Chapter 12 - Advances in sensing plant diseases by imaging and machine learning methods for precision crop protection

Abstract

Plant diseases are one of the primary causes of major economic losses in the agriculture industry worldwide. The continuous monitoring of plant health and early detection of pathogens are crucial to reduce the disease spread and help effective disease management. The traditional methods of plant disease management generally rely on the spraying of chemical pesticides in the entire field, irrespective of its real requirement or not. Such blind application of these chemicals leads to undesirable effects on soil chemistry and microbiota. Following the second green revolution utilizing genomic advancements, smart or precision farming is changing the agricultural landscape at a very fast pace across the world. Precision agriculture relies on the implementation of modern-day advanced imaging and information technologies in disease identification. These intelligent and noninvasive methods use near real-time observations to protect crop damages caused by plant diseases. From a huge landscape of precision agriculture, the present chapter concentrates on the imaging-based approaches for biotic stress detection in plants. In this chapter, the machine learning methods including support vector machines, neural networks, and deep learning are also highlighted for the detection of plant diseases. These algorithms help in making smart decisions for the actual requirement and the adequate application of crop protection resources. Both imaging and machine learning methods are powerful and unparalleled tools for sustainable agriculture. They effectively detect biotic stress in plants and can provide data directly from different geographical scales.