Evaluation of photo imaging methods for vegetation condition assessment

Armando Apan, Govinda Baral, Ernest Dunwoody, Lucy Richardson, Kevin McDougall

Research output: Chapter in Book/Report/Conference proceedingConference PaperOtherpeer-review

Abstract

The aim of this project was to assess the feasibility of a photo imaging approach in the assessment of vegetation condition attributes in comparison with the Queensland Department of Environment and Resource Management (QDERM) Vegetation Condition Assessment approach. The project applied both the QDERM BioCondition monitoring method and an alternative photo imaging analysis approach to the assessment of vegetation condition attributes on different regional ecosystems. The study focused on the comparability of results and the cost-effectiveness of the photo imaging approach in comparison to the standard BioCondition assessment method.

Six regional ecosystems (RE) sites, of varying structural complexity, were selected for this study. They contained a diverse range of different vegetation attributes. Each site was assessed using both approaches. A low-cost Canon PowerShot SX10 IS camera was used to collect the photographs. It is equipped with a 20x Optical Zoom lens with a focal length of 5.0-100mm that allows shooting a scene from wide-angle to telephoto. For the ground cover data collection, two vertical-down photographs were taken of each quadrat. To capture canopy cover, vertical-up photographs were taken at 5m intervals along the transect midline. A two pole photographic method was developed to estimate tree canopy height.

This study found that the use of photo imaging methods to measure most attributes of vegetation for the BioCondition approach is technically possible. However, their application for operational use in ecosystems with closed vegetation canopies is not feasible. The estimation of the vegetation condition variables is constrained, in various degrees, by several factors. The estimation of shrub species richness in grassland and open canopy forests are the only attributes that have potential for operational use. Canopy cover estimates from vertical-up photographs produced comparable Tree Cover Rank results compared to the manually based crown cover estimate method. The photographic technique also has good potential for estimating major classes of ground cover in quadrats. Canopy height can be estimated more easily by using a laser range finder than a photo imaging method. In the future, when the cost of data acquisition becomes less expensive, the suitability of a LiDAR system could be considered to quantify the desired vegetation attributes.
Original languageEnglish
Title of host publicationProceedings of the Surveying and Spatial Sciences Conference 21-25 November 2011, Wellington, New Zealand
PublisherScion Publishing
Pages243-258
Number of pages16
ISBN (Electronic)9780478110421
Publication statusPublished - 2011
Externally publishedYes
Event2011 Surveying and Spatial Sciences Conference: Innovation in Action: Working Smarter - Wellington, New Zealand
Duration: 21 Nov 201125 Nov 2011

Conference

Conference2011 Surveying and Spatial Sciences Conference: Innovation in Action: Working Smarter
CountryNew Zealand
CityWellington
Period21/11/1125/11/11

Keywords

  • photo imaging
  • biocondition
  • vegetation assessment
  • low-cost camera

Cite this