"Weed Mapping in Maize Fields Using Object-Based Analysis of UAV Images"

Cool article from the open access scientific journal PLOS One

Abstract

The use of remote imagery captured by unmanned aerial vehicles (UAV) has tremendous potential for designing detailed site-specific weed control treatments in early post-emergence, which have not possible previously with conventional airborne or satellite images. A robust and entirely automatic object-based image analysis (OBIA) procedure was developed on a series of UAV images using a six-band multispectral camera (visible and near-infrared range) with the ultimate objective of generating a weed map in an experimental maize field in Spain. The OBIA procedure combines several contextual, hierarchical and object-based features and consists of three consecutive phases: 1) classification of crop rows by application of a dynamic and auto-adaptive classification approach, 2) discrimination of crops and weeds on the basis of their relative positions with reference to the crop rows, and 3) generation of a weed infestation map in a grid structure. The estimation of weed coverage from the image analysis yielded satisfactory results. The relationship of estimated versus observed weed densities had a coefficient of determination of r2=0.89 and a root mean square error of 0.02. A map of three categories of weed coverage was produced with 86% of overall accuracy. In the experimental field, the area free of weeds was 23%, and the area with low weed coverage (<5% weeds) was 47%, which indicated a high potential for reducing herbicide application or other weed operations. The OBIA procedure computes multiple data and statistics derived from the classification outputs, which permits calculation of herbicide requirements and estimation of the overall cost of weed management operations in advance.

Read the resthere

Views: 3453

Comment by Joshua Johnson on October 16, 2013 at 8:43pm

Their are so many amazing benefits from crop observation using drones.  It's not only better for the earth because less pesticides are used but it also saves the farmer money from the use of less pesticides and helicopter rental fee's (Which are extremely expensive).  I can't wait to see what 3D Robotic's has in store for the drone community when you start releasing consumer based agricultural drones, Infrared Cameras, etc. 

+1 For agricultural drones!

Comment by LanMark on October 16, 2013 at 9:37pm

pretty cool.  Looks like Tetracam imagery.


Moderator
Comment by Gary Mortimer on October 16, 2013 at 10:46pm

As soon as NVIR cameras become cheap and widely available there will be an explosion in this as a bought in service.

Comment by Andrew Rabbitt on October 17, 2013 at 12:03am

I agree up to a point Gary, but I think a better model is to keep the whole thing in the amateur domain to bypass the slow and cumbersome regulatory reform processes.  Farmers interested in this technology will be already investing hundreds of thousands in machinery, so what's a few grand for a "toy plane" to map their own fields at their leisure?

Comment by koen.hufkens on October 17, 2013 at 1:22am

For some applications you don't even need the spectral data. A lot can be derived from texture alone (this goes for forests as well as grasslands), or just feeding an NDVI image into an unsupervised method. I attached a little proof of concept run I did earlier on the field I fly on.

http://74.220.219.112/~khufkens/2013/08/29/uav-vegetation-monitoring/

This method has been used in a high resolution satellite remote sensing context, but it holds it's own in a UAV context as well. A big problem with these images is often view angle effects, a lot of work has to be done on this part. However, for general weed whacking this will do :)


Moderator
Comment by Gary Mortimer on October 17, 2013 at 3:02am

Thanks Koen I shall read up, @Andrew I don't know many farmers around here that would bother learning the skills. They would rather I think buy the service in. The next hard part will be getting paid by them ;-) 

Comment by koen.hufkens on October 17, 2013 at 3:04am

@Gary, If you don't have access to pay-walled scientific literature let me know. I can send you the original articles.


Developer
Comment by Mark Colwell on October 17, 2013 at 3:13am

Koen, I am interested in your research too. Can you send me a copy please? I am designing  a nadir mount for Sony NEX-5 cameras. I started Remote Sensing study last fall at local community college, Hope to get a GIS certification soon. Thanks

Comment by koen.hufkens on October 17, 2013 at 3:16am

@Mark, It's not really my research / topic yes, the methodology no. I'll send you a copy of the literature...

Comment by Dries Raymaekers on October 17, 2013 at 3:38am

Koen, I agree on the textural analysis which at this moment is not yet fully explored and will help in the data reduction efforts. I see you are from Gent, if interested, pass-by on the Beo-day in November, an event for the remote sensing community in Belgium. There will be a session specifically on UAV (demo - data analysis -legislation etc..).

http://eo.belspo.be/Docs/DocExt/BEOday-2013-Draft-Programme_small.pdf

Comment

You need to be a member of DIY Drones to add comments!

Join DIY Drones

© 2019   Created by Chris Anderson.   Powered by

Badges  |  Report an Issue  |  Terms of Service