# Project export: Iridisee

This document was generated by HackStack to give an AI agent context about a hackathon project. Sections are labeled with their provenance; content marked as truncated was cut to keep this document small.

## Project metadata

- Hackathon: Cal Hacks 11.0
- Tagline: Lightweight video feed overlay to aid the colorblind, targeting AR platforms
- Devpost: https://devpost.com/software/iridisee
- GitHub: not linked
- Team: contributor stats unavailable

## Devpost submission (written by the team)

### Inspiration

I had read about AR overlays giving subtitles for the deaf in normal conversations and wondered if this technology could be used to aid any other disability. Being colorblindness myself, to find the answer all I had to do was look around.

### What it does

Generates an overlay for video streams to label specific color groups a user has difficulty differentiating. The web version allows users to snap a picture to be analyzed and labeled.

### How we built it

We used python and the openCV library to generate color contours and process the video data. The portable version is hosted on a website using reflex to enable greater access to the utility.

### Challenges we ran into

The main challenge of developing a CV algorithm looking for color groups is fine tuning the search for accuracy in different lighting conditions and preventing double classifications. We had to abbreviate the algorithm and optimize it for web deployment.

### Accomplishments we're proud of

We were able develop the underlying routine in a few hours and it gave us ample time to tune and port Iridisee for different platforms.

### What we learned

Computer Vision is extremely flexible and easy to use. Seeing it first hand opened a whole new world of possible applications in the AR space. Web development is not to be underestimated. The project was first written as a standalone application, writing it for the intended platforms from the beginning would have saved a lot of time.

### What's next

The image processing software is currently 2 dimensional only in a 3D world. For true integration with AR, generated masks will be specially consistent with the space by processing two different image feeds and resolving the label placement to a single point. This technology lends itself to mobile development which we don’t not get around to.

## README (from the GitHub repository)

No README available.

## Detected evidence (automated analysis)

No repository was indexed for this project. Claimed technologies below could not be checked against code.
- Python (language) — claimed on Devpost, not found in the code

## Codebase structure

No repository index available.

## Key source files

No repository index available; no source files included.