Evaluation of automatic cloud removal method for high elevation areas in Landsat 8 OLI images to improve environmental indexes computation

César I. Alvarez, Ana Teodoro, Alfonso Tierra

Research output: Contribution to conferencePaper

3 Scopus citations

Abstract

© 2017 SPIE. Thin clouds in the optical remote sensing data are frequent and in most of the cases don't allow to have a pure surface data in order to calculate some indexes as Normalized Difference Vegetation Index (NDVI). This paper aims to evaluate the Automatic Cloud Removal Method (ACRM) algorithm over a high elevation city like Quito (Ecuador), with an altitude of 2800 meters above sea level, where the clouds are presented all the year. The ACRM is an algorithm that considers a linear regression between each Landsat 8 OLI band and the Cirrus band using the slope obtained with the linear regression established. This algorithm was employed without any reference image or mask to try to remove the clouds. The results of the application of the ACRM algorithm over Quito didn't show a good performance. Therefore, was considered improving this algorithm using a different slope value data (ACMR Improved). After, the NDVI computation was compared with a reference NDVI MODIS data (MOD13Q1). The ACMR Improved algorithm had a successful result when compared with the original ACRM algorithm. In the future, this Improved ACRM algorithm needs to be tested in different regions of the world with different conditions to evaluate if the algorithm works successfully for all conditions.
Translated title of the contributionEvaluación del método de remoción automática de nubes para áreas de alta elevación en imágenes Landsat 8 OLI para mejorar el cálculo de índices ambientales.
Original languageEnglish (US)
DOIs
StatePublished - 1 Jan 2017
EventProceedings of SPIE - The International Society for Optical Engineering - , United States
Duration: 1 Jan 1981 → …

Conference

ConferenceProceedings of SPIE - The International Society for Optical Engineering
Country/TerritoryUnited States
Period1/01/81 → …

Fingerprint

Dive into the research topics of 'Evaluation of automatic cloud removal method for high elevation areas in Landsat 8 OLI images to improve environmental indexes computation'. Together they form a unique fingerprint.

Cite this