Accurate identification of leaf diseases is crucial for maintaining healthy crops and ensuring sustainable agriculture. Manual detection methods are often time-consuming, inconsistent, and dependent on expert knowledge. To overcome these limitations, advanced object detection algorithms have been employed for automated disease identification. This project leverages convolutional neural networks-based detection frameworks to analyze images of apple leaves and classify various disease types. By integrating two cutting-edge object detection models, the system efficiently detects infected regions, categorizes the diseases, and provides visual outputs highlighting affected areas. Extensive experimentation demonstrates improved detection accuracy and robustness across varying leaf patterns and lighting conditions. The approach reduces reliance on human expertise, provides scalable disease monitoring, and supports informed decision-making in crop management. The system’s architecture combines a web-based interface with image processing capabilities, allowing users to upload leaf images and receive disease identification results through a responsive platform. Experimental results indicate that the detection models achieve high precision, recall, and overall performance, validating the effectiveness of automated object detection in plant disease analysis. This methodology lays the foundation for further enhancements in precision agriculture and intelligent crop monitoring.

Keywords: Apple leaf, Disease detection, Object detection, YOLOv12, YOLOv26, Image classification, Leaf segmentation, Agricultural automation, Deep learning, Computer vision.