Our Project Story

Inspiration

Our inspiration was building a software that is budget friendly, enabling students to display their creative juices without the need of spending money on drawing tablets.

What it does

Allow students to view and convert real-life design prototypes to virtual prototype.

How we built it

We first started experimenting with object detection in OpenCV. We used the Viola Jones algorithm through the use of HAAR CASCADE Classifier for classification at first. Our project however required us to use some better means so we took to YOLO and Darknet for training a neural network that was trained on images of pen tips, which was a huge progress

Challenges we ran into

Object Detection had been the main problem that we had ran into while coding the software. Moreover, we also ran into problems while training the ML model becoz the images required to train the model were less and hence the model was less precise.

Accomplishments that we're proud of

We stepped into the field of neural networks with YOLO and DarkNet which had been a great beggining into the field of Deep Learning.

What we learned

Using Viola Jones algorithm which is very useful in all kinds of object detection.

What's next for CamScribbler

We really want to optimize project to make it more accessible for an average user and improve the machine learning model to improve the overall precision of the program.

Built With

• python