Improving the potential of pixel-based supervised classification in the absence of quality ground truth data

T he accuracy of classified results is often measured in comparison with reference or “ground truth” information. However, in inaccessible or remote natural areas, sufficient ground truth data may not be cost-effectively acquirable. In such cases investigative measures towards the optimisation of the classification process may be required. The goal of this paper was to describe the impact of various parameters when applying a supervised Maximum Likelihood Classifier (MLC) to SPOT 5 image analysis in a remote savanna biome. Pair separation indicators and probability thresholds were used to analyse the effect of training area size and heterogeneity as well as band combinations and the use of vegetation indices. It was found that adding probability thresholds to the classification may provide a measure of suitability regarding training area characteristics and band combinations. The analysis illustrated that finding a balance between training area size and heterogeneity may be fundamental to achieving an optimum classified result. Furthermore, results indicated that the addition of vegetation index values introduced as additional image bands could potentially improve classified products and that threshold outcomes could be used to illustrate confidence levels when mapping classified results.

• Publication year

  2015

• Venue

  South African Journal of Geomatics

• Publication date

  2015-11-08

• Fields of study

  Geography, Computer Science, Environmental Science

• Identifiers
  DOI 10.4314/SAJG.V4I3.6
• External record

  Open on Semantic Scholar

• Source metadata

  Semantic Scholar

• No claims are published for this paper.

• No concepts are published for this paper.

• Remote Sensing And Image Interpretation

  2016influential reference
• Closing the “Uncertainty Chain”: Enhancing Trust by Communicating Uncertainty Information in Maps

  2013cited by this paper
• ESTIMATION OF FOREST STRUCTURAL ATTRIBUTES USING ASTER DATA

  2013cited by this paper
• Comparison of ETM+ and MODIS Data for Tropical Forest Degradation Monitoring in the Peninsular Malaysia

  2013cited by this paper
• Classification of Multispectral satellite images

  2013cited by this paper
• Beyond the Surface: Current Issues and Future Directions in Uncertainty Visualization Research

  2013cited by this paper
• Optimizing land cover classification accuracy for change detection, a combined pixel-based and object-based approach in a mountainous area in Mexico

  2012cited by this paper
• EVALUATING THE NOVEL METHODS ON SPECIES DISTRIBUTION MODELING IN COMPLEX FOREST

  2012cited by this paper
• Soil-line vegetation indices for corn nitrogen content prediction

  2012cited by this paper
• A Review of Uncertainty in Data Visualization

  2012cited by this paper
• Comparison of pixel- and object-based classification in land cover change mapping

  2011cited by this paper
• A comparison of pixel-based and object-oriented approaches to VHR imagery for mapping saltmarsh plants

  2011cited by this paper
• Sources of variability in canopy reflectance and the convergent properties of plants.

  2011cited by this paper
• Is there a temporal niche separation in the leaf phenology of savanna trees and grasses?

  2011cited by this paper
• A fuzzy topology-based maximum likelihood classification

  2011cited by this paper
• Characterizing vegetation cover in global savannas with an annual foliage clumping index derived from the MODIS BRDF product

  2011cited by this paper
• Mapping understory vegetation using phenological characteristics derived from remotely sensed data

  2010cited by this paper
• Remote Sensing and GIS as an Advance Space Technologies for Rare Vegetation Monitoring in Gobustan State National Park, Azerbaijan

  2010cited by this paper
• Comparison of geo-object based and pixel-based change detection of riparian environments using high spatial resolution multi-spectral imagery.

  2010cited by this paper
• Exploring uncertainty in remotely sensed data with parallel coordinate plots

  2009cited by this paper
• ANIMAL HABITAT QUALITY AND ECOSYSTEM FUNCTIONING: EXPLORING SEASONAL PATTERNS USING NDVI

  2008cited by this paper
• Integration of Landsat TM and SPOT HRG Images for Vegetation Change Detection in the Brazilian Amazon.

  2008cited by this paper
• Large-area crop mapping using time-series MODIS 250 m NDVI data: An assessment for the U.S. Central Great Plains

  2008cited by this paper
• Comparisons of Object-Oriented and Pixel-Based Classification of Land Use/Land Cover Types Based on Lansadsat7, Etm + Spectral Bands (Case Study: Arid Region of Iran)

  2007cited by this paper
• Comparison of Vegetation Indices and Red‐edge Parameters for Estimating Grassland Cover from Canopy Reflectance Data

  2007cited by this paper
• Evaluation of feature extraction and reduction methods for hyperspectral images

  2007cited by this paper
• A ground‐validated NDVI dataset for monitoring vegetation dynamics and mapping phenology in Fennoscandia and the Kola peninsula

  2007cited by this paper
• Remote Sensing of Landscapes with Spectral Images

  2006influential reference
• USING NDVI AS A PASTURE MANAGEMENT TOOL

  2006cited by this paper
• OPTIMAL BAND SELECTION OF HYPERSPECTRAL DATA FOR TRANSGENIC CORN IDENTIFICATION

  2006cited by this paper
• Comparing MODIS vegetation indices with AVHRR NDVI for monitoring the forage quantity and quality in Inner Mongolia grassland, China

  2005cited by this paper
• A parallelepiped multispectral image classifier using genetic algorithms

  2005cited by this paper
• Land cover update by supervised classification of segmented ASTER images

  2005cited by this paper
• A COMPARISON OF OBJECT-ORIENTED AND PIXEL-BASED CLASSIFICATION METHODS FOR MAPPING LAND COVER IN NORTHERN AUSTRALIA.

  2005cited by this paper
• An assessment of the effectiveness of decision tree methods for land cover classification

  2003cited by this paper
• Remote sensing for biodiversity science and conservation

  2003cited by this paper
• Image classification of a northern peatland complex using spectral and plant community data

  2003cited by this paper
• Modelling and visualizing multiple spatial uncertainties

  1997cited by this paper
• A Modified Soil Adjusted Vegetation Index

  1994influential reference
• Potentials and limits of vegetation indices for LAI and APAR assessment

  1991cited by this paper
• A soil-adjusted vegetation index (SAVI)

  1988cited by this paper
• The factor of scale in remote sensing

  1987cited by this paper

• Detecting and distinguishing between apicultural plants using UAV multispectral imaging

  2023cites this paper
• Assessing impact of agroecological interventions in Niger through remotely sensed changes in vegetation

  2023cites this paper
• Evaluation of Different Satellite Images for Urban Land Use Analysis: A Case Study of Greater Manchester, United Kingdom

  2022cites this paper
• Inter‐Group Social Behavior, Contact Patterns and Risk for Pathogen Transmission in Cape Buffalo Populations

  2021cites this paper
• Fifteen-year resilience against further degradation of Succulent Karoo vegetation in South Africa

  2020cites this paper
• Assessing the Impacts of Climate Change on Natural Resources in Erbil Area, the Iraqi Kurdistan Using Geo-Information and Landsat Data

  2019cites this paper
• Effects of land use changes on water quality in Eerste River, South Africa

  2016cites this paper