Satellite Image Segmentation Photovoltaic Potential Estimation
Using Google Cloud Machine Learning Engine to built a model that help in estimating photovoltaic system design. An end-to-end python based deep learning application.
Tags: Artificial Intelligence, Computer Vision, Deep Learning & Artificial Intelligence, Machine Learning, Science
Scheduled on thursday 11:20 in room lounge
About the Author: • 10 years of experience in the solar industry (majorly Europe and East Africa) • 3 years of experience in Software Development and Artificial Intelligence • Presentation @ Geopython 2018 on Classification of Satellite Images • Masterthesis on the Estimation of the Potential of Roof Top Solar Systems in Luxemburg • Diplomathesis on Measurement and Simulation of a solar pumping station in Egypt
The used technologies are python based and include: MongoDB tensorflow Flask google.cloud python API
A dataset of labelled satellite images is created. Several networks are trained and tested on this dataset. The network is deployed on a production server.
The results of the classification/segmentaion are used to feed python based photovotlaic simulation libaries. The output is displayed and the results (the potential) evaluated.