One of our member's sister works in the consultancy business, and as a low-level job, she was given the task of taking handwritten forms, and transferring it to a readable, editable file such as a .docx. She then told our colleague about how much time she and the company would save if there were some application that could scan these handwritten forms, and convert them to editable files. We then decided to solve this problem for her, and others like her in similar positions and industries.
What it does
The desktop application takes an uploaded image of your physical text, such as notes, assignments, forms and problem sets. It then outputs a .doc file of the information, available for download.
How we built it
We built this application using OpenCV for image recognition, ApachePOI for creating the .doc files, JavaFX for the GUI and Git for maintenance of code between the team.
Challenges we ran into
Using and importing libraries, as well as getting them to function properly, such as the ApachePOI classes was one of the most difficult aspect of developing a deployable solution. Managing versions of repositories and transferring code between team members was also quite a hitch, and was much more difficult than we anticipated.
Accomplishments that we're proud of
Creating a consistent framework between the collaborators, strategizing a complete solution plan, creating a model which can understand the human handwriting, as well as a backend that can process and create an editable file.
What we learned
We learned about the importance of using the right tools from the start, having a well-planned workflow/pipeline, as well as how necessary StackOverflow is!!!
What's next for Hotdocs
Integrating mathematical symbols, graphs, diagrams, and the formatting of all of these together on a page in a cohesive/aesthetically pleasing manner.