Enhancing Cabbage Plant Protection: CNN Algorithm for Insect Identification

People frequently find it straightforward to identify insects that feed on cabbage plants based on the insects’ characteristics. However, challenges arise when employing a computer for precise bug categorization and identification. The implementation of Deep Learning (DL) utilizes the Convolutional Neural Network (CNN) technique, renowned for its effectiveness in categorizing and recognizing objects in digital photo data. The CNN algorithm delivers notable results in photo identification, simulating the human visual brain’s image recognition through artificial neural networks in the Python programming language. To enrich the user experience, frameworks like Keras and TensorFlow are employed in neural network development. Our research will use the CNN algorithm to determine insect species in cabbage images. Achieving high accuracy involves coupling the CNN algorithm with the VGGNet architectural model. Notably, utilizing an epoch value of 150 resulted in a 97.75% accuracy rate during the training model technique; however, the validation accuracy was slightly lower at 94.25%. The corresponding loss and validation rates were 0.0613% and 0.2945%, respectively. Moreover, based on the model (150 epochs), the testing outcomes in the study, involving 48 images beyond the dataset and incorporating additional information, resulted in an accuracy rate of 91.66%.