Title :
Classification system for blister pack of pills
Author :
Pardo, C.E. ; Sosa, L.F. ; Gutierrez, E.A. ; Jimenez, F.R.
Author_Institution :
Univ. Santo Tomas Tunja, Santo Tomas, Colombia
Abstract :
In this paper we present a classification system for blister pack of pills. For the development of the system was taken into account three types of Blister Packs, where we performed a classification process considering a Segmentation-based Fractal Texture Analysis system, likewise, was used Otsu multi-level algorithm, as nonparametric technique for thresholds detection. The features extracted from the Fractal Texture-based Segmentation Analysis system, are used as a database of entrainment to a classification system based on k-nearest neighbors.
Keywords :
drugs; feature extraction; image classification; image segmentation; image texture; nonparametric statistics; Otsu multilevel algorithm; classification process; classification system; entrainment database; feature extraction; fractal texture-based segmentation analysis system; k-nearest neighbors; nonparametric technique; pill blister pack; segmentation-based fractal texture analysis system; threshold detection; Biomedical imaging; Feature extraction; Fractals; IEEE members; Image color analysis; Image segmentation; Vectors; Blister Pack; KNN; Otsu Multi-level; SFTA;
Conference_Titel :
Circuits and Systems (CWCAS), 2014 IEEE 5th Colombian Workshop on
Conference_Location :
Bogota
Print_ISBN :
978-1-4799-6838-1
DOI :
10.1109/CWCAS.2014.6994614
