Share Email Print

Proceedings Paper

Review of filtering methods in mobile vision from ground vehicles in low light conditions
Author(s): Cassiano Paixao Pais; Fernando D. Carvalho; Fernando Carvalho Rodrigues; Mateus da Silva
Format Member Price Non-Member Price
PDF $17.00 $21.00

Paper Abstract

We discuss the application of nonlinear median, alpha-trimmed mean filters and moving average to images from an intensified CCD camera which is mounted inside a vehicle on the dashboard. Discrete filtering strategies based on alpha-trimmed mean as a prefilter to prepare an image to automatic moving average has been done. In fact, the smoothing characteristics of the order statistics filters and the integrated value of the moving average have been combined to produce a good-looking image to the human eye. In some cases, an object shape isn't completely defined in the new image after the filters have been applied. In these cases resulting from poor acquisition due to subexposition or intensive shading, a histogram modification has been effected. The intensified camera system performs at light levels as low as 0.0001 lux, equivalent to an overcast moonless night. The regular CCD cameras operate in light levels of around 0.1 lux, scene illumination at least equivalent to the light on a full moonlit night. We show, through a series of examples, that these filters can be effectively used to partially eliminate veiling glare effect and brightness fluctuation with time, which comprise the most significant noise in an intensifier tube application.

Paper Details

Date Published: 14 February 1992
PDF: 9 pages
Proc. SPIE 1613, Mobile Robots VI, (14 February 1992);
Show Author Affiliations
Cassiano Paixao Pais, LNETI-Optica (Portugal)
Fernando D. Carvalho, LNETI-Optica (Portugal)
Fernando Carvalho Rodrigues, LNETI-Optica (Portugal)
Mateus da Silva, Ministerio da Defesa Nacional (Portugal)

Published in SPIE Proceedings Vol. 1613:
Mobile Robots VI
William J. Wolfe; Wendell H. Chun, Editor(s)

© SPIE. Terms of Use
Back to Top
Sign in to read the full article
Create a free SPIE account to get access to
premium articles and original research
Forgot your username?