Precision agriculture

One of our most recent projects involved the monitoring of nutrition health in macadamia orchards using imagery obtained from an unmanned aerial vehicle (UAV) equipped with an autopilot, GPS and digital camera.  Our study investigated the feasibility of using UAVs for this purpose with very promising results.

The UAV was used to capture high resolution, geo-referenced aerial photographs to assess the canopy health of macadamia trees.  It was found that the imagery could be analysed to investigate nitrogen levels and insect damage, and that nut variety and canopy yellowing could be clearly determined from the images.

          (a)          (b)

(a) Image showing vegetation index results with different colours indicating variation in canopy health.

(b) Vegetation index values grouped to prepare a tree-health map suitable for a fertilizer spreader.

The data obtained from the imagery was compared against ground-truth data with excellent results.  The data was able to be imported into farm management software, and then loaded into a variable rate fertilizer spreader.

Tree-health map imported into software to allocate fertilizer treatments according to individual tree needs.

Due to changing CASA regulations we were unable to complete the final year of flights for this study, so instead compared the 2009 data to 2010 data obtained via the Worldview-2 satellite.  The imagery was ‘pre-processed’ into an image suitable for analysis, and the process then repeated as per the UAV data.  Both sets of data were compared to determine which was more suitable for our client’s requirements.

A classified CCCI image from the WorldView2 imagery is shown below:

 

Tree-health map from WorldView-2 imagery imported into variable rate fertilizer setup.