Bhattacharyya distance of shape feature histograms as aid to authenticate H. R. Ocampo paintings

Authors

Rosemarie Terio ⋅ PH National Institute of Physics, UP-Diliman,
Maricor Soriano ⋅ PH National Institute of Physics, University of the Philippines Diliman

Abstract

Image processing is the most recent tool to help art specialists in artwork authentication. Present studies used wavelet and fractal distribution techniques. In this paper we attempt to use shape descriptors to find statistical signatures in the abstract paintings of the national artist H.R. Ocampo. From a total of 109 shapes extracted from seven verified paintings and 20 from a painting in question (PIQ) the frequency distributions for shape descriptors were obtained and the Bhattacharyya distances is used to measure distribution similarities. It is found that out
of 14 shape descriptors, only 3 Testing and PIQ sets and only 4 Training and PIQ sets have shown almost similar distribution.

Downloads

Issue

2013: Proceedings of the 31st Samahang Pisika ng Pilipinas Physics Congress

Article ID

SPP2013-PB-24

Section

Poster Session PB

Published

2013-10-23

How to Cite

[1]

R Terio and M Soriano, Bhattacharyya distance of shape feature histograms as aid to authenticate H. R. Ocampo paintings, Proceedings of the Samahang Pisika ng Pilipinas 31, SPP2013-PB-24 (2013). URL: https://proceedings.spp-online.org/article/view/SPP2013-PB-24.