Skip to main content
ARS Home » Research » Publications at this Location » Publication #130679
Title: DESIGN OF AN OPTICAL WEED SENSOR USING PLANT SPECTRAL CHARACTERISTICS

Author
item WANG, NING - KSU, MANHATTAN, KS
item ZHANG, N - KSU, MANHATTAN, KS
item Dowell, Floyd
item SUN, Y - KSU, MANHATTAN, KS
item PETERSON, D - KSU, MANHATTAN, KS

Submitted to: Transactions of the ASAE
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 9/1/2000
Publication Date: 3/1/2001
Citation: Wang, N., Zhang, N., Dowell, F.E., Sun, Y., Peterson, D.E. 2001. Design of an optical weed sensor using plant spectral characteristics. Transactions of the ASAE. 2001. 44(2):409-419.

Interpretive Summary: Traditional approaches to herbicide applications are based on the assumption that weeds are distributed uniformly in fields. However, most agricultural fields are spatially variable in weed infestation. Herbicide usage can be significantly reduced if the weeds can be sensed, then herbicides applied only to weeds. The spectral characteristics of stems and leaves of various crop and weed species were studied using a near- infrared spectrometer. Selected wavelengths were used to design an optical weed sensor. Laboratory tests showed that the sensor identified wheat, bare soil, and weeds with classification rates approaching 100%. These results will be used to develop a tractor-mounted sensor for field use. Farmers will benefit from this technology through reduced herbicide costs, and the impact of chemicals on our environment will be reduced through reduced applications of chemicals.

Technical Abstract: Spectral characteristics of stems and leaves of various crop and weed species were studied using a diode-array spectrometer. Five feature wavelengths were selected to form color indices as input variables to a classification model for weed detection. The feature wavelengths also served as the basis for design of an optical weed sensor. Based on experimental data, color indices insensitive to illumination variations were designed and tested on the sensor. Laboratory tests showed that the sensor identified wheat, bare soil, and weeds (several species combined) with classification rates of 100%, 100% and 71.6%, respectively, for the training data set when the weed density was above 0.01 plants/cm squared. The classification rates for the validation data set were 73.8%, 100%, and 69.9%, respectively. When the density of weeds was low, as in the case of a single weed plant, more than 50% of the weeds were misclassified as soil. .Misclassifications between wheat and weeds were not observed at any weed and wheat densities tested.