Agriculture is a critical sector for global food security, but plant diseases and nutrient deficiencies remain major challenges that reduce crop yield and economic returns for farmers. Traditional diagnosis methods depend on manual observation and expert intervention, which are often time-consuming, subjective, and inaccessible in remote regions. This paper presents an intelligent crop health analysis system that automates the detection of plant diseases and nutrient deficiencies using a hybrid deep learning framework. The proposed approach integrates Convolutional Neural Networks (CNNs) for feature extraction and classification with Generative Adversarial Networks (GANs) for synthetic image generation and dataset augmentation. The CNN model learns discriminative features such as color variations, texture patterns, and lesion characteristics from leaf images, while the GAN enhances dataset diversity by generating realistic samples, thereby addressing class imbalance and limited training data. The system is trained on a dataset containing more than 55,000 leaf images across 32 classes and is deployed through a Flask-based web application. It supports two operational modes: Basic Mode for disease identification and Advanced Mode for comprehensive crop health assessment through the integration of CNN-based predictions and rule-based nutrient analysis. Experimental results demonstrate improved classification performance, robustness, and scalability under real-world conditions. Additionally, the multilingual user interface enhances accessibility for farmers from diverse linguistic backgrounds. The proposed system provides an effective and practical solution for early crop health monitoring, enabling timely intervention, reducing dependency on agricultural experts, and contributing to increased agricultural productivity and sustainable farming practices.