Perceptual Shape Analysis: approaching geometric problems with elements of perception psychology

While the recent increase in computational power gave computers an irreplaceable spot in our lives, humans still have to rely on their innate capabilities for some tasks that no electronic device can outperform. Those processes strictly linked to survival tasks, like hearing, vision or social interaction are the fields where the human being shows an efficiency that no machine can compare to. Limiting the discussion to the field of shape analysis and recognition, this work takes inspiration from the most influential studies on human perception to demonstrate how our brains perform simple operations that allow us to unambiguously recognize a viewed object, while the current state-of-the-heart algorithms heavily rely on mathematical computations based on data, given by the shape primitives, that are of no use to a human observer. By mimicking human behaviour into the proposed algorithms, the author introduces the paradigm of Perceptual Shape Analysis, obtaining fast and efficient solutions for the main Shape Analysis problems while maintaining a complete independence on the object representation. This new approach is, in the author's opinion, an innovative and fresh point of view that could bring inspiration for the future evolution of the field.