<![CDATA[The need for food, especially rice, continues to increase. Therefore, a production increase of around 70% is required to meet the demand. In this case, supervision and care of rice from planting must be carried out efficiently, which can be done by deep learning, namely Convolutional Neural Networks (CNN). The classification was carriedout on the image of rice seedlings in the form of rice seedlings and bare land patch images. The main purpose of this research is to conduct a comparison test of the performance of each CNN model with a lightweight architecture and validate the architecture. A lightweight CNN architecture is used due to its lower architecture size but still has decent performance compared to the regular CNN model for the rice seedlings dataset. Training and testing were carried out on the Rice Seedling Dataset to determine the performance of the proposed method. The research was built using the PyTorch library and the Python programming language and resulted in 99% of accuracy, precision, recall, kappa, and F-1 Score. In addition, validation was carried out using K-Fold Cross Validation which also had the best accuracy of 99%. Therefore, we conclude that the developed model can properly classify images of rice seedlings and arable land.]]>
Copyrights © 2023
