Google Summer of Code - 2017

summary.md

Project Abstract

I proposed to

1. Add a package for image segmentation as part of JuliaImages with several algorithms -

• Thresholding - Otsu’s method and Adaptive thresholding
• K-means clustering
• Mean shift segmentation
• Watershed segmentation
• Felzenszwalb's efficient region merging algorithm
• Shi and Malik’s normalized graph-cut based segmentation

2. Add Histogram of Oriented Gradient(HOG) feature to ImageFeatures.

I also planned to add tests, examples codes and documentation for all the algorithms as well as a tutorial for object detection using HOG features.

Mentor: Tim Holy

Organization: The Julia Language

Category: Package

Work Done

All the work except normalized graph cut algorithm has been merged.
List of commits: ImageSegmentation.jl, Images.jl, ImageFeatures.jl, juliaimages.github.io
I have also added tests and example code for all the algorithms. I and Animesh Kashyap (annimesh2809) wrote the documentation collaboratively. Documentation can be found here: http://juliaimages.github.io/latest/imagesegmentation.html.
Object Detection Tutorial: http://juliaimages.github.io/ImageFeatures.jl/latest/tutorials/object_detection.html

For normalized graph cut, I had completed the code and submitted a pull request at ImageSegmentation.jl. Since this algorithm is more generally applicable, we decided to add to this a general graphs package e.g, LightGraphs.jl. I have submitted a pull request at LightGraphs.jl and am modifying the code for it to work with general graphs.
Pull Request at ImageSegmentation.jl
Pull Request at LightGraphs.jl