Many diseases affect rice crops and cause significant losses in their yield of rice crops. The early detection of these diseases will be beneficial to farmers. Although there are many techniques for diagnosing diseases of rice plants from images, this study focuses on analyzing some of these techniques. This study analyzes not only traditional machine-learning techniques but also a modern approach using cloud software. The study focuses on mainly four types of diseases – namely Bacterial Blight, Blast, BrownSpot and Tungro. These rice diseases lead to the accumulation of toxic metabolites or proteins, and altered hormone levels. This study implemented the techniques and analyzed the methods through various metrics such as accuracy, f1-score, precision. This study performed a comparative study of the aforementioned methods and attempted to determine whether traditional machine-learning techniques or modern cloud-based techniques work better. With a model accuracy of 100%, the proposed method ensures rice nutrient depletion through early detection of rice diseases.