Share
تاثير نسب ومعاملات الحزم على دقة تصنيف صور الاقمار الصناعية المتعددة الاطياف == Effect of Band Ratios and Indices on Classification Accuracy of Multispectral Satellite Images
Author name:
سيف كامل شنين
Supervisor name:
اياد عبد العزيز العاني | صلاح عبد الحميد صالح
General topic:
Physics
Specific topic:
Physics
Degree:
Master
University:
Al-Nahrain University - College Of Science - Physics Department
Language:
English
University location:
Baghdad
First pages:
26T1828 - p.pdf
Abstract:
Multispectral satellite images are holding whole information about the areaunder investigation. Many of digital processing techniques are used to extractmost of the possible information from these images, such these techniques arespectral band ratios, Tasseled Cap transformation and Principle ComponentAnalysis (PCA). These techniques are play good part for features extraction andreducing the number of spectral bands with no loses in information.The research aims to apply classifications on band ratios and image indiceswhich include PCA and Tasseled Cap transformations, of Enhanced ThematicMapper Plus (ETM+) satellite images, and to show how the accuracies of theclassification are differ from the accuracy of classification of the raw ETM+ bandsimages.The results show that the best bands ratios that adopted to represent fiveselected land features are 5/7, 4/5, 2/5, 1/4, and 2/4. Optimum Index Factormethod (OIF) has been adopted on raw bands combinations, and on the bandsratio combinations to select which band combination that contains muchinformation, where the results show that the best combination for raw bands is3 - 4 - 7, and best combination for ratios band is 4/5 - 1/4 - 2/5.Iterative Self Organizing Data Analysis (ISODATA) method has beenadopted as unsupervised classification, and Maximum Likelihood method assupervised classification. The overall accuracies are 87.18%, 89.45%, 89.26%,and 89.52% for raw bands, band ratios, Tasseled Cap, and PCA respectively,which showed that transformation techniques give good spectral enhancementsand feature extraction which are undistinguishable in raw image, and representgood tools to increase the accuracy of classification.
