25 October 2012 Performance of commercial and open source remote sensing/image processing software for land cover/use purposes
Author Affiliations +
Abstract
We aim to compare the potentialities of four remote sensing/image processing software: PCI Geomatica V8.2, ENVI 4.7, SPRING 5.1.8, and ORFEO toolbox integrated in Monteverdi 1.11. We listed and assessed the performance of several classification algorithms. PCI Geomatica and ENVI are commercial/proprietary software and SPRING and ORFEO are open source software. We listed the main classification algorithms available in these four software, and divided them by the different types/approaches of classification (e.g., pixel-based, object-oriented, and data mining algorithms). We classified using these algorithms two images covering the same area (Porto-Vila Nova de Gaia, Northern Portugal): one Landsat TM image from October 2011 and one IKONOS image from September 2005. We compared time of performance and classification results using the confusion matrix (overall accuracy) and Kappa statistics. The algorithms tested presented different classification results according to the software used. In Landsat image, differences are greater than IKONOS image. This work could be very important for other researchers as it provides a qualitative and quantitative analysis of different image processing algorithms available in commercial and open source software.
© (2012) COPYRIGHT Society of Photo-Optical Instrumentation Engineers (SPIE). Downloading of the abstract is permitted for personal use only.
Ana C. Teodoro, Ana C. Teodoro, Dário Ferreira, Dário Ferreira, Neftali Sillero, Neftali Sillero, } "Performance of commercial and open source remote sensing/image processing software for land cover/use purposes", Proc. SPIE 8538, Earth Resources and Environmental Remote Sensing/GIS Applications III, 85381K (25 October 2012); doi: 10.1117/12.974577; https://doi.org/10.1117/12.974577
PROCEEDINGS
12 PAGES


SHARE
Back to Top