A simple contribution to spatial novelty detection in stereo-vision systems for mobile vehicles in indoor motion is presented by synthesizing a Cellular Nonlinear Network (CNN) for processing perspective images with light-dependent semantic content. Enhanced contrast images are generated by means of a fuzzy evaluation cellular subnet. Spatial novelty images can be subsequently obtained by cascading proper cellular subnets. Quality contrast evaluations are then performed on spatial novelty images using a dedicated cellular circuit. A test case is reported, to show how the suggested cellular system can provide useful information for stereo vision systems.
|Titolo:||On Spatial Novelty Detection via Image Contrast Enhancement using Cellular Nonlinear Networks|
|Data di pubblicazione:||2004|
|Appare nelle tipologie:||1.1 Articolo in rivista|