A Gravitational Model for Grayscale Texture Classification Applied to the pap-smear Database
This paper presents the application of a novel and very discriminative texture analysis method based on a gravitational model to a relevant medical problem, which is to classify pap-smear cell images. For this purpose, the complexity descriptors Bouligand-Minkowski fractal dimension and lacunarity were employed to extract signatures from the gravitational collapsing process. The obtained result was compared to other texture analysis methods. Additionally, AUC measure performance was computed and compared to several LBP based descriptors presented in two recent papers. The performed comparisons demonstrate that texture analysis based on gravitational model is suitable for discriminating pap-smear images.
Keywordspap-smear database Gravitational model Bouligand-Minkowski fractal dimension Lacunarity
- 9.Jantzen, J., Norup, J., Dounias, G., Bjerregaard, B.: Pap-smear benchmark data for pattern classification. In: Proc. NiSIS 2005, Albufeira, Portugal, NiSIS, pp. 1–9 (2005)Google Scholar
- 11.Sá Junior, J.J.M., Backes, A.R., Cortez, P.C.: A simplified gravitational model for texture analysis. Journal of Mathematical Imaging and Vision 47(1–2), 70–78 (2013)Google Scholar
- 12.Mandelbrot, B.: The fractal geometry of nature. Freeman & Co. (2000)Google Scholar
- 17.Everitt, B.S., Dunn, G.: Applied Multivariate Analysis, 2nd edn. Arnold (2001)Google Scholar