Using Deep Learning for Image-Based Plant Disease Detection

Motivated by the threat crop diseases pose to global food security and the difficulty of diagnosing them where infrastructure is scarce, the authors explore smartphone-assisted diagnosis enabled by growing smartphone access and advances in deep-learning computer vision. They train a deep convolutional neural network on a public dataset of 54,306 images of diseased and healthy plant leaves collected under controlled conditions, teaching it to identify 14 crop species and 26 diseases or their absence.

The trained model achieves 99.35% accuracy on a held-out test set, demonstrating that automated image-based plant disease detection is feasible. The authors argue that training deep-learning models on increasingly large, publicly available image datasets charts a clear path toward smartphone-assisted crop disease diagnosis on a massive global scale.