Keynote Link

imgmanip

Design Using C++ Final Project
This is the image manipulation library implemented in C++. We have the following features:

• Grayscale
• Convolution
• Image Warping
• Photo Mosaic
  

Goal :
We aim to provide a library package that can treat images with some amazing effects. The GrayScale and Convolution is the basic operation for the image manipulation and widely used. The image warping can be used to make some pictures fit in different shapes. For example, if the company wants to add the advertisement on some irregular surfaces, image warping can help the company to make the logo better fit in that surfaces in the picture. The photo mosaic can be used for business communication. For example, it can be used to document the parties or events with the company's logo as the target images and the employees and event pictures as the source images.

Getting Started

Install Dependencies

Download dependent lib via this link and unzip the folder in the project directory.

Install any future dependencies into the lib folder. Current dependencies include:

• armidillo for linear algebra,
• libjpeg and libpng for decoding images
• boost GIL for reading images

Important

You will need to run ./install.sh to finish setting up these dependencies. ENSURE YOU HAVE NO WHITESPACE IN YOUR DIRECTORY PATH.

Build program

1. First generate precompiled headers
   imgmanip/imgio/pch.h.gch
2. Build the main program
   make
3. Run the program
   ./main

For future tests, write additional make entry in Makefile.

Using program

Usage options:
        ./main —-mosaic 'tgtImage.png' ‘srcDirectory'
        ./main ——homography 'srcImage.png' ['trapezoid' | 'spiral' | 'rTrapezoid' | 'random']
        ./main --grayscale 'srcImage.png' 
                        optional: --shades intNumber

         For custom usage, modify main function and choose from our wide range of tools

imgmanip

Each subdirectory in imgmanip is a "module." Remember to update Makefile when you add a new "module."

Some Demo Operations

Mosaic

Target image:

Mosaic image:

Grayscale

Target image:

Grayscale image:

Homography

Target image:

Homography image: