A recent paper has been published and examines the use of Artificial Intelligence in 'just-in-time' crop disease detection in banana.

The paper reviews research towards the development of an AI-based banana disease and pest detection system using a deep convolutional neural networks (DCNN) to support banana farmers.

Banana is a popular fruit grown all over the world and a staple food in many developing countries. In many of these areas, traditional pest and disease identification approached rely on agricultural extension specialists; however, application of technologies such as Artificial Intelligence can simply and expedite this process.

The use of Smartphone-based AI apps could alert farmers and help prevent the possible outbreak of pests and diseases. The use of these technologies in pests and diseases is beginning to spread within the agricultural domain.

In order to develop test this innovation, the research team had a dataset of 18,000 images of banana, collected by banana experts, from Bioversity International (Africa) and Tamil Nadu Agricultural University (TNAU, Southern India). The database covers healthy plants (HP), dried/old age leaves (DOL) and a balanced number of images (700 images) from five major diseases.

Three different architectures were selected and trained usinga python deep learning library called TensorFlow and its object detection Application Programming Interface (API) to automatically detect diseases in the collected images.

Learn more in the article regarding:-
Banana dataset collection and annotation,
Performance metrics and validation of developed models,
And the Confusion matrix