Optical Encoding of 3D Correlation for Color Pattern Recognition

Nicolás, Josep; Campos, Juan, Univ. Autònoma de Barcelona

doi:10.1117/3.2265060.ch13

Book: The Art and Science of Holography: A Tribute to Emmett Leith and Yuri Denisyuk

Editor(s): H. John Caulfield

Author(s): Josep Nicolás, Juan Campos

Published: 2004

https://doi.org/10.1117/3.2265060.ch13

Author Affiliations +

Abstract

The spectral distribution of the different points of a color image contains information that often is relevant for recognition purposes. For example, objects that cannot be distinguished in black-and-white images are easily discriminated when their color distribution is considered. Said another way, the problem of recognition is greatly improved when one considers the spectral distributions of the different points of the image. In addition, the color distribution of images sometimes contains information on the chemical composition of the objects present in the image. Accordingly, in recent years a variety of proposals have been made to introduce the color information of images in pattern recognition tasks using optical correlation. Additionally, some works on filter design for color pattern recognition have been proposed for color image-processing techniques. Some are the generalization of well-known filters for monochromatic images, but others are specifically designed for color pattern recognition.