Methods to Extract Annotations from Engieering Drawings and/or Maps

Design information for most of America's infrastructure is held as paper blueprints and paper specifications. It is estimated that 10 billion sheets of paper drawings exist for state and federal projects alone. Yet to perform necessary modern engineering analysis on these drawings, the blueprints must be in electronic form, not paper form. Present methods to convert these paper drawing to digital form are slow, expensive, and error prone. Making an electronic copy is 40 to 90 times more expensive than making a paper copy, a cost of $800 billion to $1.8 trillion dollars to convert all existing drawings to electronic formats. This research will develop new frameworks and algorithms to reduce the cost of converting a paper blueprint to electronic format. These new methods will reduce costs by a factor of ten and will improve the quality of the finished conversion. The research will combine recent advances in scanner technology, advances in algorithm technology, and government advances in modeling languages to make these improvements possible. Anticipated Benefits/Potential Applications - This research will provide an inexpensive, accurate, and seamless method to capture the original semantic intent of the paper blueprints, and will be an enabling technology for improving existing products and structures.