Abstract
High-resolution aerial imaging with an unmanned aerial vehicle (UAV) was used to quantify wheat powdery mildew and estimate grain yield. Aerial digital images were acquired at Feekes growth stage (GS) 10.5.4 from flight altitudes of 200, 300, and 400 m during the 2009-10 and 2010-11 seasons; and 50, 100, 200, and 300 m during the 2011-12, 2012-13, and 2013-14 seasons. The image parameter lgR was consistently correlated positively with wheat powdery mildew severity and negatively with wheat grain yield for all combinations of flight altitude and year. Fitting the data with random coefficient regression models showed that the exact relationship of lgR with disease severity and grain yield varied considerably from year to year and to a lesser extent with flight altitude within the same year. The present results raise an important question about the consistency of using remote imaging information to estimate disease severity and grain yield. Further research is needed to understand the nature of interyear variability in the relationship of remote imaging data with disease or grain yield. Only then can we determine how the remote imaging tool can be used in commercial agriculture.
Highlights
2010–11 seasons; and 50, 100, 200, and 300 m during the 2011–12, 2012–13, and 2013–14 seasons
–0.87 to –0.42, respectively. (R–G)/(G+R) was positively correlated with disease index, while S did not show a consistent correlation with and ‘à’ represent the years 2010, 2011, 2012, 2013, and 2014, respectively; R, G, B, I, H, S means the value of red, green, blue, intensity, hue, and saturation acquired from aerial photography, respectively; lgR, lgG, lgB, and lgI means the logarithms of the value of red, green, blue, and intensity, respectively; G/R, G–R, and (G–R)/(G+R) are composite variables that combine the values of red and green
Unmanned aerial digital images were acquired at Feekes growth stage (GS) 10.5.4 from 50, 100, 200, 300, and 400 m above the ground in five consecutive seasons
Summary
2010–11 seasons; and 50, 100, 200, and 300 m during the 2011–12, 2012–13, and 2013–14 seasons. Airborne sensor data have been widely used to survey disease development, such as cereal rust (Colwell 1956), bacterial blight of field beans (Wallen and Jackson 1971), cotton root rot (Toler et al 1981), and spot blotch of barley (Clark et al 1981). These studies depended mainly on photographic film due to the limitation of technology at the time, which made image processing and information extraction complicated. Image analysis of aerial digital photographs has appeared in the agronomic literature in the last 20 years. Everitt et al (1999) used airborne digital imagery to detect oak wilt disease. Martins et al (2001)
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
