ColourMe

Inspiration

Our inspiration for ColourMe stemmed from a desire to create a more accessible way for blind and colorblind individuals to understand the colors around them. Working with vision and color recognition technology seemed like an interesting challenge that we had not yet explored, and we saw an opportunity to create a useful tool that could improve the daily lives of people with visual impairments.

How we built it

Initially planned on using hardware such as color sensors, we moved away from the idea because it wasn't as accessible and would be difficult to carry around when, for example, shopping. Instead, we opted for our phone and decided on a web app over an app as it would make it easier for both Apple and Android users. As we were looking for color scanner APIs to help us, we stumbled upon Google Vision that gave us the most dominant color in a picture. Using that, we used OpenAI to interpret the data.

Challenges we ran into

Learning about APIs was definitely the most difficult thing. We learned so much about https protocols and how to implement APIs into web apps. At first, we were using node to reach the API, but it wouldn't connect to the HTML so we had to scavenge for different methods and finally fell onto fetch API.

Accomplishments that we're proud of

We are most proud of our attention to detail. Our primary goal was to make sure that the user experience would be as seamless as possible. For that, we tried to choose high contrasting colors and made sure that buttons are big enough for it to be easy to use.

What we learned

Our biggest takeaway from this project was taking more breaks.

What's next for ColourMe

We want to be able to add more settings to control the voice and/or stop it from talking, as there are also colorblind people that may not want the audio. We also want to include more languages. Using vision AI, we can also expand to object detection so that the user knows that the AI is looking at the correct object.

Built With

• css
• fetchapi
• google-cloud-vision
• html
• javascript
• openai
• visionapi