About the Project

Hi! We present you all THE THIRD EYE aka T3E which is a prototype made to make a change in the lives of the blind people who after so much tech advancement still face problems. We are using Yolo V3 object detection Classifier to classify the surroundings and with the help of GTTS or Google's Text to Speech Conversion we aware the person about the forthcoming environment which will aware and alert the person wearing our prototype.

What Inspired Us?

So, living in a country where 40 million people including 1.6 million children are blind or visually impaired. And India is country which is home to world's 20% visually impaired people so this made us to make a difference in their lives in a positive way. So we came up with an idea to help them with the use of current advance tech to make them happy and to make their life much easier than ever.

What did we use?

We used AI THINKER ESP32 CAM as our source for real time video streaming which is attached to our prototype and then with the help of arduino we further made it work as a Camera Web Server where we embedded our code.

Challenges We Faced

So the main problems which we were facing was to stream the esp32cam on our code which took us so much hours because we were facing so many issues. Then came on making the hardware connection properly to embed in the prototype which also took us several hours to fix. But in the end everything was done with determination and successfully we could complete it. So there's no denying that our code would have worked much more efficiently since we the working of code required 16gb ram but we had 8gb and then CPU core requirement was hecta core but had quad core which resulted in lagging of the output video.

Future Goals

Being an AI enthusiasts and a tech geeks we look forward on making it much more user friendly for the blind wherein we could use much more cost effective cams and we can also use the distance measurement between the person wearing it and the object in the surrounding which will be more useful. We look forward to change their lives and make them more happy. To make our prototype reach the audience and make it look more good on real specs. To add Gesture Recognition as a feature.

Working ---> How to proceed ?

So using the arduino ide we streamed the esp32 cam as shown in the video and and then coding the main part i.e to classify the objects in the surroundings and then converting the text part to speech using gtts api. One must be sure to use correct ip addresses for the stream.


Built With

Share this project: