Advanced Imaging


Advanced Imaging Magazine

Updated: January 12th, 2011 09:49 AM CDT

Component Integration: The Software Story

Faster algorithms, more usability, 3D and fitting better with multicore processors are just a few trends
Figure 1
Figure 2
Figure 3
Figure 4
Figure 5

By Barry Hochfelder

Because imaging and machine vision are multipurpose technologies that are widely used in industry, science, medicine and surveillance, the hunger for advancements is huge, says Lutz Kreutzer, Manager PR & Marketing for MVTec (Munich, Germany).

“For imaging software, it means developing faster algorithms and to better use current hardware environments, such as the benefits provided by multicore computers, as well as new matching technologies to robustly and reliably find objects or work pieces even in images with strong perspective distortions.”

Kreutzer adds that 3D vision methods and processing of extremely large images are becoming more important. Advancements also support identification applications, such as bar and data codes that are steadily increasing in speed and tolerance.

Bruce Tannenbaum, Image and Video Processing Application Marketing Manager of The MathWorks (Natick, Mass.), sees the impact multicore CPUs is having on software. “First, customers are expecting that software vendors take full advantage of the underlying hardware,” he says. “Second, it is becoming increasingly challenging for engineers to “roll their own” software for high performance as multi-threaded or parallel programming is much harder than traditional programming. Third, it changes expectations such that customers now expect their software to easily scale from multicore processors to clusters of machines.”

He also points out the trend to larger amounts of data. “With many imaging applications generating more than four gigabytes of data, it is now quite common to require a 64-bit addressing architecture. However, other aspects of the PC architecture are not keeping up with the rate of data expansion. Architectures in visualization tools, data access, and data processing that worked well in 32-bit applications are now less effective and new approaches are needed.”

Subscribe to our RSS Feeds