Skip to main navigation menu Skip to main content Skip to site footer

Interpretability in the Field of Plant Disease Detection: A Review


The early detection of diseases in plants through artificial intelligence techniques has been a very important technological advance for agriculture since, through machine learning and optimization algorithms, it has been possible to increase the yield of various crops in several countries around the world. Different researchers have focused their efforts on developing models that allow supporting the task of detecting diseases in plants as a solution to the traditional techniques used by farmers. In this systematic literature review, an analysis of the most relevant articles is presented, in which image processing techniques and machine learning were used to detect diseases by means of images of the leaves of different crops. In turn, an analysis of the interpretability and precision of these methods is carried out, considering each phase of the image processing, segmentation, feature extraction and learning processes of each model. In this way, there is evidence of a void in the field of interpretability since the authors have focused mainly on obtaining good results in their models, beyond providing the user with a clear explanation of the characteristics of the model.


Machine Learning, Classification, Early detection of diseases, Interpretability, Image processing


Author Biography

Daniel-David Leal-Lara

Roles: Conceptualization, Research, Content - Curation, Methodology, Writing - Original draft, Writing – Review & editing.

Julio Barón-Velandia

Roles: Conceptualization, Methodology, Writing - Original draft, Writing – Review & editing.

Camilo-Enrique Rocha-Calderón

Roles: Validation, Writing – Review & editing.



Download data is not yet available.