guidocalvano / CUDAatScaleForTheEnterpriseCourseProjectTemplate

This is a template for the course project for the CUDA at Scale for the Enterprise

Blur images

This is a filled in template for the course project for the CUDA at Scale for the Enterprise

Project Description

The project blurs a number of images sized 512 by 512. It reads unblurred files from one folder and output blurred versions in another folder.

It is my entry for an assignment in the gpu programming specialization coursera course from John Hopkins university.

To install dependencies run:

make install

To download the test dataset run:

make data

To build the executable run:

make build

This produces ./bin/main.exe.

To run this executable using default arguments run:

make run

To get help on running ./bin/main.exe with different arguments, just run it without extra arguments.

To clean the built executable run:

make clean

To generate a zip of proof that my code ran correctly run:

make proof

Code Organization

bin/ This folder contains main.exe after running the following command:

make build

data/ After running the following command the data folder will contain textures in the folder ./data/textures.

make data

After running main.exe using the following command ./data/output will contain blurred texture files:

make run

lib/ There are no libraries. Everything is installed using sudo apt-get install.

src/ All source is in the src folder. It consists of a header file main.cuh and a code file main.cu.

README.md This file should hold the description of the project so that anyone cloning or deciding if they want to clone this repository can understand its purpose to help with their decision.

INSTALL To install libopencv and libboost on ubuntu do:

make install

Makefile or CMAkeLists.txt or build.sh Makefile contains the make rules.

run.sh There is no run.sh. Just use main.exe in the bin folder.