Automated Phenotyping for Disease Measurement in Cassava Plants Using Smartphones

Cassava is an important security crop in most of Africa. Crop diseases are one of the leading causes of low yields in cassava particularly amongst small-holder farmers. Disease identification is commonly done by experts. In this work I discuss some methods we are using to speed up diagnosis and measurement of disease by automating the phenotyping of cassava crops. We do this using computer vision techniques implemented on smartphones. I will discuss the measurement of disease from symptoms that are manifest on the leaves and the root cross-sections of cassava plants.