Astrophysical Objects Image is a deep learning image dataset for astronomical and astrophysical research. This dataset aims to provide structured and standardized astrophysical image resources for convolutional neural networks (CNNs) and other computer vision models, for tasks such as astrophysical image classification, object recognition, and model performance evaluation.

The images in this dataset come from multiple publicly available astronomical data sources. Each directory contains 12 subfolders categorized by celestial body: asteroid, black hole, earth, galaxy, jupiter, mars, mercury, neptune, pluto, saturn, uranus, and venus. All images are stored in a standardized folder structure according to their category. Due to the inconsistent resolution, preprocessing such as resizing and normalization is recommended before training.