Digital Twin in Agriculture Sector

Abstract

In the field of agriculture, pets and diseases cause enormous economic damage to farmers due to reduced yields and high costs of pesticides. The disease can be chronic, acute, and can also accelerate. It can be expressed in many methods. The symptoms include leaf curling, morphological changes, chlorosis, premature abscission of plants, change in leaf angle, stunting, and wilting. Rapid and accurate quantification and detection of early symptoms are usually tricky. Hence, for this, the various recent sensor networks and their usability techniques have shown a high potential in diagnosing diseases and in-plant monitoring for areas with infected plants. These techniques attract significant attention in the agricultural field, where disease detection is the most critical work. The occurrence of plant diseases depends on specified epidemiological and environmental factors, and thus they often have an uneven field distribution. This book chapter provides a comprehensive overview of traditional methods, the latest trends, and advances in crop disease detection using deep learning techniques in integration with sensors. It also describes the study of recent advances in the use of optical sensors for the detection, recognition, and evaluation of diseases in the crop. Multi-spectral, optical, gravimetric, conductivity, hyperspectral, and thermography sensors are the most used detection sensors to analyze different parameters and predict the cause of disease in crops. Further, the imaging systems are used for the diagnosis and monitoring of plant diseases, as the conventional methods are considered as the “gold standard” in the detection of disease, which depends on culture-based symptoms, biochemical identifications, and morphological observation.