Remote sensing has been successfully used in research at the landscape scale and for large scale production agriculture; however, cost and low resolution has made this technology impractical for smaller producers and for field or plot-scale research. Now, recent advances in digital imaging technology mounted on unmanned aerial vehicles (UAVs) and balloons, and kites has made high resolution monitoring of agricultural production operations and research plots feasible and inexpensive (individual systems cost as low as $200). This technology has immediate application to agricultural practices that are data intensive or time sensitive, such as adaptive nutrient management. Additionally, these aerial low cost imaging platforms may reduce the labor, time, and equipment costs of biomass data collection, one of the primary factors limiting on-farm research. This tool aims to demonstrate the technology in the field, and refine the techniques for data acquisition and interpretation, including identification of crop plant and weed species spectral signatures, biomass, crop height, and percent cover , as well as identification of crop stress signatures.
UAVs balloons and Kites can be used to generate rapid and accurate crop plant performance datasets for on-farm management and field and plot-scale research. To calibrate and test the accuracy of remotely sensed data on-ground controls like spectrometer and SPAD meters http://en.wikipedia.org/wiki/Plant_stress_measurement for percent cover, biomass, spectral signatures, and other ground conditions, using the same on-the ground verification and calibration process that other remote sensing technologies use
The following organizations provide instructions and community support for open source aerial imaging
http://grassrootsmapping.org/
http://publiclaboratory.org/home
http://diydrones.com/
http://conservationdrones.org/
Public Laboratory's open source balloon hardware setup:
http://store.publiclaboratory.org/products/balloon-mapping-kit
Full construction and operation instructions and bill of materials for a fixed wing aircraft based can be found here http://conservationdrones.org/hardware/
Open Source Spectral analysis kit:
http://store.publiclaboratory.org/products/desktop-spectrometry-kit
Near Infrared Camera setups
http://publiclaboratory.org/tool/near-infrared-camera
http://publiclaboratory.org/tool/stereo-camera
Canon Powershots are often used because of the Canon Hack Developer Kit
Additional information on knitting images can be found here
http://publiclaboratory.org/notes/cfastie/10-29-2012/agricultural-mapping
Here is a Hypr3D model made with the lower elevation images http://www.hypr3d.com/models/4fbef44b7a8d8d0001000030 taken by the unmodified camera on the dual IR/visible camera rig. It only used 28 of the 149 images it was given. I think it would be good to include some of the images from the higher elevation flight and then it might be able to figure out where the other lower elevation images fit in.
The images were resampled to 1200 pixels across before submitting to Hypr3D just save bandwidth because and Hypr3d uses lower resolution versions anyway.
40 photos were added from higher elevation flights to the 149 used above and Hypr3D was able to include a total of 150 photos in a [new model which covers the entire experimental plot area: http://www.hypr3d.com/models/4fbf7a876f0c57000100005a. Photos were downsampled as above before submitting to Hypr3D. It takes a few hours to process the model.
A poster about interpretation of NDVI and visual spectrum analysis can be found here http://publiclaboratory.org/notes/cfastie/10-29-2012/agricultural-mapping
Spectrometer http://store.publiclaboratory.org/products/desktop-spectrometry-kit
Balloon Kit http://store.publiclaboratory.org/products/balloon-mapping-kit
DIY Drone fixed wing http://code.google.com/p/ardupilot-mega/wiki/Get
DIY Drone rotary wing http://code.google.com/p/arducopter/wiki/Purchase