Noxious weeds, characterized by their aggressive behavior, continue to be an increasing problem throughout the United States. Early detection is critical in order to contain and/or eradicate an infestation.
Ground-based surveys for weeds require time and money to cover a relatively small area. The data collected from ground-based surveys may only provide a limited amount of information regarding the extent and spread of any target weed. Remote sensing, using an unmanned aerial vehicle (UAV), can provide a more efficient method to map and monitor an area for noxious weeds.
UAVs cover a large area in a relatively short amount of time while providing high spatial resolution, which can allow for the detection of small objects, including weeds. Due to weight constraints, UAVs use light-weight sensors and especially four- to six-band multispectral cameras to take images in the visible and the near infra-red (NIR).
Although aerial surveys offer a more efficient method of mapping infestations, there are many weeds that may not be good candidates for remote sensing. Weeds may not be distinguishable from their surroundings, making it difficult to locate in an aerial image.
Research is being done to overcome this by using multispectral cameras which work by imaging different wavelengths of light. Plants typically reflect a large amount of near-infrared (NIR) light, which is not visible to the human eye, but is visible to multispectral cameras. These cameras have been utilized to identify areas of stress in a crop, and even provide a quantitative metric for the vigor of a plant.
ADVERTISEMENT
The Minnesota Department of Agriculture and the University of Minnesota are working together to develop an effective method to aerially monitor Palmer amaranth using a multispectral camera.
