After my company so generously gave me anApple iPhone for business purposes, I was only too intrigued to find out what could be accomplished with the device. Of course, my mind immediately strayed from the purposes I was given it for, and I was only too ready to download numerous free "apps" such as Shazam, Angry Birds, and a GPS navigation system. It was good fun while it lasted.
After tiring of these applications, I decided to investigate how products such as the iPhone were used in vision applications of more benefit to mankind. Here again, it seems I was not to be disappointed. Indeed, I easily found many other apps that allowed the iPhone to be transformed into a3-D scanner, measure radiation levels, read barcodes, and even turn the device into a low-cost microscope.
Needless to say, I was very impressed. Other applications that are now -- or soon may be -- available will allow the built-in camera to monitor a person's blood pressure by measuring facial features or monitor a baby's body motions and thus their heart rate as they lie in a crib. And all of this being performed by a device that costs less than my annual holiday to the South of France!
While the developers of such image-processing software tools are leveraging the power of off-the-shelf hardware found in computer products such as the iPhone andAndroid telephones, to accomplish the same goal in the machine-vision market is more difficult. That's because there are more than 1000 different cameras offered by over 100 different vendors that come with their own hardware, operating systems, and camera-to-computer interfaces.
And, although some vendors such asXimea are diligently working on creating smart cameras that will run multiple off-the-shelf machine-vision packages, much needs to be accomplished before machine-vision software can easily be ported across a variety of embedded vision platforms.
Clearly, developers of image-processing software for Apple and Android-based products have the advantage. By simply downloading a software development kit, they can develop an image-processing application that they will be assured will run on millions of pieces of similar hardware.
Sadly, the same cannot be said for the industrial vision marketplace. That's the reason, for example, why an industrial barcode reader is not available that can be used in multiple smart cameras. This is a great pity, since if such software could be used in multiple vision systems, it would allow software developers to maximize the sales of their applications -- thereby lowering the cost to the end user and encouraging other software developers to create novel end-user applications.
Unfortunately, do not expect this situation to change rapidly. While millions of Apple products may be sold each year, the same cannot be said for those products used in the machine-vision and image-processing market, limiting their uses for the few industrial operators prepared to invest in both hardware and software engineering.
Andy Wilson, Editor in Chief