International Journal of Science, Engineering and TechnologyAuthors: Rinku Badgujar, Sambhav Kothari, Dhiraj Tapkir, Abhijeet Prasad
Abstract: The agricultural industry is increasingly adopting Artificial Intelligence (AI) solutions to tackle the challenges of early disease identification and crop management. While previous approaches have primarily focused on single-crop disease classification, this paper presents an enhanced version of the Plant Pulse system— expanding its capabilities beyond apples to include a wide variety of crops such as cherry, corn, grape, orange, peach, pepper, potato, strawberry, and tomato. The proposed system employs a custom-designed Convolutional Neural Network (CNN) trained on over 61,000 images spanning 39 plant disease and healthy categories. Developed using PyTorch, the system demonstrates high performance, achieving 98.9% accuracy on the test set. The architecture is optimized for both accuracy and scalability, supporting real-time inference and future integration with field-deployable tools. This research builds upon our prior work [1], significantly extending its scope and applicability across diverse agricultural domains.
DOI: http://doi.org/
