Customization enhances imaging capabilities

Most machine-vision and imaging systems strive to produce high-quality images. Unfortunately, off-the-shelf vision-based hardware and software are not inherently robust enough to accommodate demanding imaging applications. To overcome that drawback, system integrators are designing highly specialized imaging-based equipment. For example, they have solved the highly complicated tasks of virtually simulating optical-vision systems, producing topographic maps of corneal surfaces, and incorporating

Customization enhances imaging capabilities

George Kotelly Executive Editor

georgek@pennwell.com

Most machine-vision and imaging systems strive to produce high-quality images. Unfortunately, off-the-shelf vision-based hardware and software are not inherently robust enough to accommodate demanding imaging applications. To overcome that drawback, system integrators are designing highly specialized imaging-based equipment. For example, they have solved the highly complicated tasks of virtually simulating optical-vision systems, producing topographic maps of corneal surfaces, and incorporating CMOS sensors and on-board processors into smart solid-state cameras.

Joining an optical, illumination, analysis, and design tool and a three- dimensional graphical design program has resulted in a software combination that can simulate and analyze almost any optical-vision system with high geometric and photometric accuracy. Designing computer models of optical-vision systems enables the visualization of all the key elements of the system`s optical performance. Moreover, say Michael Stevenson and Marie Côté of Breault Research Organization, these models serve as a method for reducing or even bypassing experimental design and prototyping.

To aid ophthalmic surgeons in correcting minor eyesight aberrations, a PC-based image-acquisition system integrates proprietary hardware and software to produce topographic maps of the front and rear surfaces of the cornea and the front surface of its inner lens. Special ray-tracing algorithms can then extrapolate the map of the rear surface on the inner lens. Based on these maps, says contributing editor R. Winn Hardin, the imaging system can develop a surgical plan for removing undesired tissue from the cornea.

Systems integrators expecting to find standardization in available smart cameras will be disappointed, says editor-at-large Andy Wilson. The problems are that some cameras integrate on-board processing, whereas others use digital-signal-processing boards, reduced-instruction-set-computing devices, or central-processing units, to perform on-camera image-processing functions. In addition, varied interfaces, such as RS-232, 1394 FireWire, and ProfiBus, and varied image sensors, such as linear, area, time-delay integration, logarithmic, or foveal, further complicate the smart-camera selection task. As a result, systems integrators must carefully evaluate all of these design choices.

More in Boards & Software