This repository has source files for UnSee which was just a C++ Project (Nothing Fancy). It uses OpenCV to read images and a custom implementation of Gaussian Blur inspired from this article I found while looking for fast blurring methods because the default one provided by OpenCV was not fast enough..
You need to have Linux(preferably Ubuntu 22.04) because I used WSL2, then install
cmake, OpenCV, and build-essentials etc.Some resilience because OpenCV generally has linking problems :D
Use the following commands :
git clone https://github.com/DarshPareek/UnSee
cd UnSee
cmake --build .
This Should generate a file(executable) named UnSee
To run it :
To run it :
./UnSeeThen just follow the prompts :D.
You can directly download
program.o from the releases and run it using the following commands : chmod +x program.o
./program.omain.cpp : contains the basic loop to spare you from re-running the executable again and again for different images and blur strengths also makes calls to fastGauss.cpp.fastGauss.cpp : accepts the image path and blur strength from main.cpp and does all the crazy math that is required to do a Gaussian Blur. I implemented this js-version into C++. Call me crazy but this is way faster from what OpenCV provides and there is still scope for heavy optimization in my code.CMakeList.txt : contains the bare-minimum stuff to build this project. It links OpenCV libraries and does a whole lot of other stuff.1. I might build a windows version. (windows is a niche operating system used by gamers)
2. I might build a backend server (in C++ or Python) and a frontend GUI in HTML, CSS and JS to allow users to upload images on the localhost and choose blur strength using a slider while the image updates in real-time.
2. I might build a backend server (in C++ or Python) and a frontend GUI in HTML, CSS and JS to allow users to upload images on the localhost and choose blur strength using a slider while the image updates in real-time.