Skip to content

Galvanize G83 Data Science Immersive Phoenix -- Capstone

Notifications You must be signed in to change notification settings

tjslezak/capstone

Repository files navigation

Geomapper - DSI Capstone

Galvanize G83 Data Science Immersive Phoenix -- Capstone

Overview

This repo contains code in support of my Galvanize Data Science Immersive capstone project. The project uses multispectral satellite imagery of the state of Arizona acquired by ESA's Sentinel-2 mission.

Satellite Imagery

AZSen2

Method

Using a "Geology" band combination of 2-4-12 and training labels consisting of the State's Geologic map units trained a neural network architecture model backend of TensorFlow DeepLab's Xception_65 using the Rastervision library to construct the experiment. The project attempts to predict the geology using semantic segmentation, or pixel-wise classification, from satellite imagery.

Test Image

Test_img

Predicted

Test_pred

Actual

Test_actual

Results

A model was trained using Sentinel 2 Satellite Imagery for the state of Arizona and the State Geologic Map. Model training took approximately 23.8 hours on a P3.2XLarge AWS EC2 Instance using a NVIDIA Tesla V100 GPU though a Docker container and acheived a recall = .71, precision = .73, and F1 score = .71.

One-Page Summary Sheet

One_page

About

Galvanize G83 Data Science Immersive Phoenix -- Capstone

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published