This project was created for my Electrical Engineering Undergraduate Honours Thesis at The University of Queensland which recieved a distinction. The aim of the project was to designed and build a semi-autonomous fertilizer distribution rover as an extension of the farmbot open source project. Feel free to view my Thesis Report.
The following technologies were used:
- Linux shell scripting was used to setup the environment on RasPi to run the python scripts controlling the rover
- Keras and TensorFlow were used to train CNN Deep Learning Networks for Image recognition to identify plants correctly before distributing the fertilizer
- OpenCV was used for preprocessing and transforming image data into Numpy arrays that could be analyzed by the Image recognition Deep Learning Model
- RasPi was interfaced with Arduino microcontroller running C++ code controlling the hardware modules e.g. motors, pumps, wifi.
Please Note more detailed setup is contained in the thesis report. That being said key installations required are:
- Tensorflow
- Keras
- OpenCV
- Numpy
- Raspi Zero W running Debian Linux
- Arduino Uno
This project is licensed under the MIT License
Inspiration and code snippets.
- boppreh for keyboard control python library