This paper provides an Edge AI powered framework for predicting plant leaf diseases in real-time applications of precision farming is discussed. The suggested model involves a compact CNN architecture hosted on the edge device to conduct inference operations. In contrast to existing cloud computing solutions that rely on transferring data to the server, the proposed model conducts image processing operations locally, which minimizes response time and reduces the need for internet connectivity. The CNN model is trained on a labeled database of plant leaves images and is further fine-tuned through methods like quantization and pruning. Experimental analysis validates the efficacy of the model, with high accuracy rates achieved under minimal latency conditions.

Keywords

Edge AI, Leaf Disease Detection, Smart Agriculture, CNN, Deep Learning, Image Processing, IoT