monomest / Image-Edge-Detection

Implementation of Laplacian of Gaussian (LoG) filtering on monochrome or colour BMP images. This is project 2 of 3 from ELEC4622 Multimedia Signal Processing, a course at UNSW. This version of the project would have received full marks.

ELEC4622 Project 2

Renee Lu, z5160268 | July, 2020

Tasks Overview

1. Task 1 LOG filtering
2. Task 2 Separable LOG filtering
3. Task 3 Comparison of different sigmas in LOG filtering
4. Task 4 Edge detection using LOG filtering
5. Task 5 Bonus Integer processing for separable LOG filtering
6. Task 6 Bonus Simple Morphology Dilation after LOG filtering

Code Location

The solution file is found in:

• Project2/Project2.sln

The code for Task 1, 2 and 3 are in, respectively:

• Project2/Task1/main1.cpp
• Project2/Task2/main2.cpp
• Project2/Task3/main3_sep.cpp
• Project2/Task3/main3_no_sep.cpp - this is an inefficient implementation
• Project2/Task4/main4.cpp
• Project2/Task5/main5.cpp
• Project2/Task6/main6.cpp

Project Properties

In Visual Studio, the non-default property configuration for the projects are:

• Configuration Properties > General > Output Directory = [executables output file path] e.g. "../../Executables"
• Configuration Properties > Advanced > Character Set = Use Multi-Bye Character Set
• Configuration Properties > Debugging > Command Arguments = [input bmp file path] [output bmp file path] [sigma] [a] e.g. "../Data/pens_rgb ../Data/out.bmp 1 1"
• C/C++ > Preprocessor > Preprocessor Definitions = WIN32; _CRT_SECURE_NO_DEPRECATE

Steps to Run

1. Open Visual Studio
2. Open the Project2/Project2.sln
3. Configure the Project Properties as above
4. Run the desired tasks using the command line arguments below:

To run Task 1 give the following command line arguments: [input bmp file path] [output bmp file path] [sigma] [a]