Automated Image Processing Module for Images Captured by Earth Observation Microsatellite Diwata-1 as Support for Ground Station Operations

Abstract

This paper shows the outline of an automated image processing module for the microsatellite Diwata-1 images. The downloaded images from Diwata-1 are automatically processed using thresholding and feature detection algorithms to determine the cloud cover percentage, land area percentage, and bodies of water percentage for each image. From these percentages, an image is classified as useable or not. The list of all useable images tagged by the automated image processing module is compiled and compared to the list of all usable images tagged by a manual operator. Results show that the automated process has an accuracy of 87% to 97% depending on the payload used for capturing the image.