Three-dimensional range geometry compression via phase encoding.

Area of Science:

• Computer Vision
• Image Processing
• 3D Graphics

Background:

• Current 3D geometry compression methods encode data in 2D color images, typically using all three color channels.
• This leaves no capacity for storing additional information like textures within the same image.

Purpose of the Study:

• To develop a novel 3D range geometry compression method that allows for simultaneous storage of 3D data and texture information.
• To leverage geometric constraints from structured light 3D scanning to reduce data storage requirements.

Main Methods:

• Utilized geometric constraints inherent to structured light 3D scanning devices.
• Developed a compression technique requiring only two color channels for 3D data representation.
• Allocated the remaining color channel for storing supplementary information, such as texture images.

Main Results:

• Achieved a significant compression ratio of 3038:1 compared to the STL format.
• Maintained a root-mean-square error of 0.47% when the output image was compressed using JPEG 80%.
• Demonstrated the robustness and effectiveness of the proposed method through experimental validation.

Conclusions:

• The novel method successfully reduces the data needed for 3D geometry compression.
• The technique enables efficient integration of 3D data and texture information within a single image.
• The approach offers a robust and high-performance solution for 3D data compression.