Event-related potential maps to perceptual (stimulus type) and cognitive (stimulus relevance) manipulations were studied in 12 healthy volunteers using 42-channel mapping. Perceptual manipulation used three types of visual stimuli: rectangles constituted by: (1) contrast; (2) different densities of monocular Dynamic Random Dots (Flat DRD); and (3) different binocular disparities of Dynamic Random Dots (Stereo DRD). Cognitive manipulation within each stimulus type consisted of presenting the rectangles horizontally and vertically, one of the two with a probability of 33%, and requesting the subjects to count and thus attend to the 'rare' rectangles. Spatial characteristics of the maps were analyzed; this allowed conclusions about the generating sources. The map series were adaptively segmented using the minima points of the grand mean Global Field Power curve. Segment strength (Global Field Power) and segment landscape (locations of extreme potentials) were assessed. Stimulus type had effects from 78 to 310 ms, stimulus relevance was effective from 210 to 1000 ms. In the 78-174 ms segment, Stereo DRD and Flat DRD stimuli produced similar map landscapes, while contrast stimuli produced different map landscapes. Attended and ignored stimuli produced contrary effects on landscapes at 210-310 ms as compared to those at 310-546 ms, indicative of different neural populations activated by attention processes during these late event-related potential segments. Interaction between perceptual and cognitive manipulation occurred at 210-310 ms when perceiving stereo stimuli and attending to relevant monocular visible stimuli produced similar map landscapes, suggesting a common brain resource during this segment for automatic figure perception and voluntary attention. The observed functional differences of the segments contribute to the identification of global functional microstates of brain electric activity.