PubMed · 3380634
The source did not provide an abstract. Follow the original record for more information.
Explore connections, maps & timelines
F Boselie. 1988. Local versus global minima in visual pattern completion.. https://doi.org/10.3758/bf03207879
Cite the original work for its findings. Save a collection to share your selection of sources.
KEEP EXPLORING
A theoretical framework is proposed to understand binocular visual surface perception based on the idea of a mobile observer sampling images from random vantage points in space. Application of the generic sampling principle indicates that the visual system acts as if it were viewing surface layouts from generic not accidental vantage points. Through the observer's experience of optical sampling, which can be characterized geometrically, the visual system makes associative connections between images and surfaces, passively internalizing the conditional probabilities of image sampling from surfaces. This in turn enables the visual system to determine which surface a given image most strongly indicates. Thus, visual surface perception can be considered as inverse ecological optics based on learning through ecological optics. As such, it is formally equivalent to a degenerate form of Bayesian inference where prior probabilities are neglected.
Ptolemy (127-148 AD) studied physiological diplopia, correspondence and the horopter. He had all the data to build a theory of depth perception through disparity detection, but left that undone. Alhazen (1000 AD) associated depth perception with the sensation of binocular convergence, just as Kepler (1611) and Descartes (1637). With the development of the concept of retinal correspondence and the fusion of the retinal images in the brain (Huygens 1667, Newton 1704) a cerebral mechanism of disparity detection became thinkable. The rise of Empiricism (Molyneux' Premise, the case of Cheselden) postponed the solution of the problem, finally reached by Wheatstone (1838). Physiological proof of Wheatstone's theory came from the experiments of Barlow et al. (1967).
A new principle of grouping is proposed that is based on elements being located within a common region of space. Demonstrations analogous to Wertheimer's original displays show that this factor strongly influences perceived grouping and is capable of overcoming the effects of other powerful grouping factors such as proximity and similarity. Grouping by common region is further shown to depend on perceived depth relations, indicating that it is influenced by processes that occur after at least some depth perception has been achieved. Further demonstrations suggest that it is dominated by the smallest background area and that it can follow a hierarchical embedding scheme. It is argued that common region cannot be reduced to the effects of proximity, closure, or any other previously known factor and therefore constitutes a genuinely new principle of grouping.