Home

Awesome

REVIEW OF EDGE DETECTORS - IPOL 2012

About

These are the source files implementing some classical algorithms of edge detection.

The implemented algorithms are:

The algorithms are implemented in C language and are compiled using cmake.

Filelist

Root directory:

.:
@build          - Suggested compilation directory.
@build/bin  	 -	Binary files folder (empty until compilation).
CMakeLists.txt  -	CMake configuration file.
README  		- 	This file.
src  			-	Source files folder.
test			-	Test images folder.

./@build/bin: (empty until compilation)

./example:

lena_harlick.png	-	Output of Haralick algorithm using lena.png as input image.
lena_mh.png			-	Output of Marr-Hildreth algorithm (Gaussian kernel) using lena.png as input image.
lena_mhlog.png		-	Output of Marr-Hildreth algorithm (LoG kernel) using lena.png as input image.
lena_prewitt.png	-	Output of Prewitt algorithm using lena.png as input image.
lena_roberts.png	-	Output of Roberts algorithm using lena.png as input image.
lena_sobel.png		-	Output of Sobel algorithm using lena.png as input image.

./src:

2dconvolution.c  	-	Implementation of the 2-D convolution.
2dconvolution.h		-	(header file).
CMakeLists.txt		-	CMake configuration file.
gaussian_kernel.c	-	Implementation of the Gaussian and LoG kernel generation.
gaussian_kernel.h	-	(header file).
iio.c 				-	IIO library C-file.
iio.h     			-	IIO library (header file).
test_fded.c      	-	First derivative edge detectors, main C file.
test_haralick.c		-	Haralick algorithm, main C file.
test_mh.c   		-	Marr-Hildreth algorithm (Gaussian kernel), main C-file.
test_mh_log.c 		-	Marr-Hildreth algorithm (LoG kernel), main C-file.

./test:

lena_gray.png  	-	Lena image (512x512px).
molino_gray.png   -	Windmill image (1000x563px).
oranges_gray.png	-	Oranges image (536x480px).

Compilation

In the @build directory:

> cmake ..
> make

An executable file is created in the @build/bin directory:

bin/:
edges_ipol
test_fded  		-	First derivative edge detectors (Roberts, Prewitt & Sobel).
test_haralick  	-	Haralick algorithm.
test_mh  		-	Marr-Hildreth algorithm, using Gaussian kernel.
test_mh_log		-	Marr-Hildreth algorithm, using LoG kernel.

Execution

First derivative edge detectors:

Usage:

> test_fded input_image threshold

Example:

> bin/test_fded test/lena.png 0.1
Input image loaded:	 512x512 image with 3 channel(s).
images converted to grayscale
Roberts image saved to roberts.png.
Prewitt image saved to prewitt.png.
Sobel image saved to sobel.png.
Edge detection algorithms based on first derivative computation done.
execution time: 0.280 s.

Haralick algorithm:

Usage:

> test_haralick input_image rhozero output

Example:

> bin/test_haralick test/lena.png 0.5 lena_harlick.png
Input image loaded:	 512x512 image with 3 channel(s).
images converted to grayscale
60122 edge points found...
Output Image saved in lena_harlick.png:	 512x512 image with 3 channel(s).
haralick's edge detector computation done.
execution time: 0.440 s.

Marr-Hildreth algorithm (Gaussian Kernel):

Usage:

> test_mh input_image sigma n tzc output_image

Example:

> bin/test_mh test/lena.png 2 13 0.1 lena_mh.png
Input image loaded:	 512x512 image with 3 channel(s).
images converted to grayscale
Output Image saved in lena_mh.png:	 512x512 image with 3 channel(s).
marr-hildreth edge detector computation done.
execution time: 0.420 s.

Marr-Hildreth algorithm (LoG kernel):

Usage:

> test_mh_log input_image_1 sigma n tzc output

Example:

> bin/test_mh_log test/lena.png 2 17 0.1 lena_mhlog.png
Input image loaded:	 512x512 image with 3 channel(s).
images converted to grayscale
Output Image saved in lena_mhlog.png:	 512x512 image with 3 channel(s).
marr-hildreth edge detector computation done.
execution time: 0.670 s.

Test images

Some test images are provided in the «test» folder. The output images using «lena.png» as input image, are also provided in the «example» directory.

Tested platforms

Changelog

Copyrights

Copyright (C) 2011-2012, Haldo Spontón haldos@fing.edu.uy

Copyright (C) 2011-2012, Juan Cardelino juanc@fing.edu.uy

License

This program is free software: you can use, modify and/or redistribute it under the terms of the GNU General Public License as published by the Free Software Foundation, either version 3 of the License, or (at your option) any later version. You should have received a copy of this license along this program. If not, see http://www.gnu.org/licenses/.

This program is free software: you can use, modify and/or redistribute it under the terms of the simplified BSD License. You should have received a copy of this license along this program. If not, see http://www.opensource.org/licenses/bsd-license.html.

This program is provided for research and education only: you can use and/or modify it for these purposes, but you are not allowed to redistribute this work or derivative works in source or executable form. A license must be obtained from the patent right holders for any other use.