Difference between revisions of "Tropical land cover assessment by vegetation sampling and description"

Latest revision as of 11:44, 8 March 2019

[edit] Macro Project Summary

This macro-project is intended to articulate research lines on visual information representation and characterization of soil usage and vegetation cover of Andean Forest and populated areas in the Central Colombia region.

Robust description of the native vegetation in this geographic context is approached by the definition of morphological (shape, area, boundaries), texture and color features over each plant's foliage. These features are learned directly from datasets specifically built for this application, and are used to feed physiological models of phenotypical expressions according to environmental conditions, beyond particularities related to interspecies differences and intra species specializations.

The dataset construction is, by itself, a critical task, so that protocols for data acquisition, sorting, annotation and navigation are stated.

[edit] Platform 1 - Acquisition and Storage: Comprehensive leaf dataset construction protocol

[edit] Platform 2 - Processing and Knowledge Extraction: Robust Colombian Vegetation feature representation - Retrieval System

[edit] Platform 3 - Access and Navigation: Colombian Vegetation Collection access and navigation platform for public policy design support

[edit] Goals

• Project 1: Phytopatologic diagnosis support in tomato crops (Boyacá/Santander) by color-based disease characterization on leaflets (Juan Felipe, in collaboration with UJTL)
• Project 2: Leaf shape characterization of
• Project 3:

Propose and validate an experimental protocol for

[edit] References

1. ...