Compression of Plenoptic Point Clouds.

Area of Science:

• Computer Vision
• Computer Graphics
• Data Compression

Background:

• Point clouds are used for real-time 3D object capture and rendering.
• Current methods average colors, losing realistic viewing-angle dependent reflections, especially for specular surfaces.
• Plenoptic point clouds represent colors from all viewing directions for each point.

Purpose of the Study:

• To develop a compression method for plenoptic point clouds.
• To adapt the region-adaptive hierarchical transform (RAHT) coder for multi-color point cloud data.
• To evaluate the impact of object specularity on compression performance.

Main Methods:

• Four methods were developed to extend the RAHT coder for multi-color voxels.
• Experiments used synthetic data to correlate specularity with compression efficiency.
• Simulations with natural data employed rate-distortion curves for evaluation.

Main Results:

• Object specularity directly impacts color disparity and coder performance.
• The proposed methods effectively compress multi-color point cloud data.
• A combination of the Kahunen-Loève transform and RAHT achieved optimal compression.

Conclusions:

• The enhanced RAHT coder effectively compresses plenoptic point clouds.
• Specularity is a key factor influencing compression performance for multi-directional color data.
• The proposed compression technique improves the realism and efficiency of 3D object representation.