Large Field-of-View Integral Imaging Pickup System

Python implementation of large field-of-view integral imaging pickup system.

“Computational large field-of-view RGB-D integral imaging system”, Sensors, November 2021
Download paper : Computational large field-of-view RGB-D integral imaging system

In this system, use the code Computational integral imaging pickup system

Process

Pre-trained monocular depth estimation model

• Trained by NYU RGB-D V2 (650 MB)

Usage

• Download depth estimation model file.
  • Go to the link above, and download model.
  • Locate file at monodepth or /your/own/path/.
• Prepare the input image.
  • Locate the input color image to inputs directory or /your/own/path/.
• Start large FOV integral imaging pickup system.

  python main.py \
      --color_path ./inputs/image_file_name or /your/own/path/ \
      --output_path ./results or /your/own/path/ \
      --model_path ./monodepth/model.h5 or /your/own/path/ \
      --is_gpu

Results of our system

• Sub-aperture Image Array

  

To-Do List

• Update codes (Depth sub-aperture images).