<div class="twins"><p></p></div> <div class="header"><h1><a href="/" style="text-decoration: none"><img height="32px" src="/favicon.png"></a> Cloud Detection and Removal Techniques for Landsat 8</h1></div> <div class="story"><div class="item paragraph"><p>View Abstract (PDF) </p></div> <div class="item paragraph"><p>View Poster (PDF) </p></div> <div class="item paragraph"><p>If you are curious about how I did some of the Python scripting with ArcGIS' Arcpy API, here are a few Github gists:</p></div> <div class="item paragraph"><p>Cutting out the clouds in each image using BQA .</p></div> <div class="item paragraph"><p>Summing the binary masks to create classification groups that show how many images in our set are cloudless .</p></div> <div class="item paragraph"><p>Creating an RGB composite of bands 4,3,2 that have the clouds cut out and set as 0 .</p></div> <div class="item paragraph"><p>Averaging together all of the RGB values for each image based on the sum of valid pixels based on the BQA mask .</p></div></div>