Using CNN and Gradient Boosting to Accurately Classify Guar Varieties in India

This study introduces a new way to group different Guar types in India. It uses Convolutional Neural Networks (CNN) and Gradient Boosting (GB) methods for this task. A complete set of about 11,000 clear pictures was carefully gathered. They include different light situations and angles from various growth stages. Strict rules were used to clean the images and a new design of the model, which uses CNNs, helped get detailed features. The model was really good and got it right $\mathbf{9 7. 7 \%}$ of the time. It showed how well it could tell small differences between types of Cluster Beans. The F1 score and mistakes that can happen are also explained here. Furthermore, a modern comparison shows that the model is much better than other methods currently in use. This study’s effects go beyond school, giving a strong grouping system important for farming people. A good model’s understanding helps make smart decisions. It improves farming methods and how resources are used by knowing the differences between varieties better. This announces a new start where big computer learning methods are mixed with farming studies. It promises to change the way the farm is, helping it be better and make more food at once. This study is an important step for accurately sorting Cluster Bean types. It provides a good tool to help farming progress. Its success not only makes farming better but also helps us to come up with new ideas and work together on different subjects. This will help create a future where using smart technology in agriculture and making educated choices.