efficient detection and tracking methodologies in this research project.

This project aims to determine the spectral signature of the declared noxious weed St. John’s wort (hypericum perforatum L.) through the application of remote sensing techniques to satellite imagery incorporating known locations. The intention here is to enhance practices undertaken by local control authorities and other stakeholders by enabling effective remote detection and tracking capabilities. Specific project objectives include acquisition of freely available, regularly captured remotely sensed data, utilisation of freely available geographic information system / remote sensing software packages and techniques, mapping existing occurrences of St. John’s wort throughout the Bega Valley Shire local government area, and investigation of the potential to use the techniques adopted in St. John’s wort identification and tracking programs.

The project methodology included identification of background pertinent to noxious weeds (specifically St. John’s wort), review of remote sensing techniques, identification of the spectral signature of St. John’s wort through compilation of a list of known locations and application of remote sensing land cover classification techniques, and review and discussion of the potential to develop identification and tracking programs.

Results obtained through application of the project methodology ranged in accuracy through different seasonal points of data capture. Key project limitations contributing to the accuracy of results include data resolution and site accessibility. Privacy implications removed the potential to ground truth results during the analysis phase, and the relatively coarse resolution of the freely available satellite imagery used had a correlation to accuracy of results obtained via the land cover classification technique utilised. Conclusions drawn reiterate the implications of data and site access limitations inherent to the project, as do the recommendations for further study.


Statistics for USQ ePrint 31388
Statistics for this ePrint Item
Item Type: USQ Project
Item Status: Live Archive
Additional Information: Bachelor of Spatial Science (Honours) Major Surveying project
Faculty / Department / School: Current - Faculty of Health, Engineering and Sciences - School of Civil Engineering and Surveying
Supervisors: Apan, Armando
Date Deposited: 20 Jul 2017 00:52
Last Modified: 20 Jul 2017 00:52
Uncontrolled Keywords: noxious weed; tracking methodology; remote sensing
Fields of Research : 09 Engineering > 0909 Geomatic Engineering > 090906 Surveying (incl. Hydrographic Surveying)
URI: http://eprints.usq.edu.au/id/eprint/31388

Actions (login required)

View Item Archive Repository Staff Only