Code Record

2018-10-01

[DOI: 10.21982/9k29-dm62] Automatic Morphological Attribute Profiles
Cavallaro, Gabriele
Morphological attribute profiles are multilevel decompositions of images. This code provides a method for automated selection of filter parameters to provide representative and non-redundant threshold decomposition of the image. This is based on Granulometric Characteristic Functions (GCFs).

Code Link: Automatic_Morphological_Attribute_Profiles.zip
Code Size: 18564075 bytes
Code MD5 Checksum: 644347d4c824a46fb0753b373f72a441


Appears in: G. Cavallaro, N. Falco, M. Dalla Mura and J. A. Benediktsson, "Automatic Attribute Profiles," in IEEE Transactions on Image Processing, vol. 26, no. 4, pp. 1859-1872, April 2017.
doi: 10.1109/TIP.2017.2664667

Code Languages: C++, Matlab

To compile code: - For all operating systems that are supported by MATLAB:
1. Open a terminal (or a Command Prompt)
2. Navigate to /your_path/Morphological_Attribute_Profiles/object-files/OpenME/src
3. Make sure you have a compiler (e.g., gcc --version)
4. Compile the sources with: gcc main.c fimage.c mw.c list.c shape.c list.c flst.c tree.c flst_reconstruct.c -c -fPIC
5. You will get these object files: main.o fimage.o mw.o list.o shape.o list.o flst.o tree.o flst_reconstruct.o
6. Copy these object files into /your_path/Morphological_Attribute_Profiles/object-files/
7. Open MATLAB
8. Go to the main folder /your_path/Morphological_Attribute_Profiles
9. Run >> mex morphological_attribute_profile.cpp object-files/mw.o object-files/list.o object-files/fimage.o object-files/shape.o object-files/flst.o -Iheader -Iheader/ -DNDEBUG
10. Check if the binary MEX file work. Run >> morphological_attribute_profile

Sensor Categories: SAR, Optical Imager

Instrument Processing and Calibration Categories: SAR, Optical Imager

Keywords: Automatic attribute profiles, filter parameters selection, tree representation, mathematical morphology, remote sensing, image processing