Oak wilt, a highly contagious and lethal disease threatening oak trees across North America, has traditionally been detected through manual inspections, which are often time-intensive and insufficient for timely intervention. This study introduces an automated oak wilt detection system leveraging aerial imagery captured by unmanned aerial vehicles (UAVs) and a convolutional neural network (CNN) to classify environments as “Oak Wilt” or “Not Oak Wilt.” The model, initially trained on a dataset of 271 oak wilt and 310 non-oak wilt images, was further refined with an expanded dataset of 1,051 images and tested on 9,981 images sourced from Lake Forest Cemetery State Park, Mulligan’s Hollow State Park, P.J. Hoffmaster State Park, and Warren Dunes State Park. A customized CNN architecture was developed, comprising three convolutional layers, three max-pooling layers, and fully connected layers. To enhance accuracy, the system incorporates data augmentation and geotagging. Achieving an overall accuracy of 86.72%, the model effectively detects early-stage active oak wilt by analyzing color changes in oak tree canopies. Further performance improvements were achieved by integrating Reinforcement Learning from Human Feedback (RLHF), enabling the system to adapt based on user inputs. The model is deployed through a web platform developed with VueJS and Flask, enhancing accessibility and scalability. By enabling early detection through aerial image analysis, this approach minimizes reliance on manual surveys and contributes to proactive forest conservation efforts.