aylmerbritto/fastcode-image-resizing

Fast Code for Image Interpolation

This is a part of submission for the course 18-645: How To Write Fast Code

Workspace

We conducted our experiments on the ECE cluster. The ECE machine is ece004.ece.local.cmu.edu .
The path to our workspace directory is /afs/ece.cmu.edu/usr/arexhari/Public/645-project

Code Execution Instructions:

Nearest Neighbors

1. make bi-linear-run is to demonstrate the functional correctness of our nearest neighbor interpolation algorithm

• It makes use of a inputs/640-480.jpg and writes the output to results/640x480-nn.jpg

2. make bi-linear-performance is to demonstrate the performance of our code. After the code is run you can see the csv results at plots/kernelPerformanceNN.csv
3. make parallel is to demonstrate the parallelisation of our code. After the code is run you can see the csv results at plots/blParallel.csv

Bi-Linear Kernel

1. make nn-run is to demonstrate the functional correctness of our nearest neighbor interpolation algorithm

• It makes use of a inputs/640-480.jpg and writes the output to results/640x480-bl.jpg

2. make nn-performance is to demonstrate the performance of our code. After the code is run you can see the csv results at plots/kernelPerformance.csv
3. make parallelNN is to demonstrate the parallelisation of our code. After the code is run you can see the csv results at plots/nnParallel.csv