Grease Gallery (Schneider Prize)

Slide 5
Slide 2
Slide 1
Slide 3
Slide 4

Inspiration

Inspired by the sponsored hack, to build an app that can be run on a Raspberry Pi that can analyze an image in order to read text from it. The images are a standard format with color and statistical information on samples of grease from industrial machinery.

What it does

Webapp where an image can be uploaded, in order to analyze text and store it as useful information in a database. The app includes an analysis page with the ability to display some information about the data. Since the app is essentially a website, it can run on a browser on any Raspberry Pi, Windows PC, or Mac.

How we built it

The programming was done using typescript and next.js framework. The website and databased are built and deployed using Vercel. The database uses postgreSQL. Text is extracted from an image using the Optical Character Recognition (OCR) library tesseract.js.

Challenges we ran into

Finding a library capable of OCR, and implementing it. Accessing information from the database to use on frontend elements for analysis. Bringing together frontend and backend components of the project.

Accomplishments that we're proud of

Having a full fledge website up and running.

What we learned

A lot about writing SQL queries and doing web development.

What's next for Grease Gallery (Schneider Prize)

First, functionality to analyze a subset of all of the data uploaded to the webapp. The frontend is set up for this, but we ran out of time to complete this feature. Second, some validation for OCR, or a user-friendly to edit data that may not be converted accurately. Finally, adding images to the database, ideally a 'thumbnail' that is a cropped version of the grease sample colors and an uncropped version that includes the data in the image, possibly for comparison to what is recognized by the OCR.