Deep Point Cloud Compression of Indoor Environment Scenes

Holographic-Type Communication promises real-time transmission of holographic content, ensuring fully immersive experiences for remote users. However, it requires accurate capture, efficient compression, and low-latency transmission of voluminous three-dimensional data. This paper specifically focuses on the three-dimensional point cloud compression of large indoor environment scenes, using deep autoencoder architecture. Additionally, the paper proposes an algorithm for point cloud augmentation to enhance model compression efficiency.