Aerial cactus identification captsone project for Udacity Machine Learning Nanodegree 2019.
VIGIA project is a project based in Mexico developed to assist government led efforts to preserve natural areas. The traditional surveillance mechanisms have been insufficient to log and monitor the impacts of climate change and human activities on the flora and fauna. The goal of VIGIA project is to build an automatic surveillance of such protected areas. To build such a system various tools and technologies are used such as unmanned aircrafts (drones) equipped with a camera or cameras to take aerial imagery of the protected areas. The imagery can then be fed into a computer vision recognition system to distinguish flora/fauna, log the number of a certain protected species, etc. Specifically in this project/competition, the task of computer vision recognition system is to detect a certain species of cactus. To lead the way in protecting the flora (specifically columnar cactus) in protected natural areas, aerial imagery taken from unmanned aircraft can be fed into a system which utilizes state-of-the-art computer vision and machine learning methods to assist in the recognition task.
To install requirements
pip install -r requirements.txt
See jupyter notebook for the implementation details.