Object Detection and Segmentation through Geo-Spatial data (GE)
Sujay Kumar
Data Modelling Analyst
Data Scientist
ML Engineer
Python
PyTorch
React
This identifies and locates objects in an image using UNet Architecture. The goal is to accurately and efficiently identify and track objects to improve safety, enhance productivity, and enable new capabilities in industries such as surveillance, autonomous vehicles, and robotics.
This project focuses on rooftop segmentation using Google Earth (GE) imagery. Leveraging geo-spatial data, Tensorflow, and Google Cloud Platform (GCP), the project aims to facilitate advanced drone delivery methodologies and conduct comprehensive data analysis research.
Table of Contents
Introduction
**In an era of increasing urbanization, optimizing the utilization of spaces becomes crucial. This project, "Rooftop Segmentation using Google Earth (GE)," tackles this challenge by harnessing the power of geo-spatial data, TensorFlow, and the Google Cloud Platform (GCP) to pave the way for advanced drone delivery methodologies and comprehensive data analysis research. With the rise of urban populations and limited available space, the efficient use of rooftops becomes pivotal. By leveraging the capabilities of Google Earth imagery, we aim to identify and segment rooftops accurately. The goal is to provide insights for potential utilization, such as drone delivery drop-off points or urban gardening spaces. To achieve this, we employ deep learning techniques and sophisticated data analysis.
**
Features
Utilizes Google Cloud Platform (GCP) for scalable processing.
Applies geo-spatial data and Tensorflow for advanced segmentation.
Dataset sourced from LandStat resources.
Employes the UNet architecture for robust rooftop segmentation.
Usage
Set up your Google Cloud Platform credentials for access to GCP services.
Run the rooftop segmentation script.
