SpellPix

• 

  This is our hand written text input with spelling mistakes.

• 

  This is the output of backend with incorrect words listed and corrections suggested.

Inspiration

We noticed that there is no current software that is able to spell-check images. When we learned about the Adobe Challenge, we saw a possible blend of these two ideas. In Adobe, many people use graphics and PDFs and being able to proofread these elements can give individuals peace of mind. This can be applied to graph captions, written reports, posters and more.

What it does

Takes an image with text, extracts the text and checks the spelling of the text mapping the incorrect spelling to the correct spelling.

How we built it

We used the Adobe add-on API, Google Vision API, and a language database to see text in images, spell check it, and input it into Adobe.

Challenges we ran into

Combining the front end to the back end. We were not able to call the functions used for spell check and text extraction in the index.js file. We believe it is because of the Adobe SDK import. We tried to solve this bug for hours and tried many methods but we still were not successful.

Accomplishments that we're proud of

For the first time, we all used and called cloud APIs and configured all of the authentications. We were able to complete our backend and our front end. We were also able to figure out and create a great Adobe addon plugin.

What we learned

Learned all about Javascript, HTML CSS and Google Cloud API implementation. Created and used the adobe plugin development interface, Adobe CLI and SDK.

What's next for SpellPix

We would love to be able to combine our front end to our backend and get it to work on the Adobe add-ons. We also would like to be able to highlight the text that is incorrect and have a popup menu to suggest corrections that the user can click and change. On the image itself, progress has been made to highlight and box the words.

Built With

• css
• google-cloud
• google-vision
• html
• javascript
• node.js
• typo