Inspiration
Today, about 350 million people in the world struggle with colorblindness, but there are not many measures taken to accommodate for these individuals.
What it does
This is where ColorID comes in. With this app, a user can use their Android device to take a picture of the environment and determine the true color of given object.
How we built it
To build this application, we had a team of four Rice computer scientists, and we operated mainly on Android Studio.
Challenges that we ran into
We struggled the most with obtaining accurate results for identifying the colors. We attempted to minimize this error by taking the minimum squared error of the detected RGB value with each of the colors available for identification. Also, we displayed the three closest colors to the actual pixel value, and we deemed these three colors as being the three most likely colors of the actual user-perceived color. However, we did face some hardware limitations. For example, sometimes we would take a picture of something that was obviously purple, but when the camera took the picture, this object would look completely blue in it, and the application deemed it as being blue.
Accomplishments that we're proud of
We are proud of having created a working product. While there is further work and development to be done, the application is a good reflection on what we can accomplish given the amount of time.
What we learned
First and foremost, we all learned a lot about Android programming. We also learned a lot about the various representations for colors in a digital format and how to compare colors using vectors.
What's next for ColorID
In the future, our team will continue to develop and improve ColorID, hoping to improve performance and accuracy. We will also add more features to the application, allowing users to specify certain regions that they would like to identify, as well as providing a more detailed analysis of the picture's chromatic composition.
Log in or sign up for Devpost to join the conversation.