Waltham, MA—Always looking for ways to explore state-of-the-art development approaches, AMC Bridge has researched the possibilities of applying a convolutional neural network (CNN) to formula reconstruction from an image. The result of the research is presented in the Machine Learning Formula Scanner technology demonstration.
ML Formula Scanner converts formula images to editable expressions and saves the conversion to a Microsoft Word document.
To implement ML Formula Scanner, AMC Bridge has developed an algorithm that uses CNN to recognize symbols in the uploaded images and reconstructs formulas from them. The CNN is trained to work with a wide variety of combinations of Greek and Latin letters, numbers, and special mathematical symbols. The algorithm, developed by AMC Bridge, processes the recognized symbols and reconstructs complex mathematical expressions, such as fractions and integrals.
To see the current functionality of the ML Formula Scanner technology demonstration, watch a short demo video.
If you are interested in learning more about the demonstrated technologies and how they can be utilized for your organization’s needs, please contact us to discuss the details.
About AMC Bridge
AMC Bridge is a vendor of choice for software development services in the areas of computer-aided design, engineering, manufacturing, and construction. Since 1999, we have been delivering solutions for CAD, CAE, CAM, PDM, BIM, and PLM applications. We have participated in the development of commercial software products and custom solutions for the engineering markets based on a variety of platforms, from desktop and web to mobile and cloud.
