This research utilized the potential of UAV-based remote sensing imagery and deep learning models in the efficient identification of diseases and pests among rice crops to advance precision agriculture. Over 20 two-acre rice fields for the duration of six-month surveillance were made with real-time imagery using UAVs installed with high-resolution cameras at 48 MP. The UAV system, consisting of lightweight fixed-wing designs with brushless DC motors and Li-Po battery power, was able to transmit real-time videos to a local station for image zooming and disease analysis. The dataset was enriched with temporal environmental data such as temperature, humidity, soil moisture, and pesticide application in the collected samples of diseased and healthy crops. The advanced Convolutional neural network (CNN) models that were used included VGG16, VGG19, and ResNet50. Recurrent neural network (RNN) was utilized for temporal data processing. Among them, the VGG19 + RNN model performed the best at an accuracy of 98.78%, followed by VGG16 + RNN at 94.50% and ResNet50 + RNN at 88.65%. Diseases such as bacterial blight, sheath blight, rice blast, brown spot, leaf scald, and bacterial leaf streak were diagnosed at early stages, which enabled timely interventions such as targeted pesticide application and nutrient optimization to prevent yield losses. This study highlights the possibility of AI-driven UAV systems in revolutionizing crop health monitoring by reducing labor-intensive methods and improving sustainability through precise disease management. The discovered findings form the basis for scalable, real-time detection of diseases in agricultural areas and the potential for future improvements in automated crop monitoring systems.