Star Galaxy Classification Using Deep Learning

The classification and identification of stars and galaxies, play a crucial role in astronomical research and exploration. This system presents deep learning-based approach for star galaxy classification utilizing Convolutional Neural Networks (CNNs) and the VGG16 architecture. By leveraging the hierarchical nature of the VGG16 architecture, we aim to capture intricate patterns and distinctive features inherent in star and galaxy images. To train and evaluate the proposed model, a comprehensive dataset consisting of labeled star and galaxy images is collected from kaggle. During the training process, the VGG16 model is fine-tuned using transfer learning, utilizing weights pretrained on a image dataset. The VGG16 approach achieved an accuracy of 95.56% and CNN achieved 94.15% accuracy. The model exhibits promising performance, effectively discriminating between stars and galaxies based on their unique visual characteristics.