Broccoli Classifiers: CNN and AdaBoost Models for Differentiating Indian Varieties

This study looks at three main kinds of broccoli - Pusa KTS-1, Green Magic, and Calabrese. It uses a method called Convolutional Neural Networks (CNN) along with AdaBoost ensemble methods to see how these different types grow in Indian farms. They used a set of 1500 carefully taken pictures to teach and test how well the model works. A new kind of computer program called CNN and another one named AdaBoost joined together to make a very good score. They were able to tell different types apart with 97.10% accuracy by looking closely at small changes in pictures. Table I shows the model's strong accuracy, recall, and F1 scores. This confirmed that it is reliable at correctly grouping each kind while catching many different examples from every category fairly. Figure 3 showed the new method was better than the old ways. So, it is at the top for classifying different types of broccoli. Beyond just farming, this study could lead to wider uses in automated sorting systems. These changes can affect how farmers grow and sell their crops. The success of the study shows how well machine learning models can make plant variety classification more accurate and automate it. In the future, the model can be improved by adding more data and making it better. This will lead to higher accuracy when using it on farms or with new tech gadgets. This study is a big step to change how computer systems classify types of vegetables. It brings important help in farming practices and making markets different from each other.