Gamba Grass Management

Team Name

GreenNT

Team Members

Project Description

Gamba grass (Andropogon gayanus) was introduced into the Northern Territory as a pasture grass in the
1930s. Subsequently, plantings occurred throughout pastoral and agricultural areas of the Top End.
Since this time gamba grass has proved to be highly invasive, establishing itself in savanna woodlands,
riparian areas and vulnerable vegetation communities.
A combination of impacts on fire regimes, soil nutrient and soil water cycles are dramatically changing
these native plant communities. Wildlife communities and our environment are also at risk from habitat
loss.
Machine Learning methods already have demonstrated great potential in proving prediction and
computational power in agricultural data analysis. The next few years will undoubtedly see more
of this tools tailored and applied to agriculture.

The proposed model will works like this:
Step 1: Input Grass images
Step 2: pre-processing of selected data
Step 3: Segmentation of Images (Size, colour, texture)
Step 4: Extraction of Features (colours, size of grown up leaves)
Step 5: Identify the area for spreading chemical.
Step 6: Use google sprinkler for spreading chemical on specific area from remote access.
Step 7: Monitor Model and sprinkler monthly basis and use GPS tracker for tracking the device.

Data Story

The data gather from Northern territory Government (NTG) data portal .
The dataset we used was - Vegetation Sites Database (VSD) vegetation sites - survey conducted in a different timeframe by the by the Department of Environment, Parks and Water Security NT to gather the vegetation data from all part of NT.
We used National Map of Australian Government to extract the vegetation data from different sites.