University of Twente Proceedings

Login

Forest cover change analysis by object based method using spot and rapideye images

Share/Save/Bookmark

Gao, Yan and Gonzalez, Ignacio and Lopez-Sanchez, Jairo Gabriel and Skutsch, Margaret and Paneque-Galvez, Jaime and Mas, Jean Francois (2016) Forest cover change analysis by object based method using spot and rapideye images. In: GEOBIA 2016 : Solutions and Synergies., 14 September 2016 - 16 September 2016, University of Twente Faculty of Geo-Information and Earth Observation (ITC) .

[img] PDF
110kB
Event: GEOBIA 2016 : Solutions and Synergies., 14 September 2016 - 16 September 2016, University of Twente Faculty of Geo-Information and Earth Observation (ITC)
Abstract:In this paper we present forest cover change analysis by object based method with SPOT-5 (2007) and RapidEye (2013) images for an Ejido in the state of Jalisco, Mexico. We identified three classes in images of each date: 1) forest 2) degraded forest and 3) non-forest. An object based image analysis was applied to first segment the images into objects, and then classify the objects into the above three classes. We compared the results from this object based model with a model based on pixel based method. Classified images from both methods were evaluated with verification data composed of 254 random points and object based methods obtained slightly higher overall accuracy than the pixel based methods. Forest cover changes were analysed by constructing a model in DINAMICA (3.0.6) in which the forest classes of two dates were compared, and the forest cover changes were derived including deforestation, degradation, regeneration, and revegetation. The results show that although the overall accuracies of the classifications show no significance difference by McNemar’s test, except the classifications between MLC and MD for 2007, the obtained forest change results show big variance between adopted methods and it is rather difficult to compare them. The future study will apply test data for forest change classes and decide the best change results with the highest accuracies.
Item Type:Conference or Workshop Item (Paper)
Uncontrolled Keywords:Segmentation, Object-based image classification, SPOT-5, RapidEye, Forest degradation, Deforestation, Dinamica
Link to this item:https://doi.org/10.3990/2.377
Conference URL:https://www.geobia2016.com/
Export this item as:BibTeX
EndNote
HTML Citation
Reference Manager

 

Repository Staff Only: item control page