NSF Award
9626261
9626261 Ostrem This Small Business Innovation Research Phase II project will develop a novel OCR reader within the framework of an integrated technology aimed at providing ordinary people with the means to interact with computers in a more natural and productive way. The technologies required to make this possible include on-line hand-writing recognition, Optical Character Recognition (OCR), image processing, speech recognition, and the methodology for handling interactions between input modalities. This Phase II work will take a first important step in this direction - bringing on-line handwriting recognition and OCR into a common algorithmic framework. If successful, three significant benefits will result: 1) Humans will interact with computers in a more natural way; 2) The state-of-the-art of handwritten OCR readers will improve; and 3) Efficient new commercial products will be developed. Recently, the United States Post Office, a major user of advanced OCR systems, stated that "as significant advancements in OCR technology are needed to bring the read rate (of handwriting) to the desired level, innovative approaches are sought to achieve those advancements." From a commercial standpoint, a product combining OCR and on-line handwriting recognition in a common framework with shared resources will be smaller, more efficient, less costly, and easier to maintain than two separate products. Improved user-interfaces and OCR performance are of significant interest to major computer manufacturers because they will lead to increased hardware sales and new software products.
