Skip to content

aidanmontag13/MultiPlaneDisplay

Repository files navigation

MultiPlaneDisplay

How does it work?

Multi-Plane Display uses image processing and Pepper's Ghost illusions to enable any photo to be viewed with real depth, creating a "2 and a half D" Illusion.

Workflow

  1. Depth segmentation
    A depth map is extracted from the image using Depth Anything v2 and used to segment the scene into foreground, middleground, and background layers.

  2. Viewer tracking
    Monocular head tracking estimates the viewer’s position relative to the display.

  3. Shadow correction
    Shadow maps are dynamically adjusted based on the viewer’s position to prevent off-axis ghosting artifacts.

  4. Optical display
    The layered images are displayed through a parascope mirror array, producing the virtual display stacking effect.


Input Image

Processed Image

Viewable Result

Gallery

Features

  • Automatic image processing with input from USB drive
  • Compatibility with .jpg, .jpeg, .png, and .tiff
  • Automated occlusion maps with real-time headtracking
  • Automatic Display turn-off and idle mode when unobserved for 30 seconds

Parts List

About

This is a repository for code and schematics for multiplane display concept

Topics

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Contributors 2

  •  
  •  

Languages