Severity Calculation of Tar Disease in Sorghum Using Deep Learning

Sorghum, a crucial cereal crop, is confronted with challenges from diseases such as Tar spot, Anthracnose and Rust impacting production. In response to the scarcity of relevant data, our project proposes a Deep Learning solution utilizing the ResNet architecture to classify Tar disease severity in sorghum leaves. Facing the challenge of limited data, we meticulously curated a benchmark dataset comprising healthy and affected sorghum leaves. To enhance our system's intelligence, we selected ResNet, a specialized form of Deep Learning known for understanding intricate leaf patterns and performing well with a small dataset. The proposed methodology incorporates image masking to emphasize disease-related features and severity assessment through automated detection and quantification of black spots. The project aims to deliver a robust and reliable tool for accurately identifying and assessing Tar disease severity in sorghum, providing valuable insights for effective agricultural management.