CS 445 Final Project - Focus Stacking

An implementation of Laplacian Focus Stacking and 3D Depth Mapping from an image stack.

Getting Started

Prerequisites

• Ensure you have a Python 3 environment (Tested with 3.11) set up with the following packages:
  • matplotlib
  • numpy
  • opencv
  • sciki-image
• An environment.yaml config file is provided for conda environments

Usage

To generate all project results, run:

python generate_results.py

• Results can then be found in ./results_naive for naive approach, ./results_focus for laplacian, and ./results_stack for 3d depth mapping
• Exact arguments used to generate results for each method are defined in the generate_results.py script

About

• depth_mapping.py - Script containing 3D depth mapping implementation
• naive_approach.py - Script containing naive focus stacking implementation
• focus_stacking.py - Script containing laplacian focus stacking implementation
• generate_results.py - Script containing arguments

Authors

• Bennett Wu
• Deeya Bodas
• Victor Will

Acknowledgments

A list of resources and references used in the development of the project

• Laplacian Focus Stacking: Adelson, Edward & Anderson, Charles & Bergen, James & Burt, Peter & Ogden, Joan. (1983). Pyramid Methods in Image Processing. RCA Eng.. 29.

• Depth Mapping: J. Wlodek, K. J. Gofron, Y. Q. Cai; Achieving 3D imaging through focus stacking. AIP Conf. Proc. 15 January 2019; 2054 (1): 050001. https://doi.org/10.1063/1.5084619

• ECC Image Alignment: https://learnopencv.com/image-alignment-ecc-in-opencv-c-python/

• Bug Images: https://github.com/bznick98/Focus_Stacking

• Ant Images: http://grail.cs.washington.edu/projects/photomontage/

• Depth Map Visualization: https://depthmapviewer.ugocapeto.com/