Classification in High-resolution Aerial Imagery using Identification of Cloud

DOI

No DOI

Published In

Volume 1, Issue 4 (2021)

Published Online

July-August 2021

Pages

01-04

Article Type

Research Article

Check for updates

No DOI available for cross-reference

Cloudsandshadowsposesevereproblemsindiscernment of the scene and recognizing confirmation of articles in aerialphotography.Thechangesinillumination, ensuedbythepresence of cloud and the shadow,aresome of the reasons thatlead to ambiguity, while doing picture division leadingtodetectionoftargetedobjects.Conventionalmethodsareefficient in perceiving thick fogs in contrastive establishment, butperform ineffectually in the impression of slim fogs, different cloudsand their shadows. Reference pictures for the data are required inmostcases,andseparatealgorithmsarepursued,toidentifyclouds and shadows in an image, which most likely will not be reachable inall circumstances. Techniques used in this paperto recognize cloud andshadows,obviatingtheneedforreferenceimages,areimageenhancement,analysisofcolorhistogramofinputimages,adoptionofautomaticthresholdingandmathematicalmorphologyontheinputimage.Theproposedalgorithm,wasfoundtobefast,andexperimentedonvariousimagesthatcontainedmultiplewhitecloudclustersofdifferentshapes,thickness and their shadows. The algorithmwas endorsed with anaccuracy of 94.6% and 87.2% for unmistakable evidence of fogs andshadows, separately.