Disease detection in agriculture is critical in severe weather conditions. Keeping an eye on many cotton fields is challenging for farmers. Deep learning and image processing identify agricultural ailments before they spread and further damage. Cotton crop with the most important correlation quantities among pest disease existence are inspected by predictors. In this study, transfer learning methods for analysis are checked. The machine learning approach developed accurately distinguishes healthy and unhealthy plants. As a result, the proposed model may be used to monitor a wide range of areas for more rapid analysis and action, leading to increased productivity. It is possible to anticipate a timeline using ARIMA. A long-term study reveals that the value swipes widely during short and long periods due to random influences. It is also inspected how the recommended approach is affected by using several cotton crop characteristics. In addition to separating green crops from other crops in the field, this procedure can also separate harvests such as Cotton from other crops.