# Simple Geometries

Sept. 1, 2008
In 1884, Edwin Abbott wrote a satire: the science fiction novella Flatland: A Romance of Many Dimensions.

In 1884, Edwin Abbott wrote a satire: the science fiction novella Flatland: A Romance of Many Dimensions. His target was the social hierarchy of Victorian England, but his brilliant description of how we perceive dimensions has outlasted that culture. In the 2-D world of Flatland, a humble “Square” narrator guides readers through the implications of life in two dimensions. A visiting “Sphere” from Spaceland then introduces him to the third dimension, which leads to social disruption, repression, and revelation as the Square tries to show his fellow inhabitants a new reality.

I recall being entranced by the 1965 animated film Flatland, narrated by the late Dudley Moore. A more graphically dazzling version of the film was released in 2007—with excellent trailers available on YouTube.

Yet most engineers working with cameras and software for machine vision are essentially dealing with a Flatland-like world of images, where attributes such as shape, pattern, and value are usually captured and displayed in two dimensions. Fortunately for those who need the benefits of an additional dimension, the 3-D world of machine vision is both useful and non-threatening.

3-D machine vision systems can be used in numerous ways to guide automated equipment and robots. Our cover story, by contributing editor Winn Hardin, describes how a manufacturer of water pumps is using a 3-D vision-guided robotic workcell to unload bins full of stacked parts, reducing labor costs and improving productivity. Software provides dynamic information about part location, orientation, and the best ways to maneuver the robot and pick up a part.

3-D systems are also being used by the US Navy in a system to retrieve unmanned boats. An article by contributing editor Joyce Laird shows how two cameras located in enclosures about 8 ft apart provide 3-D images that are used to guide a robotic crane in its mission to anticipate, capture, and retrieve these unmanned boats. Given the random and sometimes violent movements of the sea, this is not an easy task.

Last, in one of his Technology Trends articles, editor Andy Wilson writes about a more futuristic 3-D system that may save lives on a battlefield or in a hostile environment. The Battlefield Extraction-Assist Robot in development by Vecna Technologies is humanoid in appearance and, so far, able to carry a 185-lb mannequin. The company is now seeking to evaluate different 3-D technologies that will allow the robot to autonomously traverse rough terrain.

The benefits and applications of 3-D in machine vision systems are already tangible and very promising. In Flatland, the Sphere banished the Square back to Flatland after he suggested that there may be other dimensions beyond 3-D. In the R&D world, many developers are already looking to incorporate multispectral and temporal image information into their systems, extending 3-D imaging concepts into multiple virtual dimensions. To avoid being banished to the past, developers of next-generation machine vision and imaging systems should encompass rather than reject these multidimensional advances.