The use of random forest and object-based image classification for thematic map creation: A case study of Kelantan, Malaysia

Publication Type:
Conference Proceeding
Citation:
38th Asian Conference on Remote Sensing - Space Applications: Touching Human Lives, ACRS 2017, 2017, 2017-October
Issue Date:
2017-01-01
Full metadata record
Files in This Item:
Filename Description Size
The Use of Random Forest and Object Based Image Classification.pdfPublished version1.17 MB
Adobe PDF
© 2017 ACRS. All rights reserved. This paper presented results of creation of thematic map for Kelantan city using an integration of object-based analysis and random forest algorithm. The thematic map was generated based on SPOT image with a spatial resolution of 5 m. Object-based analysis was used for its advantages over per-pixel based methods. On the other hand, random forest algorithm was selected because it was suggested in very recent literature and it has advantages ofgood generalization.A thematic map with sixclasses was generated with an overall accuracy and kappa index of 92% and 0.90, respectively using the proposed method. Other details are included in the paper.
Please use this identifier to cite or link to this item: