Imagine an image with some geometric shapes drawn on it - for instance curved contours, triangles, squares and the like. We can quickly and easily answer questions about the image by just looking at it, such as: is there any triangle inside the closed contour, are the red and blues squares connected, and the like. Answering a query of this type requires the application of a 'visual routine': a sequence of steps, applied to the image, and leading to the correct answer. The sequence changes from task to task, and may be composed, for instance, of finding a red shape, identifying its shape, following a contour to a neighboring shape, and so on. This illustrates a remarkable capacity of the brain's visual system, to quickly put together the analogue of a computer program and apply it to the image, during the analysis of properties and relations between components of the image. This capacity plays a major role in scene understanding, visually guided manipulation, and more abstract visual thinking.
A Weizmann institute scientist was the first to study the nature of
these visual routines composed by the brain, their elementary building
blocks and how they are put together.
Over the years, properties of visual routines have
been studied by using methods, including physiology, psychology, and
computer simulations, by different laboratories around the world.