Demonstration on Environmental Noice Removal Techniques For Raster Data Via User-Friendly Software Interfaces

Abstract

Environmental noise in the form of influences from cloud cover and/or smoke is often a critigal problem faced when deploying analyses using rernotely sensed infoimation. Such 'noise' will impact on the quality of the analysis. Hencg, it becomes necessary to remove such 'noise' in the pre-processing stages of the GIS-remote sensing analysis. whilst various noir"-r"n1oval methods have been developed by the purist image processing communiry a common problem faced by general practitioners in environmental science is that these methods are complex and very hard to comprehend and apply off-the-hand. Considering these difficulties, a simple yeicomprehensive methodolory is presented in this paper, where a tutorial spatial model using MS Excel is applied to explain the principle(s) behind spatial noise removal. using digitized data based on Landsat images for the Vavuniya district, the paper provides a detailed step-bystep discussion on how environmental noise could be removed using Boblean logic operators in"the MS Excel platform. Further, a detailed description of how this 'noise removal concept' can be transferred into the BtiOAS Imagine software platform for remotely sensed analysis is also given using the (original) Landsat image (ry) for Vavuniya district u, un "*urnple. Whilst this prototyping method is established as an excellent .\r"nu" for validation, refinement and improven;'ent of the analytical procedure, it is also stressed that future research in-these l."gu.d, should focus on developing software interfaces in advanced mathematical programming languages for enabling development of efficient and user-friendly tools for environmental noise removal'