/sun-eye

Machine Learning tools for hyperlocal weather prediction. :eye:

Primary LanguageJupyter Notebook

Sun Eye

Collection of Machine Learning algorithms and tools used for hyperlocal weather prediction

Each folder contains a different project

How to use

  • Install requirements.txt in each project before running with pip install -r requirements.txt

☁️ Feature Detection

Preview of Feature Detection

Analyze clouds and cloudiness from a stream of images or video.

  • Returns number of clouds and cloudiness factor in JSON format
  • Supports MP4 and static images (JPEG, PNG).
  • Hypnotizing visualizations during analysis

🍱 Mosaic Generator

Convert a time lapsed image collection

Preview of Feature Detection

Into a single image strip that contains the complete time lapse

Preview of Feature Detection

  • Used for training Convolutional Neural Nets
  • Mesmerizing images, put it on your wall. wow.

Algorithms

Caclculate cloud height from relative humidity and temperature

T_x = Temperature of x
h_clouds = height of clouds
T_dp = dew point
RH = relative humidity

eq-dewpoint

eq-height

Calculating width of viewport in meters

c = altitude of clouds (m)
z = width of viewport at clouds altitude (m)
alpha = focal view angle of camera (deg)

eq-viewport-width