Visual Perception Flashcards
(15 cards)
Gestalt Perception
- Human vision is biased to perceive structure
- Whole shapes, figures and objects
Gestalt
* German for “unified whole”
* Essence or shape of an entity’s complete form
* Pattern, configuration
Gestalt Principles
- Gestalt principles are laws of human perception
- How we find order in what we see
- How we recognize groups, relationships and patterns
- How we see individual elements as a whole
- How we simplify complex images when we perceive objects
Perception of Objects: Emergence
- “Seeing the big picture”
- Perceiving the whole
without having to analyse
the individual part
Perception of Objects: Closure
- Seeing the complete shape even when only parts are visible
- The mind fills in the gaps
Perception of Objects: Continuity
- Grouping elements that
follow the same path - Seeing a continuous shape even if partly occluded
Perception of Groups: Proximity
- Objects that are closer
together are perceived as a group
Perception of Groups: Similarity
Objects that share visual
characteristics are
perceived as grouped
Perception of Groups: Common Region
Objects that are within the same region are perceived as one group
- In many guidelines, this
principle is called
“Enclosure”
Perception of Image Structure:
Figure/Ground
- To simplify an image, the brain separates foreground (“figure”)
from background (“ground”) - Important, as more attention is
given to the foreground
preattentive processing
Preattentive processing of visual information is performed
automatically on the entire visual field
* quickly, effortlessly and in parallel
* without focussing visual attention
Pop-out features
- Pop-out features (also called out preattentive features) are visual properties that
can be perceived without focussed attention
Pop-out features for data visualisation
- Pop-out features help communicate
information efficiently - Some features pop out more than others
- Colour is stronger than shape
- Motion is effective
(but can be annoying)
Visualisation: Basics
In visualisation, we represent different types of data:
Categorical: Named labels (e.g., colors, types).
Ordinal: Ordered values (e.g., rankings).
Quantitative: Numeric values with measurable magnitude.
We use perceptual channels like position, size, color, brightness, and shape to visually represent this data.
Visual Encoding: Colour Hue and Value
Value (lightness) is good for showing order, so it’s useful for ordinal data, but less effective for continuous variables. We perceive contrast better than exact values.
Colour (hue) is unordered, making it suitable for categorical (nominal) data.
Rainbow color scales are common but not ideal, as they can mislead or confuse interpretation.