IntelliCane

Inspiration

Our inspiration for the project stems from our experience with elderly and visually impaired people, and understanding that there is an imminent need for a solution that integrates AI to bring a new level of convenience and safety to modern day navigation tools.

What it does

IntelliCane firstly employs an ultrasonic sensor to identify any object, person, or thing within a 2 meter range and when that happens, a piezo buzzer alarm alerts the user. Simultaneously, a camera identifies the object infront of the user and provides them with voice feedback identifying what is infront of them.

How we built it

The project firstly employs an ultrasonic sensor to identify an object, person or thing close by. Then the piezo buzzer is turned on and alerts the user. Then the Picamera that is on the raspberrypi 5 identifies the object. We have employed a CNN algorithm to train the data and improve the accuracy of identifying the objects. From there this data is transferred to a text-to-speech function which provides voice feedback describing the object infront of them. The project was built on YOLO V8 platform.

Challenges we ran into

We ran into multiple problems during our project. For instance, we initially tried to used tensorflow, however due to the incompatibility of our version of python with the raspberrypi 5, we switched to the YOLO V8 Platform.

Accomplishments that we're proud of

There are many accomplishments we are proud of, such as successfully creating an the ultrasonic-piezo buzzer system for the arduino, and successfully mounting everything onto the PVC Pipe. However, we are most proud of developing a CNN algorithm that accurately identifies objects and provides voice feedback identifying the object that is infront of the user.