Abstract

Subject of study. A physics-based expert model of initial features for the recognition of anthropogenic 3D objects in a monostatic laser location system is proposed. Method. The model is based on an intelligent analysis of data obtained using digital simulation modeling of temporal profiles of pulsed reflectance profiles of the object. Informative features are formed using the method of principal components. Main results. A cluster structure in the space of principal features is demonstrated and investigated. The parameters of several algorithms for the clustering and classification of 3D objects are identified using machine learning methods and their quality is tested in an informative space. Practical significance. The stages of solving the clustering and classification tasks for anthropogenic objects by a monostatic laser location system are described in chronological order.

© 2022 Optica Publishing Group

Distortion-tolerant 3D recognition of underwater objects using neural networks

Robert Schulein, Cuong Manh Do, and Bahram Javidi
J. Opt. Soc. Am. A 27(3) 461-468 (2010)

Fast and accurate 3D object recognition directly from digital holograms

Mozhdeh Seifi, Loic Denis, and Corinne Fournier
J. Opt. Soc. Am. A 30(11) 2216-2224 (2013)

3D object recognition through processing of 2D holograms

Behzad Bordbar, Haowen Zhou, and Partha P. Banerjee
Appl. Opt. 58(34) G197-G203 (2019)

Generalized Hough transform for 3D object recognition and visualization in integral imaging

Ariel Fernández, Julieta Umpiérrez, and Julia R. Alonso
J. Opt. Soc. Am. A 40(4) C37-C45 (2023)

Intelligent-assistant system for scleral spur location

J. E. Gómez-Correa, L. M. Torres-Treviño, E. Moragrega-Adame, M. Mayorquin-Ruiz, C. Villalobos-Ojeda, C. Velasco-Barona, and S. Chávez-Cerda
Appl. Opt. 59(10) 3026-3032 (2020)