The added problem of the synthesis of orientations for each 1-DE. As it turns out the processing of edge orientations is done in the cortex itself. Signals are passed from the retina down the optic fibre to the lateral geniculate body of the brain and thence to the visual cortex. Hubel and Weisel in 1963 discovered cells in the visual cortex that are responsive only to lines or edges of specific orientations (Fig.1). They narrowed down their findings to two main types: simple cortical cells and complex cortical cells. Each type of cell had the same specificity of orientation sensitivity as mentioned earlier but whereas the simple cells responded to orientations in specific locations on the retina the complex cells were not so discriminating. They continued firing even though a stimulus was moved, as long as the orientation angle was maintained.
Fig 1: shows a folded section of the visual cortex which contains cells that are responsive to edge orientation. It was found, by Hubel and Weisel that like cells form columns. One is led to suppose that the position of a cell in the column corresponds to a definite part of the visual field.
What is conjectured to happen during observation of a chiral field is that cells will light up at sequentially differing depths in each column of cells. Initially this is unremarkable, but since the illusion not only seems to spin but also creates a perception warped space as the eye travels around the image, subsequent firings of these cells reveal a pattern of activity that explains the warping. Imagine that initially the first column's topmost cell has lit up and as we go sequentially from column to column the cells that light up are deeper and deeper within their columns. Then the next lot of signals arrive and fire the next cell down in each column. A fluctuating plane of lit up cells will be created that travels in a wavelike fashion up and down that section of the cortex. This wave is the pulse of the warped space seen by the observer.
The image in Fig 2: shows a small section of a simple spacial layout.
The lines' angle changes in a regular pattern as we go down the column.
As he eye scans down the column a line of different inclination falls upon
the same part of the retina (presuming that each line is focussed upon
in turn). The cells of the cortex, however are not always the same ones
that fire since they are sensitive only to lines of specific angles. So
as the eye travels down the image cells are fired up in the cortex but
in different columns and possibly different depths.
RETINAL PROCESSING OF VISUAL IMAGES. Charles R. Michael Scientific American may 1969
Eye Brain and Vision; David Hubel
High-Level Vision : Object Recognition and Visual Cognition; Shimon Ullman
: A Computational Investigation into the Human Representation and Processing
of Visual Information; David Marr
Last Revised: 30/12/99