Windmill Bird Rescue (WBR)

Bird
Not bird

Inspiration

In North America 300,000 plus birds die annually to windmills, becoming a threat to energy and wildlife.

What it does

takes pictures, uses computer vision to analyze picture to see if it is bird or not, sound buzzer for time interval if bird predicted

How we built it

Python, arduino code, java

Challenges we ran into

Learning computer vision tensorflow, keras. Keras installation, Unsupported RaspberryPi , using 32 bit jdk, finding female female connectors

Accomplishments that we're proud of

The finished product.

What we learned

We went into this completely blind, we learned computer vision, arduino and python compatability, java development kits, arduino code for taking pictures

What's next for WBR

Optimizations, mobility, camera improvements, live video feed, image data base for further training models, higher accuracy in prediction

Built With

Submitted to

HackRPI

Created by

I worked on computer vision training for the bird prediction algorithm to create and train a neural network and model. In addition, I worked image file testing of adruino camera pictures.

Christopher Hu
I worked on the hardware and software of the arduino uno and arduino ov7670 to allow for pictures to be taken and analyzed

mattlevine68
I worked on Hardware with the circuitry of the OV7670 camera module, and also on Software with controlling the buzzer connected to the Arduino, with a Python script.

Kenechukwu Nwodo
Helped out with tensorflow installation/mplementation and attempted to use GPU optimization. Also helped with the buzzer set-up

Steven Cano

Updates

Christopher Hu started this project — Mar 17, 2019 10:51 AM EDT

Leave feedback in the comments!

Log in or sign up for Devpost to join the conversation.