Abstract:
In recent times, there have been different varieties of bananas in our country, which we
cannot understand when we see them easily. When we go to a shop or market, we cannot
tell which type of banana it is, which is a big challenge for us. If we do not know the banana
variety properly, we cannot go anywhere and tell anyone that we need this type of banana.
Now, when you go to buy bananas but you don't know the types of bananas, you want one
kind of banana, but they give you another kind of banana, and you can't understand
something. That's why we have a big problem. So, we have determined the different
varieties of bananas through the machine, which makes it very easy for us to recognize the
varieties of bananas. To solve this problem, we have tried to solve it by using some
algorithms through machine learning. Implementing different machine learning techniques,
such as pre-trained models, this comparative study analyzes the performance of different
machine learning algorithms. The trial results verify that the ResNet50 model has achieved
the highest accuracy of 98.09% compared to other VGG16, Xception, and InceptionV3
models.
