Abstract

The generalized Hough transform is a well-established technique for detecting complex shapes in images containing noisy or missing data. We present an efficient optical implementation of this transform using an electrical lens with variable focal length and a rotating pupil mask matching the pattern to be found. The proposed setup works under fully (i.e., both spatially and temporally) incoherent illumination and can handle orientation changes or scale variations in the pattern. Validation experiments showing its real-time application are presented.

© 2015 Optical Society of America

Real-time pattern recognition using an optical generalized Hough transform

Ariel Fernández, Jorge L. Flores, Julia R. Alonso, and José A. Ferrari
Appl. Opt. 54(36) 10586-10591 (2015)

Fully invariant generalized Hough transform by out-of-focus multiview sensing with pupil array

Ariel Fernández, Juan M. Llaguno, Gastón García González, Facundo Lezama, and Julia R. Alonso
Appl. Opt. 58(28) 7766-7773 (2019)

Image segmentation by nonlinear filtering of optical Hough transform

Ariel Fernández, Jorge L. Flores, Julia R. Alonso, and José A. Ferrari
Appl. Opt. 55(13) 3632-3638 (2016)

Optical implementation of the Hough transform by a matrix of holograms

P. Ambs, Sing H. Lee, Q. Tian, and Y. Fainman
Appl. Opt. 25(22) 4039-4045 (1986)

Diffractive optical Hough transform implemented with phase singularities

Filippus Stefanus Roux
Appl. Opt. 33(14) 2955-2959 (1994)