With the amazing advances in image processing, computer vision has a unique position to help those without vision themselves. We wanted to create a companion and guide to serve as the blinds eyes as they adventure through the world.
What it does
Instead of a walking stick(or whatever its called) we offer a selfie stick, this can provide much more rich vision through the phone camera. By running the images gathered through some image processing with Microsoft APIs and OpenCv, we offer the blind an intelligent view of the world. They can identify the things in front of them and be warned when things approach a little too fast. With a friendly computer voice guiding them, the blind can better face the world and finally see the world we take for granted.
How we built it
We implemented Microsoft APIs and Open Cv to process images captured by a phone camera. We used Android Studio to create the front end which would connect to our server by AWS. We wrote an image processing function to try to notice the difference in movement in the frame according to the object to help perform our task
Challenges we ran into
We first tried to create our application using react native and swift ,but when we tried to integrate Microsoft's API's and our server errors continually persisted. React native also wouldn't allow us to use the camera function and we asked Facebook engineers for assistance ,but we could not find a solution. Swift would only receive information from server and not be able to transfer images. Time became a major factor towards the end because solutions to our problems were never found so we had to restart our front end.
Accomplishments that we're proud of
We are proud that after continuously having to shift our front end, we were able to find a platform to run our program smoothly for the most part. Having software that has the capability to assist the disabled is very much needed in our society today. Through image processing, we were able not only to analyze objects in the way, but also how close the object is during the path to warn the user.
What we learned
We learned the endless capabilities of image processing and that there are many roadblocks in utilizing this capabilities. While a function may be simple, a lot of thinking, frustration, and subtle gratitude goes into a single block of code. We have been tested and tried past our limits. Thank you.
What's next for Virgil
Virgil with more time and technology might be able to use regular camera's from phones to do 3-D image processing to be more accurate with our end goal to help blind people have the ability to "see". Also we believe we can use this technology also to help map out a way for anyone in real time to get somewhere. Virgil has the ability to maybe measure distances and warn about obstacles way better with an 3-D image processing capability.