SBIR-STTR Award

Current morphometry and image analysis systems can be broadly divided into those which display an image and require the user to trace the points or lines considered significant, and those which apply complex algorithms to locate significant objects autonomously. We see a need for an intermediate type of system in which the user could rapidly indicate the approximate location of significant objects and allow the system to apply a simpler restricted edge finding algorithm to determine the exact line to use. Such a system would be much faster to use than "dumb" morphometry systems and much less costly to obtain or use than "intelligent" image analyzers which currently cost hundreds of thousands of dollars. By using a new VLSI microprocessor specialized for signal processing tasks we believe we can design a system of this type to sell for $15,000 to $20,000. This would bring powerful image processing technology within the reach of many more investigators.

Current morphometry and image analysis systems can be broadly divided into those which display an image and require the user to trace the points or lines considered significant and those which apply complex algorithms to locate significant objects autonomously. We have developed a prototype of an intermediate type of system in which the user can rapidly indicate the approximate location of significant objects after which the system applies an edge following algorithm to determine the exact boundary of the object. This system will be much faster to use than "dumb" morphometry systems and much less costly than "intelligent" image analyzers costing hundreds of thousands of dollars. This system uses a new VLSI microprocessor specialized for signal processing tasks and innovative image processing circuitry we have designed. We believe we can design a system of this type to sell for approximately $10,000. This would bring powerful image processing technology within the reach of many more investigators.