Optical Character Recognition (OCR)
Optical Character Recognition (OCR) is that branch of computer science concerned with optically reading text from various written sources such as books and signs. OCR algorithms translate an image consisting of symbols, letters and numbers into an electronic computer file. For example, an OCR system allows the user to scan a book or a magazine article and have that image translated into a word processor document which can then be stored, edited, etc. The usefulness of this technology in, for example, the legal profession, in security ID and in postal sorting is obvious. Advanced OCR systems are able to interpret text in a wide variety of fonts, but handwritten text remains problematic.
StockerYale supplies the high-brightness, high-uniformity illuminators required for OCR. The illuminator is a highly sensitive component of the overall OCR system. In particular, it must be carefully designed to enhance contrast. The optimum illumination wavelength depends on the chemical makeup and on the color of the symbols and background, and on the spectral sensitivity of the detector. In order of perform OCR on moving targets, the illumination must be pulsed or sufficiently bright to allow for rapid camera shuttering.
One of StockerYale's advantages for some OCR applications is that our chip-on-board technology allows us to place the LEDs very close to the surface that is being read. The chip-on-board approach involves an array of closely spaced (sub-millimeter) light-emitting dice on a printed circuit board. Such illuminators are inherently highly compact. Since so little physical space is taken up by the module, it can be positioned almost anywhere, i.e. at a very short distance from the text surface. This closeness allows for extreme brightness. These high levels of brightness and uniformity would not be possible if one were to use an illuminator comprising a host of individually lensed LEDs.