Application of machine learning: a recent advancement in plant diseases detection

Abstract

The world population, and thus the need for food, is increasing every day. This leads to the ultimate question of how to increase food production with limited time and scarce land. Another obstacle to meet the food demand includes the stresses a plant goes through. These may be abiotic or biotic, but the majority are biotic, i.e., plant diseases. The major challenge is to mitigate plant diseases efficiently, more quickly and with less manpower. Recently, artificial intelligence has turned to new frontiers in smart agricultural science. One novel approach in plant science is to detect and diagnose plant disease through deep learning and hyperspectral imaging. This smart technique is very advantageous for monitoring large acres of field where the availability of manpower is a major drawback. Early identification of plant diseases can be achieved through machine learning approaches. Advanced machine learning not only detects diseases but also helps to discover gene regulatory networks and select the genomic sequence to develop resistance in crop species and to mark pathogen effectors. In this review, new advancements in plant science through machine learning approaches have been discussed.

Keywords

• diagnosis
• disease
• hyperspectral imaging
• machine learning