Scalable Coding of Plenoptic Images by Using a Sparse Set and Disparities.

Area of Science:

• Computer Vision
• Image Processing
• Optics

Background:

• Focused plenoptic cameras capture spatial and angular scene information using microlens arrays.
• This technique generates high-resolution images with significant data redundancy.
• Efficient compression is crucial for storage, transmission, and rendering of plenoptic data.

Purpose of the Study:

• To propose a novel lossy coding scheme for efficient plenoptic image representation.
• To reduce the bit rate of plenoptic images while maintaining reconstruction quality.
• To develop a scalable coding structure for plenoptic image data.

Main Methods:

• A lossy coding scheme representing plenoptic images via a sparse set and disparities.
• Reconstruction using disparity-based interpolation and inpainting.
• Using the reconstructed image as a prediction reference for full plenoptic image coding.
• Implementing a three-layer scalable coding structure.

Main Results:

• Achieved over 60% bit rate reduction compared to High Efficiency Video Coding (HEVC) intra coding.
• Demonstrated over 20% bit rate reduction compared to HEVC block copying mode.
• The proposed scheme offers efficient compression for plenoptic images.

Conclusions:

• The proposed lossy coding scheme effectively compresses plenoptic images.
• The method leverages disparity information for efficient representation and reconstruction.
• The resulting scalable structure facilitates flexible plenoptic data handling.