Lemon cultivation is an important component of horticultural agriculture, contributing significantly to bothdomesticconsumption and commercial markets. However, lemon plants are highly vulnerable to a variety of leaf diseases suchascitruscanker, anthracnose, leaf curl virus, spider mite infestation, and other fungal and bacterial infections. These diseases adverselyaffect leaf health, fruit quality, and overall yield, leading to substantial economic losses for farmers. Early and accuratedetectionof such diseases is critical for effective disease management and for minimizing the excessive use of pesticides. Traditional diseaseidentification methods rely on visual inspection by experienced farmers or agricultural experts, which is time-consuming,subjective, and often impractical in large-scale or rural farming environments where expert access is limited. To addressthesechallenges, this paper proposes a real-time lemon leaf disease detection system using digital image processing andartificialintelligence techniques. The system is designed to automatically capture, process, and analyze leaf images under real-fieldconditions. Live images of lemon leaves are acquired using a Raspberry Pi camera module, enabling continuous monitoringwithout manual intervention. The captured images are pre-processed using Open-CV and NumPy-based techniquessuchasresizing, noise removal, background elimination, color normalization, and contrast enhancement to improve image qualityandensure robustness against varying lighting and environmental conditions. Following pre-processing, the system performs feature extraction to identify key visual indicators of disease, includingcolorvariation, texture irregularities, lesion patterns, and shape deformation. These features are then analyzed usingatrainedconvolutional neural network (CNN) model optimized for real-time performance. Lightweight deep learning architecturesandTensor-Flow Lite are employed to ensure efficient execution on low-cost hardware platforms such as the Raspberry Pi, makingthesystem suitable for field-level deployment. The model is capable of accurately detecting and classifying multiple lemonleafdiseases without the need for expert supervision. Keywords: Lemon leaf, disease detection, real-time monitoring, image processing, CNN, Tensor-Flow Lite, smart agriculture,plant disease classification, feature extraction, Raspberry Pi, precision agriculture