Share Email Print

Optical Engineering

Object recognition with the hybrid evolutionary algorithm and response analysis in security applications
Format Member Price Non-Member Price
PDF $20.00 $25.00
cover GOOD NEWS! Your organization subscribes to the SPIE Digital Library. You may be able to download this paper for free. Check Access

Paper Abstract

The recognition of an object in a scene is a common and important task of electronic imaging arising in many defense and security applications. When the image of the sought object is significantly distorted, and the image of the scene is cluttered, noisy, and contains many objects, the commonly used methods based on correlation and comparison of the feature vectors of the images can show poor performance. The approach utilizing the particular model of the hybrid evolutionary algorithm based on image response analysis is proposed to solve the object recognition problem formulated as the global optimization problem. The computational experiments with two-dimensional grayscale images show that the proposed approach can solve complex object recognition problems. It is able to discriminate between objects having a high degree of similarity, and to detect the sought object in the large cluttered and multi-object scene.

Paper Details

Date Published: 1 October 2004
PDF: 11 pages
Opt. Eng. 43(10) doi: 10.1117/1.1790504
Published in: Optical Engineering Volume 43, Issue 10
Show Author Affiliations
Igor V. Maslov, Columbia Univ. (United States)
Izidor Gertner, CUNY/City College (United States)

© SPIE. Terms of Use
Back to Top