Agricultural Informatics: Assessing CNN and Decision Tree Effectiveness in Indian French Bean Variety Classification

This study delves into the application of CNNs and DT to classify prominent French bean varieties in India, including Selection-9, Pusa Parvati, and Arka Suvidha. A vast assortment of almost eight thousand carefully chosen high-resolution photos illustrating different stages of development, lighting arrangements, and viewpoints. The study’s overarching goal is to create a reliable classification system that can autonomously extract meaningful information from photos using CNN and visually distinguish between different varieties of beans using DT. With an overall accuracy of 98.7 percent, the model proved it was expert at correctly classifying the three French bean kinds according to their outward look. measures like accuracy, recall, and F1 score were used to evaluate the model’s performance. These measures revealed the model’s strengths for each group. Specifically, they shed light on its ability to classify different items. On top of that, the Misclassifications were exposed by the Confusion Matrix. The results of this investigation showed where the model could use some improvements. By developing a computational tool for the accurate classification of numerous French bean varieties, this work improves agricultural practices. Farmers, researchers, and anybody else involved in the worldwide distribution of crops greatly benefit from this procedure of supplying accurate information about the type and origin of crops. By doing so, we can be sure that everyone taking part in this crucial endeavor has all the information they need to make educated choices. Negative outcomes may result from neglecting this task. As a result, you must participate in and back this initiative, which is called Give Something Back (GSB). This lays the groundwork for future innovations in crop differentiation and shows how effective it is to use advanced educational strategies to improve agricultural operations. Future work can concentrate on improving the model’s performance in a wider range of environmental scenarios and increasing the amount of data collected.