Combinations solve tough imaging problems
George Kotelly, Executive Editor
Off-the-shelf machine-vision and imaging hardware and software are often integrated with related products and technologies to solve demanding visual problems. In some situations, a related technology, such as x-ray, optics, or infrared, is used to obtain initial imaging data. Then, the machine-vision or imaging system takes over and processes, analyzes, and displays the resulting data. In other situations, special software is applied to ordinary imaging data to obtain specific details.
In a new approach to three-dimensional (3-D) digitizing, as described by senior engineer Michael Petrov, two companies have combined resources to incorporate scanning control and image-processing hardware and software into a single scanner design. This scanner integrates off-the-shelf cameras, structured lighting, custom image-processing software, and multiline laser scanning to obtain high visual and geometric object quality in its 3-D models (see p. 50).
To protect military facilities from intruders, a PC-based perimeter-security system combines television, IR, and imaging technologies, says contributing editor John Haystead. Cameras are mounted on a remotely controlled pan-and-tilt head that surveys a large security zone for intruder detection and identification. The system automatically classifies and displays the target (see p. 36).
The final nondestructive test of sealed automobile air-bag inflators by a combined x-ray and machine-vision workstation verifies the number, amount, positioning, and packaging of installed components. According to senior controls engineer David A.Wilson, an image intensifier converts the x-rays to digital images for scanning by a machine-vision system to ensure that each inflator is thoroughly checked (see p. 26).
A common deconvolution technique can achieve impressive deblurring results in images obtained by telescope-linked cameras, according to contributing editor Lawrence Curran. He describes a PC-based telescope that can clarify real-time, color-processed, filtered, and enhanced star-type images using a custom CCD camera and maximum-entropy software (see p. 42).
To boost the imaging capabilities of multiprocessing systems, reports editor at large Andy Wilson, the latest DSP boards are supporting both distributed and shared-memory architectures and several operating systems (see p. 56). Imaging-system designers can now implement a variety of topologies for DSP.
