https://github.com/suft/differ
https://github.com/suft/differ
Last synced: 10 months ago
JSON representation
- Host: GitHub
- URL: https://github.com/suft/differ
- Owner: suft
- License: mit
- Created: 2016-11-11T15:20:48.000Z (over 9 years ago)
- Default Branch: master
- Last Pushed: 2016-11-11T16:39:30.000Z (over 9 years ago)
- Last Synced: 2025-04-09T11:49:23.139Z (about 1 year ago)
- Language: JavaScript
- Size: 7.29 MB
- Stars: 1
- Watchers: 1
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
- License: LICENSE
Awesome Lists containing this project
README
# Differ
## Summary
An algorithm written in Node, to spot the differences between two images and produces a grayscale image highlighting the differences. The algorithm uses percent differences in RGBA values between pixels of the two images and determines outliers by comparing values against a certain tolerance.
## Differences
* Liquid in the barrel
* Rope tied to the boat
* Fox's nose
* Roof of the well
* Vegetable
* Wheat
* Plowed soil
* Sheep's Ear
* Gazebo railing
## Thinking Process
When I first started the problem, I was able to spot several differences just by inspection, however I decided to write an algorithm to ensure I would spot all the differences. My first attempt at writing the algorithm involved comparing hexadecimal colour values of the pixels of both images, to my surprise, my program was outputting a 94% difference between the two images. Since the pixel colour comparison wasn't yielding any results, I opened the image in an editor and used an eye-dropper to find that the pixels which supposedly the "same" varied slightly while the pixels which were "different" varied quite a lot. So, I converted the hexadecimal colour into RGBA and calculated the percent difference between between the red aspect, green aspect, blue aspect, alpha aspect of each pixel. From there, I used the interquartile range to identify outliers.
## Time Spent
* Total of 4 hours
* Choosing what language to use
* Looking for a pixel manipulation library
* Hexadecimal colour comparison approach
* Calculating a percent tolerance to detect outliers
* RGBA percent difference approach
## Dependencies
Jimp [[https://github.com/oliver-moran/jimp](https://github.com/oliver-moran/jimp)]
## Authors
Sufien Tout [[https://github.com/suft](https://github.com/suft)]