Abstract: Highlights•Analysis of the transfer learned models on rice leaf disease identification using 10 state-of-the-art DCNN models.•Development of ML-based identification models on deep features extracted from the aforementioned DCNN models.•Attention-driven preprocessing of rice-leaf disease dataset using Dynamic Mode Decomposition based ROI extraction algorithm.•Analysis of the impact of DMD preprocessed images on transfer learned DCNN classifiers and ML models.•The effectiveness of DMD pre-processed images on on-field rice leaf images was evaluated using transfer-learned DCNN models.
Loading