No abstract

Human diagnosis of horticultural diseases comes with added monetary costs in the shape of time, cost, and acces-sibility, with still considerable possibilities of misdiagnosis. Most common plant diseases present visually recognizable symptoms like change in color, shape, or texture. Deep learning is known to work with such accuracy and precision in recognizing patterns in such visual symptoms that rivals human diagnosis. We specifically designed a deep learning–based multi-class classification model AppleNet to include extra apple plant diseases, which has not been the case with other previously designed models. Our model takes advantage of transfer learning techniques by implementing ResNET 50 Convolutional Neural Network pretrained on image-net dataset. The knowledge of features learned by ResNET 50 is being used to extract features from our dataset. This technique takes advantage of knowledge learned on a larger and more diverse dataset and also saves precious computational resources and time in training on a relatively lesser data. The hyper-parameters were uniquely fine-tuned to maximize the model efficiency. We created our own dataset from the images taken directly from the trees, which, unlike the publicly available datasets created in a controlled setting with smooth (white) background, has been created in a real world environment and includes background noise as well. This helped us train our model in a more realistic way. The results of experimentation on a collected dataset of 2897 images with data augmentation demonstrated that AppleNet can be efficiently used for apple disease detection with a classification accuracy of 96.00%. To examine the effectiveness of our proposed approach, we compared our model with other pretrained models and a baseline model created from scratch. Results of the experiment demonstrate that transfer learning improves the performance of deep learning models and using pretrained models based on residual neural network architectures gives remarkable results as compared to other pretrained models. The mean difference in classification accuracies between our proposed model AppleNet and other experimental models was 21.54%.