CitationDownload as .RIS
Emerald Group Publishing Limited
Copyright © 2003, MCB UP Limited
About 18 years back, I was just leaving a restaurant having wined and dined a group of potential investors when one of them asked me in an unguarded moment "what is the most exciting thing you would like to be working on?"
Now when these chaps ask you a question it is a good idea to think on your feet and come up with something that sounds good. So I said "holographic imaging".
A long time has passed since then and as far as I have been able to find out, no one has actually made too much progress in this area. At least not in the way that I imagined or in the way that I still consider to be of great interest.
This issue has the themes "Machine Vision + Laser Scanners". Both are subjects in their own right, however, it is the combination of the two that is of particular interest to me.
Lasers are frequently associated with 3D ranging and measurement systems, however, in all the systems that I know, the sensor is either 0D (e.g. a single element photodiode), 1D (e.g. a CCD linear array or PSD), or 2D (e.g. a CCD or photodiode array).
How about a 3D vision sensor?
A crazy thought perhaps, but by no means impossible. I am imagining a sensor that is made up of lots of layers of 2D sensor arrays with "large" open gaps between pixels that can let light pass to the layers below. Or perhaps some other means whereby a sensor with actual physical depth can be interrogated and its contents processed.
Old computer memories used to look like this. Each board (we are talking before or in the very early days of the pcb) would comprise an array of horizontal and vertical wires each spaced a few millimetres apart. Where each vertical wire crossed a horizontal wire they would pass through a tiny ferrite ring. I am not quite sure how they worked but they basically read the magnetic properties of each ring and gave them a binary value. Stack a load of these boards side by side and you effectively had a 3D memory.
Is anyone working on a 3D sensor? And if so what are the intended applications? I have not worked out the details, but the concept sounds to me as though it might be just the thing for holographic imaging.
As I write, it is now looking as though the war in Iraq is drawing to a close. Whatever be your views on this war, or wars in general, I can only consider it beneficial as technology has allowed targets to be hit with generally high precision and greatly reduced civilian casualties.
The success in this area only serves, however, to highlight the problem of "Friendly Fire". While people are in the loop, mistakes, particularly in the heat of battle are inevitable. Surely machine vision and laser scanners can be used to insert an additional check between the pilot and the trigger button? Friendly fire has always been a problem and probably reached a peak in the Second World War, but these days we have the technology to help us eliminate it.