Publications

@article{gonzalez-castro_automatic_2015,

title = {Automatic classification of skin lesions using color mathematical morphology-based texture descriptors},

author = {Víctor González-Castro and Johan Debayle and Yanal Wazaefi and Mehdi Rahim and Caroline Gaudy-Marqueste and Jean-Jacques Grob and Bernard Fertil},

url = {https://www.spiedigitallibrary.org/conference-proceedings-of-spie/9534/953409/Automatic-classification-of-skin-lesions-using-color-mathematical-morphology-based/10.1117/12.2182592.short},

year  = {2015},

date = {2015-01-01},

journal = {Twelfth International Conference on Quality Control by Artificial Vision 2015},

volume = {9534},

pages = {53–59},

abstract = {This paper presents an automatic method for classifying skin lesions in dermoscopic images using color texture analysis. It combines mathematical morphology for local pixel descriptors with Kohonen Self-Organizing Maps (SOM) for clustering and global texture description, eliminating the need for segmentation. Two approaches—classical and adaptive morphology—achieve similar AUC scores (0.854 and 0.859), surpassing dermatologist predictions (0.792).},

note = {Publisher: SPIE},

keywords = {dermoscopic imaging, machine learning, skin lesion classification, texture analysis},

pubstate = {published},

tppubtype = {article}

}

@article{gonzalez-castro_automatic_2015-1,

title = {Automatic classification of skin lesions using geometrical measurements of adaptive neighborhoods and local binary patterns},

author = {Víctor González-Castro and Johan Debayle and Yanal Wazaefi and Mehdi Rahim and Caroline Gaudy-Marqueste and Jean-Jacques Grob and Bernard Fertil},

url = {https://ieeexplore.ieee.org/abstract/document/7351095},

year  = {2015},

date = {2015-01-01},

urldate = {2015-01-01},

journal = {2015 IEEE International Conference on Image Processing (ICIP)},

pages = {1722–1726},

abstract = {This paper presents a method for characterizing and classifying skin lesions in dermoscopic images to detect melanoma. It uses Local Binary Patterns (LBPs) on geometrical feature maps derived from General Adaptive Neighborhoods (GAN). An Artificial Neural Network evaluates performance, showing that the GAN-based approach outperforms classical LBPs and dermatologists' predictions in ROC curve analysis.},

note = {Publisher: IEEE},

keywords = {dermoscopic imaging, LBP, local binary pattern, skin cancer detection},

pubstate = {published},

tppubtype = {article}

}