Optical Character Recognition : Jacob Rabinow

Profits from consulting work at Rabinow Engineering were largely funneled into explorations in the emerging field of optical character recognition (OCR). Rabinow’s vigorous research into OCR had begun during his days at the National Bureau of Standards, when he developed a reading machine.

Vital to the successful operation of the reading machine was the “best match” principle. The machine employed this principle by projecting the positive image of each character produced by a portable typewriter onto the negative image of all alphanumeric characters. The black area of each character was projected onto transparent areas on a disc of photographic film. Behind the film was a disc scanner and behind that was a photocell. If the dark character image matched a transparent character image, the amount of light reached the photocell was small. After each positive was projected onto all negatives, photocell readings were compared and the best match was determined. This machine is now in the collection of the Smithsonian Institution.

Most of Rabinow’s subsequent OCR machines, including those using resistor networks tot recognize characters, operated on the “best match” principle. Rabinow’s OCR clients included the U.S. Army Signal Corps, the Bank of America, the Ryder Truck Co., Remington-Rand, RCA, and Almex. In 1964, Rabinow sold Rabinow Engineering, which by then had 100 employees, to Control Data Corporation but continued to do OCR research and development as a Vice President of Control Data.

Objects :
Optical Character Recognition Scanner, ca. 1954-1972
Special Font & Optical Character Recognition Machine Immune to Over-Printing and Over-Writing, ca. 1972