OCR Systems

OCR Systems, Inc.

Company type	Private
Industry	Optical character recognition technology
Founded	1969 (1969) in Bensalem Township, Pennsylvania, United States
Founder	Theodor Herzl Levine
Defunct	July 1992 (1992-07)
Fate	Acquired by Adobe Inc.
Headquarters	Bryn Athyn, Pennsylvania , United States
Area served	International
Key people	Gregory Boleslavsky (Vice President), Vadim Brikman (Vice President)
Products	System 1000, ReadRight software
Number of employees	30+ (1988)

OCR Systems, Inc. was a U.S.-based company specializing in both hardware and software solutions for optical character recognition (OCR) technology. Established in 1969 by entrepreneur Theodor Herzl Levine in Bensalem Township, Pennsylvania, the firm introduced its inaugural product, the System 1000, in 1970. This device gained widespread adoption among prominent corporations such as Texas Instruments, AT&T, Ricoh, Panasonic, and Canon for applications including bill processing and mail sorting. Although the company faced early challenges—such as losing a distribution agreement and the bankruptcy of a Canadian partner—it rebounded under the strategic guidance of vice presidents Gregory Boleslavsky and Vadim Brikman, who had recently emigrated from Soviet Ukraine.

During the late 1980s, OCR Systems broadened its offerings by launching ReadRight, an OCR software initially created for DOS platforms and subsequently adapted for Windows. The company expanded its market presence through key distribution partnerships with firms like Japan’s Chinon Industries and scanner manufacturers including Canon, Hewlett-Packard, and Skyworld. By 1988, OCR Systems employed over 30 staff members and reported revenues reaching $3 million.

In the early 1990s, as competition intensified from software companies such as Toronto-based Delrina, OCR Systems pursued potential acquisition deals. A proposed stock exchange agreement with Delrina fell through in early 1992. Later that year, Adobe Inc. acquired OCR Systems in July for an undisclosed amount, incorporating its OCR technologies into Adobe’s growing product portfolio.

History

Founded by Theodor Herzl Levine (circa 1923–2005), OCR Systems was born out of Levine’s extensive experience in electrical engineering and academia, holding degrees from the University of Pennsylvania and having taught at multiple universities. Levine’s ambition to advance OCR technology followed his work at Philco during the 1960s.

A pivotal moment came in 1973 when the System 1000 was distributed through a partnership with 3M, significantly boosting OCR Systems’ footprint within the data entry sector. Nonetheless, financial difficulties persisted throughout the late 1970s and early 1980s. When Levine became ill and faced financial struggles around 1983, executives Boleslavsky and Brikman assumed leadership roles, driving growth through product development and expanded market reach.

Under their direction, OCR Systems secured international clients such as Allegheny Energy and Belgium’s postal service while enhancing its workforce with talented professionals from Eastern Europe and Asia. To accommodate this expansion, the company moved its headquarters to Masons Mill Business Park in Bryn Athyn.

The 1987 alliance with Chinon Industries allowed OCR Systems to bundle ReadRight software with scanners, strengthening its competitive position. Further collaborations with top scanner manufacturers reinforced its standing within the industry.

Despite these achievements, OCR Systems faced stiff competition in the rapidly evolving OCR software market by the early 1990s. The unsuccessful acquisition attempt by Delrina highlighted these challenges before Adobe’s purchase ultimately integrated OCR Systems’ technology into a larger corporate entity.

System 1000

The System 1000 centered on the 16-bit Varian Data 620/i minicomputer, equipped with 4 KB of core memory, which managed functions such as paper feeding, document layout, optical character recognition (OCR), error checking, sequencing, and output processing. Initially designed to identify 1,428 OCR fonts—including those used by Selectrics—IBM 407 print, full OCR-A and OCR-B sets, Farrington 7B characters, as well as optical marks and handwritten digits, it offered the possibility of expanding font compatibility upon request. Memory capacity could be increased in 4 KB steps up to a maximum of 32 KB, allowing support for general typewriter and photocopier font types.

The System comprised document transport mechanisms, one or multiple scanners, a CRT monitor, and a Teletype Model 33 or 35. Documents were advanced using a friction-based rubber belt. It supported scanning up to three lines per page and accommodated document sizes ranging from small utility cash stubs measuring 2.6 by 3.5 inches to ANSI letter size (8.5 by 11 inches) and stock certificates sized at 8.5 by 12 inches. Initial processing speeds reached up to 420 documents per minute per transport unit, with later enhancements increasing throughput to as many as 500 documents per minute depending on the size and complexity of the material.

A notable capability was the Document Referral feature, which notified operators about fields that could not be read by displaying blank spaces on the CRT screen for manual keyboard input. Output options included IBM 7- or 9-track magnetic tape, paper tape, punched cards, various mass storage devices, or direct connection to System/360 mainframes for subsequent data handling.

Offered at a purchase price of $69,000 or available for lease at $1,800 monthly over three years or $1,600 monthly over five years, Computerworld highlighted the System 1000 as costing less than half compared to rival systems while delivering enhanced functionality and user-friendly operation. Competitors in this segment included Recognition Equipment Retina, Scan-Optics IC/20, and Scan-Data 250/350.

ReadRight

ReadRight employed a topographical approach to letter recognition by decomposing scanned characters into constituent elements such as strokes, curves, ascenders, and descenders within a hierarchical tree structure to determine character codes. Entirely software-driven without requiring an expansion card, its final DOS iteration (version 2.01) operated within under 640 KB of RAM. In 1991, OCR Systems released a Windows-exclusive version 3.0 while continuing distribution of version 2.01. Later that year in October, they introduced ReadRight Personal tailored for handheld scanners on Windows platforms featuring real-time word updates as lines were scanned. By 1992, ReadRight transitioned fully to Windows with release of version 3.1.

Bundled alongside Canon’s IX-12F flatbed scanner, ReadRight version 2.0 received an Editor’s Choice award from PC Magazine in 1989. Nonetheless, reviewers pointed out difficulties in accurately interpreting proportional fonts such as Helvetica and Times New Roman. The subsequent version 2.01 improved font detection capabilities and was commended for its user-friendliness and minimal system resource requirements but faced challenges handling uneven column widths and embedded graphics which necessitated manual exclusion of images by users.

Version 3.1 garnered mixed reactions within business circles: it was lauded for affordability and rich features but criticized for only average accuracy and speed relative to competitors like Calera’s WordScan Plus. It struggled with correcting errors that went unnoticed initially, contained an imperfect spellchecker, and had trouble processing small-sized text or multicolumn formats with narrow margins. Additionally, the software frequently misclassified graphics as textual content. Although OCR Systems planned an update in 1992 aimed at resolving these issues, it was never launched due to Adobe’s acquisition of the company.

This product series exemplifies OCR Systems’ targeted efforts to advance optical character recognition technology amid competitive market pressures; however, inherent limitations affected both its market penetration and financial outcomes before the acquisition occurred.

References

• </refend}}