Fig. 11

Sample output after applying (2 g − r–b) color indices with the NN approach (Li et al. [19]) on our validation dataset consists of randomly selected control as well as drought plants. The circle shows the wrongly classified spike pixels

Sample output after applying (2 g − r–b) color indices with the NN approach (Li et al. [19]) on our validation dataset consists of randomly selected control as well as drought plants. The circle shows the wrongly classified spike pixels