The possibility of detecting dead plant areas in konjac using an unmanned aerial vehicle (UAV) was investigated. The intensity spectrum of the dead plant area had a unique range. Aerial images were binarized to detect dead plant areas. The dead plant rate estimated from aerial images was compared with the dead plant rate observed. Therefore, there is a significant relationship between the actual dead plant rate and the rate from aerial images (R2 = 0.829, p < 0.01). This result proposes that the dead plant rate obtained from aerial images using UAV might allow us to predict the actual dead plant rate.
