Imaging fulfills inspection needs

Among its numerous operational advantages, imaging technology excels in noncontact, nondestructive, and noninvasive inspection applications. It has proven to be highly effective in lowering costs, saving time, reducing labor needs, handling fast task speeds, and enabling automation. In addition, these imaging attributes can be focused across a spectrum of detection and identification uses.

Oct 1st, 1999

Imaging fulfills inspection needs

George Kotelly

Executive Editor

georgek@pennwell.com

Among its numerous operational advantages, imaging technology excels in noncontact, nondestructive, and noninvasive inspection applications. It has proven to be highly effective in lowering costs, saving time, reducing labor needs, handling fast task speeds, and enabling automation. In addition, these imaging attributes can be focused across a spectrum of detection and identification uses.

In the past, inspection procedures for the food-packaging industry have been time-consuming, costly, nonautomated, labor-intensive, and destructive. To overcome these problems simultaneously, an on-line food-spoilage detection system has been developed that uses magnetic-resonance imaging. This system, says contributing editor John Haystead, performs noncontact, nondestructive inspection of food packages at fast production rates.

With the increased processing power of computer technology and improved x-ray tubes and detector assemblies, computed-tomography (CT) systems have advanced human-body imaging. A novel CT system approach, says contributing editor R. Winn Hardin, obtains four x-ray transmission intensity profiles per target slice while simultaneously increasing the number of slices per scan.

Despite computing advances, image-processing systems still fall short of emulating human vision capabilities. However, based on new single-chip enhancements, researchers have built integrated circuits that mimic the neurobiological functions related to human vision. As spotlighted by editor at large Andy Wilson, neuromorphic imaging sensors are combining photoreceptor arrays with analog circuits at each pixel to emulate the human retina.

Because many image-processing tasks are computationally intensive, they are often executed by repetitive multiply/accumulators mounted on DSP boards. To match or exceed the processing performance of general-purpose personal computers, says Andy Wilson in this month`s Product Focus, image-processing-board vendors are incorporating multiple-instruction multiple-data microprocessors, very-long-instruction-word devices, and combined RISC/DSP processors.

As an added service to our readers, we have started a new section on our Web site (www.vision-systems-design.com) called Back to Basics. Three unabridged articles are now available. Your comments and contributions are requested.

More in Scientific & Industrial Research