On the Influence of Markovian Models for Contextual-Based Optimum-Path Forest Classification
Abstract
Contextual classification considers the information about a sample's neighborhood to improve standard pixel-based classification approaches. In this work, we evaluated four different Markovian models for Optimum-Path Forest contextual classification considering land use recognition in remote sensing data. Some insights about the situations in which each of them should be applied are stated, as well as the idea behind them is explained.
How to cite this document
Osaku, D. et al. On the Influence of Markovian Models for Contextual-Based Optimum-Path Forest Classification. Progress In Pattern Recognition Image Analysis, Computer Vision, And Applications, Ciarp 2014. Berlin: Springer-verlag Berlin, v. 8827, p. 462-469, 2014. Available at: <http://hdl.handle.net/11449/184745>.
Language
English
Sponsor
Collections
