Certain investigations on deep learning methods for image super resolution and disease classification in rice plants

Abstract

Globally, increasing population has raised the demand for agricultural crops. Food demand is expected to rise by 70% by 2050, according to the Food and Agricultural Organization of the United Nations (FAO). As a matter of fact, this gains a lot of attention in the research community as it focuses on smart agriculture and its subsidies. However, natural factors such as climatic change, water and land scarcity, environmental degradation, disease outbreaks, and pest infestations that are impending threats to food security. The quantity and quality of the food crops are hampered by the proliferation of diseases and pests, which cause famines, and shortfalls in crop yield, resulting in starvation and drastic economic loss. Identification of diseases and any disorder in the plants at the appropriate time is becoming a tedious task for farmers and experts. Traditionally, crop health is monitored by both farmers and pathologists through manual visual observation to identify diseases in plants. Manual diagnostic techniques always require highly skilled professionals to understand the signs and symptoms in order to provide suitable remedies, and recommendations. Moreover, the fast spreading nature of infections leads to potential damage to crops and plants. In order to address these challenges, the development of an early disease diagnostic system is essential to avoid the disease spreading from unhealthy to healthy plants newline