Making Handwritten Materials Accessible

Handwritten materials pose unique accessibility challenges because, when shared digitally, they are generally turned into a single image rather than text. This means that the content is impossible for someone using a screen reader to access and can't be digitally manipulated to enhance readability. It can be time-consuming to make those materials fully accessible, but this guide provides some tips for making the process faster and easier.

Consider your Workflow

There are numerous ways to approach this task, and the optimal choice will depend on factors like your discipline, the nature and complexity of your course material, and your lecture style. Moving forward, it may help significantly if you test out some of the different methods on this page and make adjustments to the way you take notes to make the process easier over time. For creating digitized notes, thinking carefully about how you structure your notetaking will help make the process easier. For example, taking those notes directly in PPT and using the built-in Ink-to-Text feature may be more efficient than taking notes in other formats and trying to remediate the resulting PDF. If you would like to talk to someone to talk about what may be optimal for your personal workflow, you can contact the Instructional Designer assigned to your unit.

Digitizing Handwritten Notes

A Note About This Section

Below, we provide guidance on using free AI tools to speed up the process of making handwritten note files more accessible. This is a work in progress and this section will be updated as we discover new tools that work better than what we currently recommend.

What it Takes for Handwritten Notes Files to be Accessible

Making handwritten notes accessible requires all of the following:

All text has been accurately digitized.
All drawn figures and diagrams have alt text. They should also have long descriptions where appropriate.
All equations are accessible. This generally requires additional steps beyond transcribing the text to ensure the equations are read correctly.

Doing this work manually can be a challenging process, but there are AI tools that can be helpful. The degree of accuracy from the AI tools is dependent on a number of factors including your handwriting quality and the complexity of the notes. They tend to do a lot better with cleanly organized pure text than with more complex materials. At present, we haven't found any freely available tools that allow you to upload a full slide deck and get back a completely accurate text file, but here are some procedures that might help.

Before you Start Working

If you have a lot of handwritten notes to make accessible, this process can be a lot of work. Before you start making your files accessible, you may want to consider whether the entire set of notes need to be made available to students. Sometimes it may be sufficient to make only part of the notes available to students (for example, a clear outline including main headings, topics, and definitions along with key figures and equations). This approach may be particularly important if you create a new set of notes every semester instead of re-using them across courses.

Option 1: Working Slide by Slide

It may seem more time consuming to make notes accessible in small chunks, but we've found that AI tools tend to produce better output when working on small pieces at a time. If you have a single set of handwritten notes that you re-use across semesters, this process is likely to produce higher quality initial output that requires less editing.

Open up an AI tool that allows file uploads. There are a number of AI tools freely available for this process. In our testing, both Google Gemini and ChatGPT provide reasonably good output.
Take a screenshot of the first slide you want to have transcribed.
Upload that screenshot file into the AI tool along with a prompt. For example, you might say "Please transcribe the text in this image. For all equations, use LaTeX formatting. For any images or diagrams, please provide both alt text and long descriptions."
The resulting output will be pure text, so you'll need to copy & paste into something like a Word or PPT document. Before using the output, be sure to edit for accuracy.
For any images or diagrams, you'll need to copy from the original file into the new file and add the alt text and long description.

Option 2: Working from a Full Slide Deck

Free AI tools (and even the paid version of ChatGPT) seem to struggle with accurately transcribing long presentations. These steps are likely to produce the best outcome, but the output will still require significant editing.

Ensuring the document is in PDF format.
Open the file in Adobe and run OCR so the handwriting is turned into selectable text.
- To do that, open the All Tools menu, select Scan & OCR, under Recognize Text select "In this file", ensure "All Pages" is selected, then click "Recognize text"
Open up an AI tool that allows file uploads. There are a number of AI tools freely available for this process. In our testing, both Google Gemini and ChatGPT provide okay output.
Upload the PDF into the AI tool along with a prompt. For example, you might say "Please transcribe all text in this file. For all equations, use LaTeX formatting. For any images or diagrams, please provide both alt text and long descriptions. Arrange the output slide-by-slide formatted as closely as possible to the original file"
The resulting output will be pure text, so you'll need to copy & paste into something like a Word or PPT document. Before using the output, be sure to edit for accuracy.
- For those familiar with using LaTeX, you might try telling the AI to use LaTeX formatting for everything, then re-create the slide deck using the Beamer document class.
For any images or diagrams, you'll need to copy from the original file into the new file and add the alt text and long description.

Option 3: Using PowerPoint's 'Ink to Text' Feature

If you use PowerPoint and take notes directly on individual slides using a digital pen or stylus, your handwriting can be turned into text using the 'Ink to Text' feature within PowerPoint. It does a relatively good job of transcribing text, although it tends to separate the text across text boxes that will need to be combined & edited. Also, you will still need to add alt text and long descriptions for images as well as taking additional measures to ensure the accessibility of your equations. The 'Ink to Text' feature often automatically pops up as a lightning bolt icon when you click on an area that has handwritten text. If you need to use it manually, you can select the area of interest using the lasso tool, then select 'Ink to Text from the Draw tab.