A Survey on the Detection of Diseases in Plants Using the Computer Vision-Based Model

Abstract

The Agriculture sector acts as the backbone of our country. It has not just remained as simply to feed the increasing population. Rapid growth of the population day by day has resulted in a high demand of agricultural yields. Agriculture sector is struggling to cater to the needs of ever-growing demand of the population. India’s economy greatly depends on the productivity of the agricultural yields. Due to irregular rainfall, drought or different climatic conditions, most of the time agricultural crops suffer from a variety of plant diseases. These diseases affect not only the production quantity but also the quality of the crop yield. This results in catastrophic or chronic losses, due to which Indian farmers suffer to recover from bank loans and finally attempt suicide. To avoid such situations, a means has to be discovered to appropriately identify the diseases affecting the plants in the early stage. Diagnosing diseases incorrectly might lead to improper chemical sprayed on the plantations, which results in ill impacts on the environment, wastage of costs with a significant loss in economy. The present diagnosis of disease is based on human vision, and it is a time-consuming process and expensive. Making use of a computer vision-based model to detect the diseases affecting the plants would help in increasing the accuracy and efficiency. Huge variances in the disease symptoms, genetic variations and conditions of light on the plants, all decrease the accuracy. Both the productivity and quality of the crop yield can be increased by detecting the diseases affecting the crops at an early stage so as to reduce the loss incurred at a later stage. This paper contains a survey of the advanced techniques used to detect different diseases affecting the plants.